Governance and Agility in Product Development Organizations
Graham Oakes Ltd., United Kingdom
Agile42, Finland
Abstract
In a globalized, technology-driven world, consumer and citizen expectations are changing rapidly. Likewise, the capabilities of potential partners and competitors are constantly shifting and realigning. These trends place increasing pressure on organizations to adjust the way they develop new products and services. They must evolve their product portfolios rapidly, while constantly seeking feedback from customers and partners. This may involve substantial changes to organizational processes and boundaries, e.g. opening up innovation processes through techniques such as crowdsourcing. Above all, it requires organizations to deal with uncertainty and complexity. And that requires them to deal with governance, for most strategies for dealing with uncertainty and complexity involve redistributing power within the organization. Governance sets the terms for allocation and legitimate use of power.
This chapter explores the role of governance in product innovation. It examines the benefits of clear governance, both to organizational effectiveness and to the wider societal good. It discusses people’s perceptions of governance. It describes common governance models and the trade-offs they entail. And it looks at these questions from the perspective of agility: how does effective governance support an organization to create and learn from change while rapidly creating value for the customer?
The authors take the perspective of practitioners working within product innovation organizations. They present their observations of the challenges that such organizations are facing (or perceive themselves to be facing), then attempt to place these observations into a theoretical context by reference to models of complexity and agility. They then explore the implications of this thinking for effective governance of product innovation.
Keywords: accountability, agility, change, complexity, crowdsourcing, Cynefin, decision making, governance, open innovation, product development, uncertainty.
Introduction
Most organizations, whether public, private or not-for-profit, exist in order to deliver a suite of products and services to customers, citizens or other stakeholders. As the world transforms towards an information society, customers have more information on hand and can make better buying decisions. Companies still compete on price and quality, but differentiation through product and service innovation is an increasingly important element of their continued survival. Such innovation is often characterized as a complex, emergent process, especially in domains that deal with fast-paced change, rapidly evolving competition and “wicked” problems.
Consider, for example, the case of AB Corporation1:
AB operates mobile phone networks in most European countries, along with the all the necessary supporting infrastructure—service centres to provide round-the-clock support to its customers, retail chains to sell handsets and related services, e-commerce systems, etc.
AB constantly needs to reconfigure each of these elements in response to a rapidly changing marketplace. For example, rapid shifts in mobile phone technologies require it to invest heavily in networks. Its service plans must be updated regularly to accommodate shifts in the way people use their phones, new entrants into the marketplace, etc. Partnerships with device manufacturers, retailers, app developers and other partners form and reform as the market shifts. Revenue from “legacy” sources (voice and texting) declines while compensating revenue from data services grows less quickly.
Amongst all this change, the central product development unit decides to develop a new service. It works with a device manufacturer in East Asia to develop new handsets and to tailor their operating system for AB’s customers. It acquires several European app developers to develop customised applications for the handsets. It builds a large in-house team to develop “cloud” services to add value to the applications. The overall target is to provide an environment that will retain more customers in a highly competitive market, and then enable AB to sell more, higher value services to each customer.
Unfortunately, the project to develop the handsets is delayed due to communications problems between the European and Asian teams. The “cloud” service receives poor feedback from initial user testing in several countries. The different app developers cannot agree on a common approach to deal with concerns such as customer data storage and privacy. AB’s operating arms in several European countries refuse to support the new service, claiming that it does not address significant features required for their markets. In markets where it is introduced, customer service operators do not receive adequate training, so cannot sell and support the service effectively. The overall programme is closed down, and rarely mentioned again within AB Corporation.
Forces for Change
This case study is not atypical. In our experience, as practitioners working within product innovation teams in a variety of contexts, organizations around the globe are experiencing (or perceive themselves to be experiencing) rapid shifts in their “marketplace”—whether a commercial marketplace, the citizenry to which a government provides services, or the stakeholders supported by not-for-profit organizations—due to a large number of interconnected factors, the primary factor being technological change.
Digital and communications technologies are evolving rapidly, creating new opportunities in the marketplace while rendering existing products and services obsolete. Changes in one area frequently cause cascading changes in other areas. For example, emergence of “cloud” technologies has required new software development approaches and toolsets, and created new affordances for user interface design, while also raising substantial concerns in areas such as security and privacy. It is difficult for organizations to understand the impact of such changes and hence to direct an appropriate response.
This technological change has given rise to information-driven economies. Information-centred products offer new ways to create value—customers are prepared to pay for services, for personalisation, for “experiences” as well as, or instead of, physical artefacts. Organizations build further value by connecting and integrating rather than by manufacturing. Much of this value comes from design rather than from capital-intensive manufacturing and logistics, allowing small, nimble organizations to compete on an even footing with larger ones.
Advances in communications technology and medical science interact with demographic shifts. Populations in the developed world are aging, creating requirements for new services in domains such as health care, while requiring fresh attention to be paid to concerns such as accessibility. At the same time, younger generations are bringing very different experiences and expectations to the marketplace, e.g. for “always-on” availability of services and support, and for “any time, any where” access to information. These shifts empower some people, but create complications for the organization(s) delivering the services. These complications can in turn lead to added pressure being placed on staff, e.g. to work longer hours or unsocial shift patterns.
Coupled with technological change has been a rapidly increasing demand in people’s expectations as to what constitutes “good” service. Customers and citizens expect services to be delivered with ever increasing levels of speed, availability and reliability, with improved user experience, in a personalised way, and at ever decreasing cost. Again, this creates challenges for the organizations providing such services: how can they deliver such high levels of service while also delivering acceptable performance to shareholders, members or taxpayers, and while respecting the needs of staff and suppliers?
In parallel, global shifts in supply chains and manufacturing capabilities are taking place. Improved communications and transport technologies have made it easier to coordinate global supply chains, while many countries have built advanced capabilities in product design and manufacturing, service delivery, etc. These capabilities have co-evolved with customer expectations—customers expect more because they see that their suppliers can do more, and suppliers improve their capabilities because they can see customer demand for better products and services—but they also create dramatic power shifts in the supply chain. Centralised, top-down direction is rarely feasible in such complex supply chains, and neither is it acceptable to suppliers that have developed advanced capabilities in their own right. And again, questions of the rights of vulnerable staff within the supply chain arise.
Demographic changes and improved customer analysis methods give rise to new markets. Growing sophistication in analysis of customer data allows organizations to identify ever more micro-segments. This fragments the product portfolio, and raises questions about how accountability is partitioned across these often-overlapping segments. At the same time, rising wealth in nations such as India, China and Brazil is focusing attention on “bottom of the pyramid” customers, which central product development teams are poorly placed to understand. Overall, this combination of broadening and fragmenting markets means that it is no longer a case of people in “advanced” facilities telling those in “less advanced” ones what to do, but more one of sharing ideas and learning across peers. It is not clear that traditional corporate governance structures are best suited to such peer-to-peer organizations.
Furthermore, the capabilities of new competitors and partners are growing. Many organizations have been founded on the premise that they contain sufficient in-house skills and resources to succeed through tight internal control of product development and delivery. This premise no longer holds. Many trends are increasing the range and capability of potential partners and competitors. For example, advanced information and communication technologies enable small companies and networks of companies to develop new products in very short timespans. To compete effectively, organizations need to work with such emerging organizations, integrating them into their innovation teams and processes. Again, this creates complexities for decision-making and coordination—who within the web of partners has power to make critical decisions; who else must they involve in making these decisions; etc?
In all, this adds up to an impetus to address ever more complex problems. Many of the niches for simple products have already been filled—organizations must address ever more complex problems if they are to find a niche for themselves. At the same time, customers prefer conceptually simple solutions that are easy to use and fit neatly into their busy lifestyles. Organizations are left with the conundrum of finding simple solutions to complex problems. This is often resolved by hiding product complexity behind a simple façade, a feat which requires the organization to manage complexity exceedingly well.
Coping With Change
In order to thrive in these new marketplaces, organizations must constantly extend and evolve the suite of products and services that they provide to their customers. They typically do this through some combination of the strategies outlined in this section.
Organizations may attempt to produce a range of product variants and extensions, each targeted on a small micro-segment of their customer base. If such variants can be produced rapidly and without complicating and slowing down development of the overall product catalogue, they provide a powerful way to compete in a market that is fragmenting and evolving rapidly. Likewise, organizations may try to capture more value from their customer base by creating value-added services to support their product range. As well as increasing profitability, this can help meet rising customer expectations, but again at the risk of complicating the product portfolio and associated questions of ownership, accountability and decision-making.
Many companies are attempting to increase the pace at which they deliver new and improved products. Formerly, companies could spend several years on developing a new product version — recall Microsoft Windows 95, 98 and 2000? Today, a large number of organizations have moved from multi-year projects through regular yearly releases to a quarterly or bi-monthly rhythm, an order of magnitude faster. In the extreme, many web-based service providers now aim to deliver updates to their sites several times per day. This allows them to test new products and hence respond rapidly to emerging customer preferences, to experiment with and learn about new technologies and partners , etc. And again, it creates many questions for control, accountability and decision-making—how is product quality assured in the face of such rapid evolution; who prioritises objectives and resources across multiple parallel streams of development; etc?
Likewise, many organizations are trying to bring customers and partners into the innovation process, for example, using “crowdsourcing” to work with citizens and customers to clarify their needs and co-create new services that fill those needs. Thus LEGO is actively nurturing “AFoL” or Adult Fan of LEGO communities, from which it can source ideas for new and innovative designs. Or organizations may bring specialist expertise into their innovation process through methods ranging from setting challenges onto “open innovation” networks to forming more conventional partnerships with specialist firms. These approaches help the organization increase understanding of its customers and expand the range of expertise it has available to deliver new products. Use of lightweight processes such as crowdsourcing and open innovation challenges can also increase the speed with which an organization responds to shifts in markets and technologies. And again, it raises questions of ownership of intellectual property, accountability, the locus of decision-making, etc.
Organizations are also doing all they can to seek rapid feedback from stakeholders and customers. Many companies no longer rely on six- or twelve-month product release cycles, but instead aim to test customer expectations then implement and market new products within a few months. Deploying small incremental changes allows them to collect early data about sales and usage, and hence to validate assumptions before committing significant resources. This also enables them to drastically cut their inventory of “work in progress”, enabling rapid turn-around and reducing operating risk significantly.
To enable the above changes, organizations are adopting new product innovation processes using thinking drawn from domains such as lean product development and agile software development. This typically involves devolving some elements of decision-making to small, self-organising teams, on the assumption that this will increase the pace and diversity of product innovation, allow teams to get closer to specific customers, and make it easier to integrate specialist suppliers into the process. However, “empowering” small teams also raises questions about the boundaries of responsibility and accountability (which decisions belong to the team, and which to managers and executives overseeing them?), setting and coordinating priorities across the organization, allocating scarce resources, and developing and maintaining “corporate” assets (e.g. brand, product line architecture). The plethora of “agile” processes that has been described also raises questions about who chooses which process to adopt, and how to coordinate across different teams if they each choose their own process.
These are all strategies for dealing with uncertainty and complexity—a larger product portfolio gives more chances to find the “sweet spot”, as does more rapid introduction of new products; including partners and customers in processes extends the range of skills and perspectives that can be applied to any problem; devolving decision making to small teams enables rapid response to new information as it arrives. The overall trend is to adopt strategies that enable an organization to extend the range of information which it can gather from the marketplace and increase the pace at which it can respond to this information. These are the keys to surviving in an ever more dynamic environment. (“The Lean Startup”, Reis (2011), with its emphasis on “pivot points”, is a good example of this focus on rapid acquisition and response to information, as is the general shift towards “agile” software development, discussed below.)
Such strategies also change the power structure within the organization. Approval structures built for annual product releases cannot cope with monthly, weekly or daily product increments, so approval power must be devolved to new structures. As revenue shifts from products to services, so power shifts from manufacturing arms to service delivery teams. Each alliance requires some elements of decision making to be shared with new partners. Open innovation processes raise questions about ownership of intellectual property, accountability for product defects, etc. Empowering the team implies disempowering product and project managers, and so on.
Thus effective governance is at the core of product and service innovation. Organizations cannot improve the way they innovate without rethinking they way they govern themselves. The balance of this chapter will look at this question.
Challenges for Governance of Product Innovation
Of course, rethinking the way an organization governs itself is hard. People are rarely keen to give up power they have accrued for themselves within existing structures, and many people may be reluctant to take on unfamiliar responsibilities and accountabilities. Establishing governance structures to support product and service innovation processes runs into many challenges.
For a start, many advocates of “agile” approaches take a negative attitude to governance, disputing the very need for it. Equating governance with the imposition of bureaucracy, top-down controls and a culture of compliance, they reject attempts to consider governance and focus their attention on other aspects of change. In doing so, they increase the likelihood that inappropriate governance structures will emerge as an uncontrolled side effect of change. The resulting uncertainty around decision-making and responsibility can reduce the efficacy of the organization significantly.
We also find that different members and units of an organization can have very different perceptions of governance. For example companies that are changing rapidly may find that their previous centralised governance structures are no longer adequate, but have little time to think about new ones. The old structures remain in place while mid-level managers define their own local governance structures to get things done. This makes it difficult for people to understand where decisions are made: governance becomes fragmented and opaque, a source of contention rather than clarity.
Adding further confusion, professional bodies with an interest in product and service innovation promote a wide variety of governance models. For example, project management bodies such as the PMI (Project Management Institute) and APM (Association for Project Management) promote forms of governance that emphasize “top down” decision-making by steering groups and project managers. Other groups place greater emphasis on self-organization and “bottom up” decision-making. There are few objective sources discussing the pros and cons of these various forms, leaving practitioners in a quandary as to which form will work best in their circumstances.
In addition to the complexities mentioned above, organizations have increasingly porous boundaries. Rather than control innovation entirely within their own boundaries, they often form partnerships with other organizations for multi-organizational innovation. This leads to a variety of cross-organizational forms—joint ventures, public private partnerships, special purpose vehicles for specific projects, etc.—that can be difficult to govern, especially when the partners bring very different cultures and decision-making styles to the relationship. The rise of crowdsourcing, open innovation and co-creation with customers further complicates questions about the locus of decision-making and the ownership of intellectual property.
Even without the above complications, choosing an appropriate governance model involves complex trade-offs. No decision-making structure gives a perfect combination of speed, flexibility, context awareness, organizational consistency, etc. Centralised control structures, for example, help ensure that common standards are applied across the organization, but also create long chains of command that slow down decision-making and separate decision-makers from information on what is happening “on the ground”. Conversely, devolving decision making to small teams enables them to be responsive to local concerns, but makes it hard to apply consistent priorities across the organization. Organizations often invest a lot of effort trying to find the right balance, when the best model is probably to maintain a dynamic balance: shifting the locus of decision making as different decision attributes come into prominence. Again, this raises questions about who decides who has the authority to make any specific decision.
Of course, changes to power structures and decision making processes associated with product and service innovation will run into all the challenges associated with managing organizational change— without a clear impetus, people will question the need for change; without clear communication, different people will interpret objectives and ways of doing things differently; without appropriate training, people will execute new processes poorly; and so on. When something as fundamental as governance is being addressed, these changes will happen in a highly emotive and political environment, exacerbating the challenges.
This is because changes to governance are inherently about changes to power structures. People will be reluctant to relinquish their existing power, resulting in politics and power struggles. They may also be reluctant to take on new responsibilities and accountabilities, especially if commensurate resources and rewards are not forthcoming. So they will resist change. On the other hand, such struggles are already going on around product and service innovation in many organizations—the drivers identified above (pace of technological change, changing customer expectations, etc) are already triggering changes that affect the established power structures. An organization is more likely to steer these changes in a beneficial direction if it addresses the governance issues head on.
Yet organizations are often prone to cultural patterns that preclude effective discussion of governance. Low levels of confidence and trust drive many of these patterns. For example, managers may attempt to micromanage and pre-empt the decisions of specialist staff when they lack confidence in the capabilities and motivation of those staff. Those experts may then lose confidence and withdraw from decision-making (e.g. by deferring decisions to their managers), reinforcing their managers’ suspicions. Low trust levels can also lead to information hoarding (people hold onto information to protect themselves) and over-analysis (people spend a lot of time seeking “perfect” answers, even in complex situations where high levels of uncertainty and change favour experimental over analytical decision-making styles). All of these behaviours are antithetical to effective governance of product innovation, which requires high degrees of information and expertise sharing, and rapid assimilation of and response to new information.
Some of these challenges are definitional—without a clear definition of governance, people cannot address it effectively. Some of them are due to the difficulty of defining an appropriate governance model in complex, rapidly changing environments. And some of them are due to the inherent difficulty of managing organizational change. (Which, of course, itself involves aspects of governance—who, for example, is empowered to determine the direction of change?)
Definitions
Governance
The Institute on Governance (www.iog.ca) defines governance as follows:
Governance determines who has power, who makes decisions, how other players make their voice heard and how account is rendered.
This emphasizes four aspects of governance:
1. Governance is inherently about power—how it is distributed within the organization, and how it is used to drive behaviours and achieve organizational goals. This encompasses the various types of power that may come into play: the power people hold by virtue of their position in the organization, the resources they control, their expertise, their experience, their charisma, their personal strength, etc.
2. Power is exercised through decisions. Powerful people are able to influence what objectives the organization chooses to focus on, how it chooses to allocate resources in pursuit of these objectives, what structures, processes and systems it chooses to set up, and so on. Good governance ensures that these decisions are legitimate—that the appropriate people are involved in making each decision, and that “due process” is followed in the course of making the decision. It also ensures that decisions are made in an efficient way, and that the organization focuses an appropriate level of attention and resources on each decision—for example, that important decisions receive more attention and due diligence than trivial decisions.
3. Good governance attends to the concerns of all stakeholders. It ensures that the voice of all stakeholders is heard and that they can influence decisions in an appropriate way.
4. Good governance also ensures that people are held to account for the decisions they make. Although many people interpret “accountability” as being synonymous with “blame” and hence punishment, we interpret it as being closer to “feedback”: good governance ensures that the organization tracks the outcomes of decisions and acts to modify them where they are not having the desired effect. (Of course, where individuals consistently make bad or illegitimate decisions this may result in punishment, but that is only one way to improve the outcomes of decision-making, and not always the best.)
We also like one of the Institute on Governance’s earlier definitions:
Governance is the process whereby societies or organizations make important decisions, determine whom they involve and how they render account.
This definition emphasises the importance of decision-making. Good governance identifies which decisions are important. It then ensures that the appropriate people are involved in making those decisions, that they follow an appropriate process, and that they are held to account for the results.
Complexity
Product innovation teams must often deal with high degrees of uncertainty and ambiguity. Market data is often incomplete, contradictory and subject to rapid change. Technology capabilities are constantly evolving. The factors that drive market dynamics are, at best, poorly understood. The relationships between these factors are non-linear and cyclic, changeable, subject to unpredictable time lags, and often very non-transparent. These may be characterised as conditions of complexity.
Traditional command-and-control management styles do not cope well with complexity; indeed one of their tenets is to reduce complexity and increase predictability whenever possible. However, this has the side effect of reducing innovation, which is inherently about dealing with the novel and the unknown, and hence the complex.
There are two popular models for understanding complexity, the Stacey Matrix (Stacey, 2007) and the Cynefin model by Dave Snowden (Snowden and Boone, 2007). We base our definition on the Cynefin model.
The Cynefin Framework
Cynefin, illustrated in Figure 2.1 and Table 2.1, is a sensemaking framework that helps decision-makers understand their situation, and hence make better choices. Problems can be understood within one of four domains or contexts: simple, complicated, complex and chaotic. In addition, a fifth context of disorder captures those problems that cannot be easily characterised.
Figure 2.1 The cynefin framework
(adapted from Snowden and Boone, 2007)
Table 2.1 Attributes of cynefin domains
The simple and complicated domains are ordered. They are predictable, either by anyone (simple domain) or by experts (complicated domain). Causal relationships are visible and can be used to plan actions for the future. Decision making structures can exploit this order and predictability.
The complex and chaotic domains on the other hand are unordered. They are unpredictable: causal relationships are not visible enough or stable enough for long-term planning. Decision-making must be set up to allow localised experimentation and action.
In the simple domain, causal relationships are straightforward, visible and linear. Problems have one correct solution and “best practice” always exists. Having sensed the situation, a simple categorisation allows us to determine the best response. The action we take is virtually certain to give the expected result; in the odd case when it doesn’t, troubleshooting is easy. Since the results are predictable, decisions can be scripted and processes automated. Governance based on command-and-control models works well.
In the complicated domain, people with the appropriate expertise and experience can at least identify causal relationships with sufficient analysis. Problems can be deconstructed into a number of smaller problems that are easier to solve, and the working of the whole system can be deduced from the working and interaction of the components. Problems may have multiple good answers, each with different trade-offs: it makes sense to talk about “good practice” rather than best practice. After sensing the situation we need to analyse it before we can respond. If our analysis is good, the action is highly likely to give the desired result.
Unsurprisingly, this is a domain where experts and specialists thrive. Decisions are made by groups of appropriate experts. Project and mid-level managers play an important role coordinating and supporting the work of these experts.
Unlike the simple and complicated domains, causal relationships in the complex domain are cyclical or hidden and can therefore be seen and understood only in hindsight, if at all. The behaviour of a complex system is emergent: the system as a whole behaves in ways that cannot be extrapolated from the components.
Without clear causality, it is impossible to find optimal answers. Several more or less likely solutions may exist: the best choice cannot be determined by ex-ante analysis of the situation. It is difficult to predict the impact of any decision: by the time the potential outcomes have been analysed, the chosen option may have expired or the whole question may have become irrelevant.
In the complex context, we must first probe the situation, undertaking low commitment, and ‘safe-to-fail’ experiments. We then sense the impact of these probes and respond to what we have learned. There are no best practices or even good practices, but rather we allow appropriate practices to emerge from our experimentation. If an action gives good results, we do more of it and try different variations to see if results improve. If an action doesn’t work as expected, we try something else. Research organizations typically use this kind of experimental and empirical approach to emerge a workable solution over time.
Experimentation is essential in the complex domain, as is acceptance that well-designed experiments are equally likely to fail as to succeed. It is not useful to make detailed plans beyond the immediate future in such circumstances, and centralised chains of command typically struggle to keep up with the ever-changing situation. Distributed decision-making and self-coordination is required for effective operations.
In the chaotic domain, all bets are off. The parameters and causal relationships are constantly shifting. While you may be able to exploit a pattern for a while, this is not reliable and can’t be repeated indefinitely.
In the complex domain we start by probing the situation, in the chaotic domain our probes yield little useful information. Instead, we must act decisively in order to stabilise and simplify the situation. We can then sense the situation in the “islands of stability” we have created, and then begin to respond. For example,
The Red Cross is accustomed to working in chaos, but they can only do it repeatedly by simplifying their context. When a disaster occurs somewhere in the world, the Red Cross sends in a trained team and a container with supplies and tools. The container has everything the team needs in order to survive and help victims—water purifiers, canned food, tents, power generators, medicines, communications gear, etc.—until they are lifted out a fortnight later.
The container forms a bubble of order in the surrounding chaos. Because their own working structures, processes and tools are known and ordered, the team can afford to take on complexity and chaos in their operative work. Without this little ordered domain all effort would be spent on surviving, but with the back-up from their ordered domain they can really make a difference.
This is the realm of the emergency and natural disaster. People must be empowered to act rapidly within their immediate situation, guided towards a common goal. Distributed decision-making is essential, supported by clear lines of authority and clearly defined overall objectives and operating principles.
The Borders of Complexity
Most organizations operate in either the complicated or the complex domain, simply because it’s difficult to turn a profit in the simple and chaotic domains. Commodities dominate the simple domain: work is performed for a standard (low) fee, or by the lowest bidder. Chaos, on the other hand, is so unpredictable that actions cannot be repeated in a profitable manner, making it difficult to sustain a business.
The borderline between complicated and complex is especially interesting for governance. As we move from complicated to complex, we lose predictability. This manifests itself in the form of frequent and seemingly random operational problems, for which a common root cause is impossible to discern. In such circumstances, a number of patterns of misinterpretation are possible.
If the context has been stable for a long time, people may become complacent. When changes occur, they do not correctly identify the context change or its implications. They stick to approaches and methods that worked well previously, but now mysteriously fail to have the desired effect.
Another common pattern is entrained thinking, perhaps better known as the “law of the instrument”—if all you have is a hammer, everything looks like a nail. For example, project managers are trained to use project plans, risk lists, Gantt charts and other tools that are appropriate for the complicated domain. When working in the complex domain, these tools must be replaced or significantly adapted.
Oversimplification is also common. Essential details may be abstracted away, giving the impression that a problem is simpler than it actually is. For example, in a well-written piece of software, every detail exists because it’s necessary for the program to do its purpose. Any one incorrectly designed or coded detail may cause the program to misbehave or crash, but high-level software architecture diagrams do not accurately convey this significance.
By definition, problems in the complicated domain have good solutions that can be found by experts with the appropriate knowledge. People accustomed to working in this domain may enter a state of analysis paralysis in the complex domain, where good solutions are not identifiable in advance. As experts analyse a problem, they continuously unearth details that invalidate their previous assumptions, and the definitive answer seems to be constantly eluding them. Too proud to ask for help, they doggedly continue to analyse the problem until it’s too late.
On the other hand, specialists can also become overconfident of their own ability to understand issues and identify root causes. Where causality is unclear experts may come to very different diagnoses about the problem and—each having confidence in his/her own analysis—resort to arguing about the nature of the problem and what kind of solution is needed. A more fruitful approach would be to develop a solution through experimentation and observation.
And finally, the trap of retrospective coherence is particularly nasty. When causality is unclear it is impossible to plan for every eventuality. In the complex domain, plan-based projects tend to fail due to unforeseen (and indeed unforeseeable) circumstances. The events and causal relationships leading to failure can however become visible in hindsight, e.g. in a post-mortem. Believing that the project operated in a predictable environment, the post-mortem may decide that the project plan was inadequate and recommend more detailed planning in future. This adds overheads and diverts attention from managing complexity, increasing the likelihood of future failures and thus feeding a vicious cycle.
These patterns are particularly pertinent as the pace of change accelerates, as rapid change often pushes operations from the complicated to the complex domain. For example, in slowly changing domains, designs are stable and decision-making by centralised committees is viable. In a rapidly changing domain however, the longer you hold on to a design the more likely it is to need rework, and design decisions must be made in frequent small batches nearby or within the teams that will use the design.
Agility
Walk into a factory and you perceive noise, activity and movement. But walk into a software development organization and you perceive silence, inactivity and stillness. A lot of things are happening, a lot of work is being done, but it is invisible to the naked eye. Fred Brooks argues in his famous essay No Silver Bullet (Brooks, 1975) that software is “invisible and unvisualizable”, having no single geometric representation. Software is more complex than any other human construct, and the complexity is arbitrary rather than systematic. Furthermore, software is not only subject to change but also, being easier to change than hardware, acts as an attractor for change.
Because these attributes capture the essence of what software is, Brooks concludes that there is nothing that can make software visible, simple and resistant to change. What you can’t visualise, model and understand, you can’t predict. Thus the software domain is inherently unpredictable, or to put it another way: in software projects the predictability horizon is uncomfortably near.
Nothing has changed in the four decades since Brooks wrote his essay, and the continuing lack of visibility and predictability still makes it very difficult to manage software projects. Despite well-laid plans and the best progress monitoring practices, a large proportion of software projects fail to meet either schedule or budget. And decades of research into project management have had little impact on project failure rates.
Thus a group of new software development methods emerged in the 1990s. Collectively known as “agile software development”, or simply “Agile”, these methods revolve around the idea that, if we are constantly faced with change and complexity, then we should design processes to accept change rather than suppress it. Figure 2.2 shows some of the attributes that enable agile methods to do this.
Figure 2.2 Positive feedback loops in agile software development methods
The general approach of agile methods is to refocus attention on outcomes rather than the processes that produce these outcomes, and on collaboration rather than contracts and specifications. Because software is invisible, complex and malleable, it must be reified and validated early and often, and used to elicit feedback from customers and end users. The only real measure of progress is working tested software that gives value to the customers and end users.
There are almost as many definitions of agility as there are practitioners, researchers and trainers. We favour the definition synthesised by Kieran Conboy (2009) from the basic principles of agility:
[Agility is] the continual readiness of an [Information Systems Development] method to rapidly or inherently create change, proactively or reactively embrace change, and learn from change while contributing to perceived customer value (economy, quality, and simplicity), through its collective components and relationships with its environment.
Agile methods are interesting for several reasons. Firstly, the nature of software and software development means that these agile methods embed forms of governance that may be especially relevant to the complex domain in general, and hence to governance of innovation and research.
Secondly, software is an increasingly important element of the innovation chain. A growing number of products and services are based on software, or incorporate software as a key component. Even products that don’t incorporate software in themselves are often designed, simulated and analysed using software-based tools. Thus our governance structures must increasingly account for the issues of developing software.
Agility in Practice
Agility is abstract: it is mainly a philosophy or collection of thinking models, combined with values and principles defined in the Agile Manifesto (2001). A variety of methods (almost twenty documented methods exist) supplement these with more specific practices and processes.
Some of these methods, e.g. Test-Driven Development (TDD), Agile Modelling and Pragmatic Programming, consist primarily of sets of mutually supporting agile programming practices. Others, e.g. Scrum, Adaptive Software Development, Feature Driven Development (FDD) and the Crystal process family are more comprehensive, covering software project, product and/or process management for large parts of the product lifecycle. Only a few methods, such as Dynamic Systems Development Method (DSDM) and the Rational Unified Process (RUP), cover the full life cycle including conceptualization and maintenance. More recently, a number of frameworks such as Disciplined Agile Delivery (DAD) and Scaled Agile Framework (SAFe) have tried to bridge devolved, self-organized teams into larger structures.
All agile methods share the following attributes (adapted and extended from Abrahamsson et al, 2002):
1. Incremental: small and well-tested software releases are made frequently, using build and test automation.
2. Cooperative: customers, developers and stakeholders work together with close communication, enabling distributed authority and responsibility, and rapid feedback.
3. Straightforward: the method is easy to learn and modify.
4. Adaptive: the method is able to handle (even last-minute) changes to products, processes and the organization.
5. Self-organising: the development team and stakeholders own the process, and continuously adapt it to produce more customer value in less time.
Agility in Comparison
Agility contrasts with traditional, plan-driven development methods in several ways.
First, agile methods test the current state of the product empirically and frequently, several times a day, thus increasing visibility. Does the code build and integrate? If so, does it pass all tests? If so, which features were tested? These metrics are direct, measuring the presence of value (working, useable features) rather than the absence of waste (defects and bugs). Being able to directly measure the outcome gives a reliable and current understanding of the quality level and feature set of our product, making it easier to manage the project. Projects that do not measure outcome directly must resort to secondary, indirect metrics, such as how our initial progress estimates compare to the progress of time.
Indeed, agile methods have a different understanding of progress. The product is constantly progressing, as can be evidenced from the outcome, but the concept of “100% done” has little meaning. Progress instead focuses on two questions: (1) “What can we do next to add as much value as possible to the product?” and (2) “How can we improve, so that we can add value even faster?”
Rolling plans, also known as backlogs, support the first question. Agile methods break work down into fine-grained items on demand, then queue these items for implementation according to priority, e.g. based on the perceived value of the work item, the direct cost, the cost-of-delay associated with it, or a combination of these. Because commitment to a work item is only made when implementation commences, items can be radically reordered or replaced when the situation changes.
The answer to the second question can be found in continuous improvement, an approach known from Lean manufacturing. As stakeholders become accustomed to seeing value steadily and reliably being added to the product, they become less interested in how the work is done or by whom. Thus the development team can take ownership of the process, to freely modify it and distribute the work as they see fit. The surrounding organization must give them enough latitude to self-organise around retrospection and process experimentation.
Agile methods also strive for simple products. This is because software has high holding costs and a low shelf life: many people believe that source code is an asset, but it is in fact a liability. Less code is also less complex, takes less time to write and test, contains fewer defects, is easier to read and understand, easier to maintain and debug, and so on.
Similarly, agile methods attempt to achieve process simplicity. Agile processes are uncomplicated, and the mechanics easy enough to understand. Meetings are few, short and frequent; roles are few and people are expected to work outside them if necessary. The process is jointly owned and easy to modify locally.
It can be seen that “agility” brings many differences to “traditional” software development methods: embracing rather than suppressing change; distributed rather than central authority; an emphasis on visibility, transparency and feedback. This shifts the locus of decision making from central managers to self-organizing teams, and requires a very different governance style.
Open Innovation, Crowdsourcing, Co-creation
Many organizations have recognized that they control only a small proportion of the world’s expertise in any given area. Customers know more about their own needs than the organization can ever know (although it may be able to help its customers discover and analyse these needs). The pool of specialists in research institutes, suppliers, think tanks, academia, etc. outnumbers the specialists employed directly by the organization. Experts outside the organization’s domain of operation may be aware of solutions that could be adapted to solve the organization’s problems, if they were aware of those problems. This recognition has led to growing engagement with a number of different techniques.
Chesbrough (2003) identified open innovation as “a paradigm that assumes that firms can and should use external ideas as well as internal ideas, and internal and external paths to market, as the firms look to advance their technology”, open innovation encompasses a variety of techniques aimed at encouraging people outside the organization’s boundaries to engage with its problems and propose solutions. These techniques include conventional joint venturing and contracting with research institutes, publication of challenge prizes, participation in informal innovation networks, crowdsourcing, collaborative product design, and so on.
In the product innovation context, organizations may use crowdsourcing to extract ideas and possible solutions from large groups of people (“the crowd”), typically via online forums. In this way companies may gain information about customer needs and preferences by interacting with communities of potential customers, or gain information about solution options and design trends by interacting with communities of innovators, designers and similar thinkers. Crowdsourcing differs from traditional research activities in that it involves more active engagement by the external community—the organization outlines the type of information it is looking for, then community members actively explore the problem and solution space for themselves rather than being closely directed by researchers.
Finally, co-creation is a process whereby the organization works jointly with customers to design and develop solutions to customer problems. Compared to the aforementioned techniques, cocreation emphasizes collaborative working—both organization and its customers are active partners in the process, working together through activities such as workshops and online collaborative spaces in order to analyze customer needs and problems, and hence develop potential solutions.
It can be seen that all of these concepts increase the permeability of the organization’s boundaries—information and ideas must flow between the organization and the outside world if it is to engage in them. This creates questions about the flow of control and decision-making—the organization must share some of its decision-making power with the external parties that participate in its innovation process. How should decision rights and processes, ownership of intellectual property, etc. be reconfigured to best support this, while still protecting the organization’s assets?
Why Governance Matters
Organizational Effectiveness
Organizations make many decisions during the course of product innovation—which markets to be in, which customers to address, which products to develop, how to design and manufacture those products, how to price and market them, etc. They must prioritize their use of limited resources, constantly choosing where to focus their budget, skills, time, specialist facilities, etc. The effectiveness and efficiency with which they make such decisions does much to determine whether they succeed. Thus effective governance contributes directly to product innovation success.
In particular, effective governance ensures four things:
1. People understand which decisions matter, and how they contribute to organizational goals. Innovation teams make hundreds of decisions every day. Many of these decisions are trivial; only a small percentage has a large impact on organizational goals. An effective governance framework will identify those important decisions and ensure that attention is focused on them. Furthermore, it will make it clear just why and how these decisions contribute to overall goals, so people are aware of the bigger picture as they make decisions.
2. The right people are involved with each decision. To make a good decision, we may need to consult with a wide range of expertise. We may need to engage with diverse stakeholders to ensure we understand their needs and perspectives. We may need to build buy-in from staff members who will execute the decision, negotiate with resource owners, obtain consent from regulators, etc. An effective governance framework will identify who needs to be involved in each type of decision, and what power they have to influence that decision. Do they have a vote, or even a veto? Or are they simply consulted in order to gather relevant information or opinions? It will also identify who is accountable for the decision, i.e. who has the ultimate authority to approve the decision, must justify the resources expended as a result of the decision, and bears responsibility for the consequences of the decision.
3. An appropriate process is used to make each decision. Following the agreed process gives legitimacy to a decision. When people can see that “due process” has been followed, they’re more likely to buy in to the outcome and hence to execute the decision effectively. Moreover, by defining the process upfront, we increase the efficiency and effectiveness of decision-making. People don’t need to spend time “deciding how to decide” for each decision, but rather can get on with the process of gathering information, conducting research, undertaking experiments and trials, identifying options, assessing those options against agreed decision criteria, etc. They are less likely to miss important steps or overlook key considerations. The organization can also provide training (e.g. by running simulations), ensuring people execute the process well when under pressure. Thus an effective governance framework will outline the process for each major decision, in sufficient depth to support people to make the decision, yet without overly constraining their actions and professional judgement.
4. Appropriate mechanisms for feedback and accountability are in place. Effective governance ensures that the organization monitors the outcomes of decisions and, if necessary, adjusts those decisions to reflect new information and learning. This is especially important in dynamic and uncertain environments (almost always the case for product innovation), where decisions must be made on the basis of incomplete and rapidly changing information. Effective decision-making is then about making tentative commitments and constant adjustments, rather than conducting “perfect” analysis and getting everything right at the outset. An effective governance framework will also define mechanisms to account for the resources expended as the result of each decision. Finally, it will ensure that these feedback and accountability mechanisms extend to the decision-making process itself—a well-governed organization constantly looks for opportunities to improve the way it makes decisions.
In summary, effective governance supports organizations to make good decisions—ones which are grounded in evidence and awareness of the current situation, which are backed by adequate analysis, and which are accepted by major stakeholders. It enables them to make those decisions in an efficient way, and it ensures that they track outcomes and hence constantly steer towards their desired goals. In the fuzzy and dynamic world of product innovation, this capability to constantly make and adjust decisions is critical to success.
Effective governance also tends to favour open and transparent decision-making. To govern well, organizations must establish effective mechanisms for monitoring and recording their decisions—who was involved, what process and criteria they used, what decision they made, and what outcomes that decision led to. Such transparency makes it easier for managers to understand what is going on within the organization, and hence to steer it effectively towards organizational objectives. It can also reduce costs associated with regulatory reporting and compliance.
Conversely, when governance is ill-defined, organizations may suffer from a number of issues. Often, decisions are delayed, either because people must spend time defining bespoke decision-making processes for each decision, or because they must spend time identifying and managing stakeholders (e.g. scheduling meetings to discuss options). At worst, people may use the lack of clarity to avoid making decisions altogether.
At they other extreme, decisions may be rushed. Delays in one part of the decision-making process often lead to shortcuts elsewhere. Organizations may scrimp on data gathering and analysis, for example, to compensate for time taken to manage stakeholders. A common pattern is to spend a long time defining “objective” decision-making processes, only to run out of time and make hasty decisions based on intuition and gut feel.
In so doing, organizations waste a lot of effort, spending an enormous amount of work on defining bespoke decision-making processes for every decision, fighting about decision rights, arguing over the minutiae of trivial decisions, constantly consulting people on decisions for which they have neither expertise nor interest, etc. Again, this diverts resources from activities that may contribute more to effective outcomes, such as gathering data or conducting experiments.
After all this, people may challenge the legitimacy of decisions, either because the right people weren’t consulted or because the appropriate process wasn’t followed. This act of revisiting decisions leads to yet more delay and wasted effort.
Lack of clear governance can also lead to inconsistent decisions. Different parts of the organization may use different criteria and information, and hence come to different outcomes for the same decision. Such inconsistency can be appropriate in highly uncertain environments where the organization wants to test different approaches, but it also increases coordination overheads and destroys economies of scale. At worst, it can cause confusion amongst customers, regulators and other stakeholders.
Finally, demarcation disputes and infighting about accountabilities all too often result in relationship breakdowns. The resulting lowering of trust levels can lead to a negative spiral—people interpret actions by their “opponents” negatively and hence respond aggressively, further damaging the relationship. In such an environment, effective decision-making becomes almost impossible.
Organizations may try to compensate for delays, waste and inconsistency by defining rules and policies to cover every situation—in other words, by introducing bureaucracy. This rarely works well in a product innovation environment, which is focused on dealing with novel situations. Attempts to apply inappropriate rules, or to customise existing rules for new circumstances, may simply lead to further delay and waste. Conversely, in the absence of clear lines of authority and decision-making processes, people may simply go their own way. This leads to anarchy. As mentioned above, anarchy may be appropriate when the organization wants to try numerous approaches in order to determine what works best in highly uncertain or complex environments. However, it creates high coordination costs and high degrees of inconsistency. Alternatively, powerful people may use lack of clarity to seize control. Despotism can sometimes lead to extremely efficient and effective governance, but only in the interests of the despot—wider organizational goals and the interests of other stakeholders may be ignored.
Ultimately, all of the above issues lead to poor decisions and hence poor outcomes. Decisions are made by default, or without adequate analysis and consideration. The perspectives of key stakeholders and technical experts are ignored. Resources are squandered on trivial decisions or on organizational infighting. No one monitors the results of decisions, so there is no way to steer and adjust when poor decisions are made, or when new information is acquired.
Organizations that don’t consciously and actively tend their governance end up spending a lot of time on it. They address it afresh for each decision as they argue about due process and decision rights and accountabilities. They then end up with little time for the decision itself. So they make bad decisions. In rapidly moving markets, product innovation teams can ill-afford the waste and delay this creates. Worse, by paying little attention to governance, organizations often find that their governance decays into inappropriate forms. In the absence of well-defined and consciously tended governance, the ground is fertile for extreme forms—anarchy, despotism or bureaucracy—to take hold. Such patterns are antithetical to effective product innovation, with its need for high levels of communication and information sharing, and its emphasis on rapid response to new information as it arrives.
Societal Benefits
As well as benefiting the organization, good governance gives benefits to the society within which that organization operates. By ensuring clear and equitable distribution of power within the organization, its staff and partners, and other stakeholders, an effective governance framework can, for example, help an organization embed the values set out by the United Nations Global Compact (http://www.unglobalcompact.org/). This addresses concerns such as:
• Human Rights. An effective governance framework will ensure that innovation teams consider the concerns of all people affected by a product’s manufacture, usage and disposal as they make decisions about the product and its lifecycle. This can help ensure that the product does not damage human rights. Thus, for example, ways in which a product might be used to violate people’s right to privacy could be identified at an early stage and eliminated or mitigated through appropriate design choices.
• Labour Rights. Likewise, good governance ensures that the concerns of staff throughout the supply chain are addressed when making decisions about product design, manufacture, support and disposal. Many organizations have used this approach to reinforce labour rights within their global supply chain, e.g. by ensuring that suppliers do not use forced or child labour during product manufacturing.
• The Environment. Likewise, good governance ensures that innovation teams consider the environmental concerns of all stakeholders. This increases the likelihood that products will be manufactured, used and disposed of in an environmentally sustainable way. The growing trend to account for lifecycle environmental performance as well as financial performance reinforces this attention to environmental sustainability.
• Anti-corruption: Feedback and accountability go hand-in-hand with transparency and openness. Well-governed organizations tend to be more transparent, and thus better placed to work against corruption. For example, when innovation teams procure components and services openly, with clearly identified decision-makers and well-defined decision criteria, then scope for bribery is much reduced.
It can be seen that a well-governed product innovation process can contribute to societal good in many ways. This is particularly true when an organization works with its partners to embed good governance across the product’s supply chain and throughout its complete lifecycle. The growing tendency to employ open innovation and a global supply chain gives innovation teams substantial scope to embed the values of the UN Global Compact across society.
An organization that is seen to respect the rights of its internal and external stakeholders, and to pay due attention to environmental sustainability and other societal concerns, is likely to gain significant indirect benefits. For example, customers are increasingly well informed and have started favouring organizations that produce their products in a fair and sustainable way. This improved customer perception can lead to increased market share, improved customer loyalty and retention, greater brand equity, etc.
In addition to customers, staff members are likely to favour organizations that treat them well and contribute positively to society. Improved staff perception can lead to greater ease of recruitment, higher staff retention, and greater willingness for staff to commit to organizational objectives. Similarly, improved partner perception means that innovation partners are more likely to engage with organizations that are seen to respect the rights of all stakeholders in the supply chain.
Further, governments, trade organizations and other regulating bodies may be more likely to consider the views of organizations that are seen to contribute positively to society, thus gaining greater influence with regulators.
Organizations also have less need to act defensively when operating in an environment where legal and human rights are respected. Thus they do not need to invest so much in protecting staff and property, creating legal protections, buying insurance etc., resulting in reduced operating costs. Likewise, some classes of investor favour organizations that are seen to act positively within society, and give them easier access to capital.
Good governance is not an overhead—it can create a virtuous circle, generating wins for all stakeholders. However, the way those stakeholders perceive governance, and what they consider to be “good governance”, may vary substantially. In order to establish effective governance, an organization must deal with these perceptions.
Governance suffers from the same problem as strategies, values and processes: the actual lived governance structures may diverge from the documented structures, which in turn may be different from the desired structures. Furthermore, there may be more than one ‘lived governance’ structure. Since organizations are made up of people with different tasks, roles, backgrounds, experiences and personalities, their perceptions of the current governance structures can in fact be vastly different.
These different perceptions can themselves undermine good governance. When people work at cross-purposes, invested effort goes to waste and more effort is needed to consolidate and rebuild the pieces. Differing visions also make ripe ground for internal dissent and political power struggles. In an organization with a clear and well-understood governance structure, people are more likely to work effectively and efficiently together.
Consider the following case study2:
A department in a multinational enterprise was given free reins and an unrestricted budget to modernise the enterprise’s product portfolio. Within months, three competing visions emerged:
1) Some people wanted to invest in a new platform that would allow the company to produce new products faster.
2) Others wanted to push several successive products out quickly and let the platform emerge and stabilise over time.
3) Others believed that the company should form a standards consortium with partners and collaborators and thereby take a leading role in the industry.
Everyone had their opinion, but on the whole, most people were concerned with getting stuff done, whatever that “stuff” happened to be.
The Director of the department was a charismatic and extrovert engineer, now promoted into a position that required a politician with the ability to juggle different perspectives. He developed a tendency to leave difficult decisions open until they resolved themselves. In this case, he left a decision vacuum: no-one had clear responsibility for resolving the issue, and few people were even aware of its importance. This vacuum allowed managers and engineers to set their own priorities. The situation involved very little malevolence or aggression: it was simply that no one knew who should make the decision between three equally good options.
This left the engineers building a changing product on a changing platform based on changing standards, further befuddled and delayed by constantly changing priorities. Unsurprisingly the first product was released on a half-baked platform more than a year behind schedule. The budding standards group was silently abandoned by the other partners and fizzled out. The department was closed down two years later.
This section explores some common patterns in perceptions of governance, and the way they undermine effective decision making in product innovation organizations.
Background
We became interested in the question of governance perceptions during a workshop on complexity in 2012. We thought it would be interesting to ask different people to draw their perceptions of governance by first listing a number of decisions important to their context, then forming a Cynefin diagram from those decisions.
The first trials surprised and delighted us. For example, a group of project managers stated that deciding on the project schedule was a complicated problem, bordering on complex. A group of engineers on the other hand saw the same decision as simple. When the two groups noticed the difference, an interesting and constructive discussion ensued.
The project managers explained that determining the schedule involved negotiating, consulting, delegating, informing and generally interacting with several stakeholders on different levels of detail over several weeks or months. The engineers on the other hand were only informed after negotiations had been concluded and thus did not perceive how much effort had gone into producing an answer. After this discussion both groups were a bit wiser: the engineers showed a greater appreciation for the work of the project managers, and the project managers said they would consult and inform the engineers when planning the next project.
On the basis of our experiences with these trials, we developed a workshop format that let groups plot the perceived complexity of different decisions on the Cynefin diagram, while simultaneously placing the same decisions on a decision/role matrix. In order to produce comparable results, we provided ready-made lists of roles and decisions that are valid across many organizations. The workshop is run in several groups within each organization, and its value lies in the differences between the resulting complexity and role matrices.
To date we have collected data from approximately 20 teams in a handful of organizations, typically R&D units consisting of 100–500 employees within larger enterprises in the telecom business (both operators and equipment manufacturers). In the workshops, mid-level managers typically form one group, quality managers another, system architects a third, developers a fourth, and so on. For the purpose of raising awareness and collecting feedback, we have also run the workshop in special interest groups such as the British Computer Society and Agile Finland and at several conferences on software agility.
While we certainly need to gather more data, patterns have already emerged that we can identify and recount here. The information presented below is based on preliminary results from ongoing research. It reflects our current understanding of how perceptions of governance might differ within an organization, and should be viewed as anecdotal.
Patterns of Perception
The workshops contain two different activities, the first activity reflecting the perception of the complexity of a decision. Preliminary analysis has revealed a number of interesting potential patterns. First, being involved in making decisions helps people appreciate the complexity of those decisions.
People who are involved in making a decision often categorise it as complicated or complex, while those who are merely informed of the results often perceive the same decision as simple. These differences can create conflicting expectations between managers at different levels in an organization.
Second, different work domains tend to have different complexity levels. For example, people and teams who work with technical issues tend to see their jobs as primarily complicated, i.e. quite predictable although requiring expert knowledge. Project planning decisions (e.g. scheduling, scoping, resourcing) are seen as either complex or complicated by the groups performing them. Project and product managers tend to see their own jobs as complex (or even chaotic). And in most organizations we surveyed, managers thought that getting people to collaborate and share information was clearly chaotic—fraught with uncertainties and unexpected errors and miscommunications.
The second activity of the workshop explores the perception of who owns a decision. Again, we found some interesting potential patterns in how power is distributed horizontally and vertically.
Among peers or groups of people with approximately the same power in the organization, we found evidence of decision vacuums. These occur when everyone thinks that someone else owns the decision, meaning that no one takes responsibility for it. We speculate that this may happen for many reasons—miscommunication, ignorance that the decision exists, a desire to avoid accountability, etc.—but the effect is the same: decisions don’t get made, or they get made by default. For example, when decisions about the prioritization of work fall into a vacuum, teams easily default to polishing their latest work while waiting for “someone else” to decide what they should do next. Decision vacuums can be a significant source of delay, inconsistency and standoffs.
The opposite of vacuums could be called decision stand-offs. Multiple parties each think they should own a decision, while being aware that others also claim some ownership. This may be resolved through negotiation or by reference to corporate hierarchies and other power structures, or it may lead to protracted argument and political infighting. For example, in one organization project managers thought they were responsible for deciding when a work item is completed to adequate quality levels. Developers also claimed responsibility for this decision. It emerged that projects spent a lot of time debating the “doneness” of work items and arguing about what constituted adequate quality, separately for each item.
In between vacuums and stand-offs, we find the concept of outliers: multiple parties believe that they alone own a decision, but do not recognise the involvement of others. For example, a company that used open source components in their product recruited several developers from open source projects. These developers were well-educated, often brilliant, and very happy to work on their pet projects, but owed more allegiance to their open source community than their employer. They therefore tended to prioritize development of functions benefiting the community over those supporting the company’s product. Priorities and schedules set by the company’s managers did not influence their decisions much, and constant reminders changed their behaviour only temporarily.
We also found evidence of layering. For example, project and product managers often attribute more decision power to themselves than others attribute to them. This is a potential source of conflict and stand-offs. Similarly, mid-level managers often attribute more decision power to operational teams than the organization does on average. This creates conditions for a vacuum—the manager thinks a team is empowered to make a decision, but the team doesn’t realize it has that power.
Teams seem to fall into one of two types. The first type is independent, making their own decisions; the second type is dependent, taking orders from project managers, product managers, specialists, etc. No teams were found to inhabit the middle ground, suggesting that the scale is bipolar rather than gradual. Each type of team, of course, needs (or is nurtured by?) a very different management style.
In this section, our analysis has focused on patterns of dysfunction (while the next section looks at models of “ideal” governance). One function of effective governance is to recognise when such dysfunctions are in play, and take action to resolve them. This is part of the fourth (feedback and accountability) aspect of governance, identified in the definitions above.
Models and Trade-offs
Organizations are more likely to establish effective governance when they tune the allocation of decision rights to the characteristics of decisions. Decisions about market requirements, for example, may require a different decision-making structure to decisions about technical solutions or product pricing. Many organizations limit their options here, for example making all product decisions in a single decision-making body. This section explores a number of models for allocating decision rights, and discusses the trade-offs that must be considered when choosing amongst these models.
Weill and Ross (2004) define six “archetypes” for allocation of decision rights. These archetypes, slightly modified for our context of product innovation, illustrate the range of ways in which decision rights can be allocated. They are as follows:
1. Business Monarchy: A central group of business executives (or a single executive) makes the decision. For example, a central committee might make decisions about which markets to address and how to allocate investment across the product portfolio. In the extreme case, the executive committee makes detailed decisions about all aspects of each product—design, “look-and-feel”, and so on.
2. Technical Monarchy: A central group of technical specialists makes the decision. For example, a group of technical architects might decide which technologies product teams will use, and how they will deploy these technologies. In extreme implementations, these technical specialists might decide on all aspects of the product portfolio. (This is common in technology-driven startups.)
3. Federal: Representatives from several organizational units come together to make the decision, typically via consensus or some voting mechanism (illustrating that allocation of decision rights is often interlinked with the decision-making process). For example, technical specialists from a number of product teams may come together to make decisions about common product line architecture. Federal models ensure that many units have a say in the decision, and hence can help build buy-in across the organization; they can also be cumbersome and prone to delay and blocking when there are substantial differences between units.
4. Duopoly: A central group of specialists works with representatives from the organizational unit affected by the decision to make the decision. For example, experts from a central technology unit might work with business managers from a product team to make decisions about product design and implementation. Or the product managers for an individual product line might work with central executives to decide on priorities for investment within that product line.
5. Feudal: Each organizational unit makes its own decision. Units may be organised by product line, functional specialism, geography or some other dimension or combination of dimensions. Likewise, they may range in size from major organizational units down to individual product teams (in which case the model grades into Anarchy). The essence of this archetype is that each unit makes its own decision without wider organizational input. Thus, for example, a geographic unit might decide which products to bring to market in its region.
6. Anarchy: Individuals or small units make the decision, based solely on their local needs and decision criteria. Anarchies can be useful when an organization wishes to operate several independent “experiments” to explore a new or rapidly changing domain, or when each team operates in a very different environment. Thus, for example, the organization may set up several “skunk works” teams to explore options for entering a new market, or it may allow teams developing products for very different markets to operate with a high degree of autonomy.
The archetypes may be combined. For example, many organizations set overall policy and define guidelines using a monarchic or federal model, then use a feudal or anarchic model to make individual decisions. Depending on how rigorously they enforce the central policy, this then gives each organizational unit more or less latitude, while still maintaining a degree of consistency across the entire organization.
No single archetype or combination of archetypes will be optimal for every type of product innovation decision. An effective governance framework might therefore choose the archetype or combination that works best for each decision type. This choice is rarely straightforward, and organizations must typically make trade-offs between a number of diverse factors.
First, each model varies in its ability to assimilate new information, affecting the speed of decision-making. In general, centralised models (monarchies, federations) have longer chains of command from operational teams to central decision makers and hence can be slow to make operational decisions. Monarchies (and especially despotisms) can however make rapid strategic decisions, as the decision-makers act with considerable authority. Devolved models (anarchy, feudalism) have shorter chains of command and hence can make faster operational decisions. Anarchies can be especially quick to react, if somewhat chaotic. Federations, with their requirement to consult widely and build agreement across multiple organizational units, tend to make decisions slowly.
Related to speed is frequency of decision-making. Some decisions need to be made regularly; some are rare; some are less predictable than others. It makes sense to automate the regular decisions where possible, e.g. by defining clear policies and setting up systems and tools to support them. In such cases, centralised models (e.g. monarchies) work well—the cost of developing policies and tools can be spread across many decisions. On the other hand, automating rare or one-off decisions is rarely cost effective. For such decisions, it makes more sense to convene the affected stakeholders as necessary and hence decide jointly. Thus federal and duopoly models may be more applicable.
Unsurprisingly, the complexity of the business domain is another important factor. Complex domains are not very predictable, and even a small detail can, if misunderstood, cause inordinate amounts of rework and delay. Keeping on top of complexity requires both high speed and a high frequency of decision-making, which may be provided by devolved models such as feudalism and anarchy. If the domain is merely complicated, devolved models do not necessarily provide enough cohesion, in which case federations or monarchies can work better.
Centralizing decision-making in monarchies and federations helps ensure organizational consistency—that consistent approaches and criteria are applied to all decisions. Appropriately configured duopolies can also perform well here. Thus these models perform well where consistency is valuable, e.g. to build consistent look-and-feel across several product lines, or to develop a product line architecture that allows components to be reused across multiple products, or to exploit economies of scale by purchasing in large batches. Feudal and anarchic models struggle to deliver the same consistency.
Compliance with regulatory requirements, mandatory or voluntary standards (e.g. ISO standards for quality, security, environmental performance, etc.), and other similar instruments tend to be easier when the organization can demonstrate consistent processes and decision-making. Thus models that favour consistency also work well here.
In order to make good decisions, it is important to have an understanding of all factors that influence the decision, also known as situational awareness. For operational decisions, this favours people who are “on the ground”—people close to the customer tend to be most aware of customer needs and other market factors, while people who are actively working with a technology tend to be most aware of its capabilities and limitations. Thus devolved models work well. For strategic decisions, awareness of wider organizational goals and strategic context may be more important, in which case models such as monarchy and duopoly may be more relevant (although even here, awareness of what is going on “on the ground” can be important).
Likewise, people may need access to expertise—the relevant technical, financial, legal, commercial or other specialists—in order to make a good decision. The chosen governance model must ensure that the relevant expertise is available to the decision makers. This often favours duopolies.
Good decisions are useless if they cannot be executed. Obtaining buy-in to decisions from the people who will execute them is therefore essential. People tend to buy in when they feel that they can influence the decision-making process, so devolved models (anarchy and feudalism) perform well here. Federal models can also score well—going through the process of building agreement across multiple units can be slow and painful, but when it is done, people are more likely to buy in to the resulting decision.
Many companies seek to optimise utilisation of specialist skills and resources. It can be easier to manage such skills and resources (e.g. legal experts, technical experts, specialist laboratory equipment) if they are placed into a central resource pool. For example, this allows people to prioritise work in a way that optimises use of scarce skills. In such cases, a centralised decision-making model (duopoly or technical monarchy) may work best. However, this can cause specialists to become separated from product innovation teams, thus losing situational awareness and creating queuing delays.
Finally, culture also plays an important role. Organizations are generally more comfortable with some archetypes than others. For example, a company founded and operated by a small number of people may be a business monarchy; technocratic organizations may be most comfortable with technical monarchies; a research university may favour feudal or anarchic models. As people generally work more effectively within comfortable forms, it can make sense to default to the culturally preferred form when there is no strong reason to use another. The problem, of course, is that many organizations use the culturally preferred form even when there are strong reasons to favour alternatives.
Table 2.2 summarizes these trade-offs. It can be seen that no archetype provides the best performance against every trade-off.
Table 2.2 Trade-offs in Governance Models
Many of these trade-offs entail balancing the benefits of centralised and devolved decision-making. Centralizing decisions helps ensure that policies and objectives are applied consistently throughout the organization. It also makes it easier to manage scarce skills and resources efficiently. However, it risks keeping decision makers remote from local operational concerns and customer needs. It also lengthens chains of command, slowing down decision-making and reducing people’s buy-in to decisions. Devolution gives the reverse: in particular, it can be a way to build buy-in, speed and situational awareness—all vital in product innovation.
Many organizations address this problem by separating policy from implementation—one group of people sets policy for a decision type (e.g. what standards to apply, what decision criteria to use), while another group then applies these policies in order to make individual decisions. This yields the four broad options listed in Table 2.3.
Table 2.3 Centralized versus devolved governance
Of course, an organization is free to use different models for different decision types. It may, for example, use an anarchic model in the early stages of product development, when the focus is on rapidly exploring many options in order to buy information about different markets. It may then centralise decision-making as it integrates the lessons learned through such market experimentation into its global product lines.
An organization might also use a dynamic mix of these models. Some organizations shift between centralized and devolved models as a conscious strategy to build internal skills and expertise. People might work initially in a central pool in order to build expertise and consistency. This pool is then dispersed to field locations in order to apply that expertise to specific products while building local knowledge. The central pool is then reformed in order to bring local knowledge back to the center. And so the cycle continues.
Finally, modern communications tools can reduce the impact of many of these trade-offs. Tools such as video conferencing, mobile phones, instant messaging, team collaboration environments, and suchlike, all serve to increase the facility with which people can communicate organizational objectives, local situational awareness, expert opinions and feedback, etc. This can enable centralised decision-making bodies to gain situational awareness and to communicate their decisions more effectively. Thus these tools can mitigate the weaknesses of centralized decision-making models.
Equally, such tools enable organizations to communicate overall objectives and strategic intent to devolved teams. Or they enable fragmented teams to coordinate their actions and hence operate with greater consistency. Thus they also mitigate the weaknesses of devolved decision-making models.
Nonetheless, it remains the case that many innovation projects fail through weaknesses in communication. Effective communication is inherently difficult, especially in uncertain, fast-moving environments. Technology can reduce the problem, but it cannot eliminate it. Thus it remains important to align governance structures with the key decisions that the organization must make.
Heuristics
So, governance is complex. We must trade-off many factors and options. We must adjust it dynamically to changing circumstances. Where does that leave us? This section, based on our experience working with organizations across the public, private and not-for-profit sectors, outlines nine heuristics for establishing and sustaining good behavior in the organization, making it easier to install appropriate governance structures for product innovation.
1. Address Power and Politics. Governance is about power structures, within the organization and between the organization and its partners and other stakeholders. Power comes from many sources—position in the organizational hierarchy, control of resources, personal expertise and charisma, support from other stakeholders, legal authority. A decision cannot be implemented if it is not backed up with the right mix of power—good governance ensures that mix of power is brought to bear.
Many governance initiatives fail because they focus on the minutiae of decision-making rather than on mobilizing necessary and legitimate power. This confuses governance with management. Governance is about ensuring that the right decision makers are engaged. Management is then about making the necessary decisions—gathering information, identifying options, making choices, etc. When governance encroaches on management, it tends to become bureaucracy. It tries to predict all eventualities and prescribe ways to deal with them. This is unlikely to succeed in product innovation, where high levels of uncertainty and the need to respond rapidly to new information are the norm.
2. Enable People to Exercise Judgement. Good judgement adds more value than rule-following. Governance bodies may be able to define prescriptive policies for simple decisions, but they can rarely do this for many of the decisions that occur in complex, uncertain environments. In product innovation, it is more important to ensure that the right people are engaged, and that they have access to the appropriate tools and information sources. This then frees them up to exercise their professional judgment when making decisions.
3. Favor devolved control. People who are close to the situation—through talking to customers, working directly with new technologies, etc.—tend to understand local capabilities and customer needs far better than central policy makers. Thus it makes sense to devolve control to individual product teams wherever possible. Organizations need to place some bounds around these teams, e.g. setting overall budgets and market objectives, to keep them aligned with organizational priorities. But innovation teams will generally deliver more value when central controls are relatively lightweight.
4. Articulate Organizational Objectives. People need to understand the wider context if they are to make decisions that correctly balance local circumstances with organizational goals. When most decision-making is devolved to local teams, the organization must communicate its objectives clearly so that teams can make decisions that align to these objectives. This also helps build consistency across the activities of different teams.
5. Create Transparency. Good decision-making requires high levels of awareness; it is paramount that relevant, fresh and trustworthy information is available when and where it is needed. Transparency also makes it easy for others to see what is happening, improving communication and accountability, and preventing misuse, corruption and undue politicking.
6. Keep Policy Clear and Simple. Organizations generally need to define some amount of central policy, e.g. to ensure consistency across teams, to support regulatory compliance. However, defining too much policy can be counterproductive. Complicated bodies of policy take too long to understand and create too much scope for conflicting rules. And they frequently fail to anticipate important issues that arise during the course of product innovation, rendering them of little value to innovation teams. Innovation gains most from simple policies that set clear boundaries and define overall principles. People can then interpret the nuances that apply to their specific situation.
7. Regular Cadence; Small Batches. When decision-making bodies meet infrequently, they create delays. For example, product teams spend a lot of time with progress blocked while waiting for decisions. And because decisions are infrequent, teams spend a lot of time crafting their inputs to meetings rather than developing the product. Even scheduling meetings into people’s diaries becomes a major task.
Large, infrequent meetings also tend to be ineffective. Because they must address a large batch of decisions, people pay little attention to each individual decision. Attention begins to wander before the meeting is finished. And because the gap between the meeting where a decision is made and the meeting where its outcomes are assessed is long, people struggle to learn from experience.
Decision-making bodies tend to be more effective when they meet frequently (precluding the queuing delays and long feedback loops) and at regular times (precluding the scheduling issues).
8. Seek Feedback and Learn From It. Even with “perfect” governance mechanisms in place, people will make mistakes. In the face of uncertain and incomplete information, they’ll make incorrect assumptions. They’ll overlook key information. They’ll be blindsided by competitors. And governance mechanisms are rarely perfect—new groups of stakeholders emerge and influence decisions without being included in formal governance structures; changed technology capabilities require adjustments to processes; policies fail to reflect changed circumstances; and so on.
An effective governance framework therefore needs to capture feedback at two levels. First, it needs to create mechanisms to monitor the outcomes of decisions and adjust the decisions when they are not having the intended effect. Second, it needs to monitor the effectiveness of the decision-making process, and adjust governance structures where they are not operating effectively.
9. Be Careful of Institutions. While socially acceptable forms (institutions) of organization and governance—e.g. the matrix organization, project management, yearly budgets etc.—may be useful, it should be remembered that they always embed past practice. These practices have evolved mostly in the complicated domain and do not necessarily perform well in the complex domain, including in product development, research and innovation.
Another problem with adopting socially acceptable forms of governance is that organizations implicitly limit themselves to whatever their competitors are doing. Successful organizations on the other hand break out of the mold, defining their own ways of working that better fit their strategy.
Conclusion
Governance is an important driver of organizational performance. Organizations which make good decisions, and which make those decisions in an efficient way, will be better placed to provide valuable products and services to their customers, citizens and other stakeholders. This applies to all aspects of an organization’s operations, but it is particularly important in product innovation, where the rapid pace of change, high levels of uncertainty, and increasing trend towards complex, cross-organizational forms, can make decision-making especially challenging.
Good governance of product innovation also has considerable scope to improve societal outcomes. By considering the perspectives of all stakeholders affected by a product’s design, manufacture, use and disposal, innovation teams can ensure that their products do not damage human or worker rights, and are environmentally sustainable throughout their lifecycle. This in turn benefits the organization—such products are more attractive to many consumers.
However, good governance only arises when an organization actively attends to and maintains its governance structures. When an organization ignores governance, ad hoc structures emerge. People spend a lot of time defining bespoke governance arrangements for each decision. At best, this creates inefficiencies. At worst, at leads to poor decision-making and allows inappropriate governance structures—bureaucracy, despotism and anarchy—to take hold.
In a world of rapidly moving technology, ever-greater micro-segmentation of markets and personalization of products, globalization of supply chains, and ever-greater competition, product innovation teams cannot afford to be constrained by inappropriate governance structures. Attention to governance is likely to become increasingly important for effective product innovation. In particular, governance models that support small, empowered, self-organizing teams to operate with rapid feedback loops and with clear awareness of and buy-in to wider organizational goals are likely to come increasingly to the fore.
References
Abrahamsson, P., Salo, O., Ronkainen, J., and Warsta, J. (2002). Agile software development methods. Review and analysis. VTT Publication 478, VTT Electronics, Espoo, 2002. URL http://agile.vtt.fi/.
Agile Manifesto. (2001). URL http://agilemanifesto.org/. Last checked on 2014-04-26.
Brooks, F.P. (1986). No silver bullet – essence and accidents of software engineering. Proceedings of the IFIP Tenth World Computing Conference. pp. 1069–1076.
Chesbrough, H.W. (2003). Open innovation: The new imperative for creating and profiting from technology. Boston MA: Harvard Business School Press.
Conboy, K. (2009). Agility from first principles: reconstructing the concept of agility in information systems development. Information Systems Research 20, no. 3, pp. 329–354. doi: 10.1287/isre.1090.0236.
Reis, E. (2011). The Lean Startup: How constant innovation creates radically successful businesses. New York NY: Crown Business.
Snowden, D.J., and Boone, M.E. (2007). A leader’s framework for decision making. Harvard Business Review 85, no. 11, pp. 68–77.
Stacey, R.D. (2007). Strategic management and organizational dynamics: The challenge of complexity to ways of thinking about organizations (5th ed.). Harlow: Prentice Hall.
Weill, P., and Ross, J.W. (2004). IT Governance: How top performers manage IT decision rights for superior results. Boston MA: Harvard Business School Press.
Additional Reading
Anderson, D. J. (2010). Kanban: Successful evolutionary change for your technology business. Sequim, WA: Blue Hole Press.
Argyris, C. (1970). Organization and innovation. Homewood, IL: Irwin.
Allen, P., Maguire, S. and McKelvey, B. (2011). The SAGE handbook of complexity and management. SAGE Publications Limited.
Atkinson, S. R. and Moffat, J. (2005). The Agile organization: From informal networks to complex effects and agility. DoD Command and Control Research Program (CCRP).
Basili, V. R., Shull, F. and Lanubile, F. (1999). Building knowledge through families of experiments. IEEE Transactions on Software Engineering, 25(4):456–473.
Brooks, F.P. (1975). The mythical man-month: Essays on software engineering. Reading MA: Addison-Wesley.
Beck, K. (2010). Extreme programming explained: Embrace change. Boston MA: Addison-Wesley.
Byrne, J. (2012). The occupy handbook. New York NY: Little, Brown and Company.
DeMarco, T., and Lister, T. R. (2013). Peopleware: Productive projects and teams. Harlow: Addison-Wesley.
Goldratt, E. M., and Cox, J. (2004). The goal: A process of ongoing improvement. Aldershot: Gower.
Handy, C. B. (1987). Understanding organizations. Harmondsworth: Penguin.
Nagappan, N., Murphy, B. and Basili, V. R. (2008). The influence of organizational structure on software qual- ity: an empirical case study. In ICSE ’08: Proceedings of the 30th international conference on Soft- ware engineering, pages 521–530, New York NY: ACM.
Poppendieck, M., and Poppendieck, T. D. (2003). Lean software development: An agile toolkit. Boston, MA: Addison-Wesley.
Reinertsen, D. G. (2009). The Principles of Product Development Flow. Second Generation Lean Product Development. Celeritas Publishing.
Schein, E.H. (2004). Organizational culture and leadership. San Francisco CA: Jossey-Bass.
Schwaber, K., and Beedle, M. (2002). Agile software development with Scrum. Upper Saddle River, NJ: Pearson Education International.
Stapleton, J., and DSDM Consortium. (2003). DSDM: Business focused development. London: Addison-Wesley.
Weinberg, G. M. (1991). Quality software management. New York NY: Dorset House Pub.
Discussion Questions
The authors claim to take a perspective as practitioners working within product innovation organizations, and hence present a number of observations about the forces which are driving change within such organizations. How well-founded are these observations? Identify other reports or research which examine today’s context for product innovation and compare the factors they describe with those discussed here. How well do the models (e.g. of complexity and agility) discussed here fit with those other factors? What other theoretical frameworks might be relevant to the forces described here? What implications do these frameworks have for effective governance of product innovation?
Describe different models for governance of product development and the trade-offs they make against attributes such as speed and consistency of decision-making. Discuss how the most appropriate trade-off might vary as a product moves from the complex to the complicated and simple domains during the course of its lifecycle.
Describe ways in which good governance of the product innovation process can support the development of products that benefit society as well as the organization that developed them. Research examples that illustrate the benefits received by society and the product development organization through promotion of good governance throughout the product innovation lifecycle and supply chain.
The chapter promotes the thesis that product innovation in uncertain, fast moving environments happens most effectively when small teams are empowered to make local decisions with rapid feedback loops. Discuss the challenges this creates for global organizations with large product portfolios. How might other, more centralized, governance models be adapted to support effective product innovation?
How can individuals, teams and organizations recognise whether they are working in the complex or the complicated domain? What indicators would be useful?
The section on ‘Models and Trade-offs’ advocates creating different decision-making structures tuned to the characteristics of different decisions. This can lead to “decision making sprawl”—many decision-making bodies controlling different parts of the innovation process, with consequent costs to communication and coordination. Discuss the trade-offs between use of a single decision- making body for all decisions and defining different bodies for different decisions.
Key Terms
Accountability: The obligation for a person to explain and accept responsibility for their actions and decisions.
Agility: The ability of a product development method to create, embrace and learn from change while rapidly creating value for the customer.
Complexity: A state characterized by hidden, non-linear and cyclical causal relationships between a system’s components and its environment. The behavior of a complex system is emergent: the system as a whole behaves in ways that cannot be extrapolated from the components.
Crowdsourcing: The practice of sourcing ideas and solutions from large groups of people (“the crowd”), typically via online forums.
Cynefin: A sense-making framework that helps decision-makers to understand their situation, and hence make better choices. Problems can be understood within one of four domains or contexts: simple, complicated, complex and chaotic. In addition, a fifth context of disorder captures those problems that cannot be easily characterized.
Governance: Governance determines who has power, who makes decisions, how other players make their voice heard and how account is rendered for decisions taken.
Open innovation: Identified by Chesbrough (2003) as “a paradigm that assumes that firms can and should use external ideas as well as internal ideas, and internal and external paths to market, as the firms look to advance their technology”.
Uncertainty: A state characterized by limited knowledge, where it is impossible to fully describe the current situation and potential outcomes.
Authors Biographies
Graham Oakes helps people untangle complex technology, processes, relationships and governance. As an independent consultant, he helps organizations ranging from startups to global corporations and intergovernmental agencies to define strategy and hence set up and support project and product development teams to deliver that strategy. His book Project Reviews, Assurance and Governance is published by Gower. E-mail: [email protected]. Phone: +44 7971 546288.
Martin von Weissenberg is a trainer and coach in agile and lean software development. After a decade in a wide range of roles and assignments in the software industry, he moved into process and organization development in 2004 and from there into agile and lean methods in 2007. In the spare time he doesn’t have, he is writing a PhD on how to organise, manage and lead for agility. E-mail: [email protected]. Phone: +358 400 314159.
____________
1 This case study is derived from our work on product innovation within AB Corporation.
2 Again, this case study is drawn from our experience working with product innovation organizations.