8. Learning Innovation: How Do Organizations Become Better at Innovating?

The Importance of Learning

In fast-changing environments, the ability to learn faster, better, and more cheaply than your competitors could mean the difference between retaining market leadership and barely surviving.¹ In Silicon Valley, it became clear to major venture capital companies that ramping up the sales volumes in startups always took longer and cost more than was projected. Sometimes the result was slower-than-expected growth, but often it resulted in the failure of the startup because of poor cash flow. Reviewing the successes and failures of startups led to the development of the Sales Learning Curve to guide better, faster, and cheaper sales growth in the crucial early days. The Sales Learning Curve concept says that the time it takes to achieve cash flow breakeven is reasonably independent of sales force staffing. Instead, it depends entirely on how well and how quickly the entire organization learns what it takes to sell the product or service while incorporating customer feedback into the product itself.²

The key to learning is not to avoid making mistakes, but to learn from them.³ Microsoft has made learning from mistakes a virtue. Version 1.0 of its products is “good enough,” but the company improves the product relentlessly until it dominates the market. It does not wait to have the “best” product, and it knows that to improve the product, it needs to listen and see how the product is used. However, organizations are not usually adept at learning. Many organizations do not learn from their mistakes regarding innovation and are stuck repeating errors and feeling the same frustrations.⁴ The key to organizational learning is to have systems in place that enable people to learn better, faster, and with less associated expense.⁵ You can often tell that an organization suffers from innovation learning disabilities after seeing several symptoms. They are disbelievers in the effectiveness of innovation; they stumble upon execution of innovation projects, favoring only incremental innovation, and they are constantly amazed that others invest in innovation at levels higher than theirs.

Company A (the name is meant to keep it anonymous) suffered from significant learning disabilities that hampered its innovation. The company had a poor track record of innovation successes: a few winners and a bunch of losers. The company’s managers exhibited a penchant for rampant incrementalism and scoffed at other companies that invested in major innovation initiatives. The company was locked in a downward spiral driven by a lack of learning and in which organizational antibodies attacked every innovation. The company lacked meaningful metrics for innovation; therefore, efforts to learn what it did right and what it did wrong were frustrated. And no special efforts to learn from its own innovation actions or those of competitors could be launched because the management team did not believe that innovation was worthwhile. The downward spiral went like this: Because the company had not innovated successfully in the past, it did not believe in the value of innovation. Because it did not believe in innovation, it did not learn about it. Therefore, the company did not innovate well. For the record, this company performed very badly relative to its industry counterparts and was eventually purchased at a discount.

Company B had similar learning disabilities. The company had a mediocre track record for innovation, a culture that tended to dismiss the value of innovation, and a high level of organizational antibodies that attacked change and innovation. In addition, management had no useful metrics or learning systems in place to assist the innovation processes. However, this company’s management team saw the long-term implications of this handicap and directed the staff to diagnose the situation and both identify and fix the sources of weakness in its innovation approach. The analysis led to an overhaul of the innovation approach. Changes included a measurement system developed to help change behavior and improve learning, enhanced innovation processes and organization, and specific efforts to combat the organizational antibodies that frustrated innovation in the company. The result was a rejuvenated company that began to develop and implement important innovations. Currently, the company has one of the highest levels of profitability in its industry and is often held up as the standard for long-term success.

Organizational learning and change go hand in hand. Because innovation is all about change—incremental, semiradical, and radical change—learning is an intrinsic part of innovation. Properly conceived and executed, organizational learning can unleash powerful forces of creativity and the development of processes to focus them into successful commercial realities.

A Model of Learning

Two major types of learning exist: Learning to Act and Learning to Learn.⁷

Learning to Act

This type of learning includes collaborative assessments of how the current systems (including the structure, processes, and resources) are working, a shared understanding of strengths and weaknesses, and a proactive effort to improve them.⁸ This type of learning takes the current strategic objectives as given and does not question them. Quality circles, where team members brainstorm about making the current manufacturing process better, are a classical example of learning to act. The team does not question whether the company should be producing PDAs or whether they should be produced in-house or subcontracted. Taking these decisions as given, the team looks for ways to improve the efficiency of the existing plant.

At a strategic level, most planning systems extrapolate the current business models into the future. This extrapolation represents Learning to Act, in which the planning effort focuses on incrementally improving the current actions. This type of planning is key to moving the company along the current strategic trajectory. However, it may run the risk of becoming a ritualistic exercise, with little, if any, value added—the worst case of bureaucracy. Becton Dickinson, one of the largest medical devices companies in the world, had a strategic planning process that ran from December through September. It incorporated strategic, operational, and financial plans. Throughout the cycle, there was significant flow of information along the organizational hierarchy, with frequent meetings. The outcome of the process was a thick book with detailed strategic analyses, plans, and financial commitments. The nature of the process and the relevance of financial plans led to a strategic plan that projected a conservative picture of the current business model into the future and incremental innovations.⁹

Learning to Learn

This type of learning consists of structured processes to assess how well the organization learns and changes. This big-picture perspective is critical to ensure that the investments in innovation are yielding maximum return and that the organization is building sustainable innovation. The Learning to Learn cycle questions the current processes as the best way to innovate. By questioning what is pursued and how, the organization is more open to new ideas and educated risks (the type of risks that can have large payoffs). At the process level, reengineering does not try to improve the way processes are currently performed. Reengineering efforts do not improve what is already done; rather, they question the assumptions behind what is currently done to come up with potentially very different ways of executing the same tasks.

A high-technology company used strategic planning as a Learning to Learn exercise. The planning cycle had two processes going on at the same time, culminating in an annual off-site meeting. The first day and a half of the meeting focused on the financial goals and programs for the upcoming year, in much the same way as a traditional strategic planning process.¹⁰ The last part of the meeting focused on exploring new technologies, market trends, and any ideas that any manager thought were relevant. Each lead was examined, and teams were formed to further explore those that were judged to be more relevant.

Innovation relies on both learning cycles; incremental innovation relies to a larger extent on the Learning to Act cycle, and radical innovation uses the Learning to Learn cycle more often.¹¹ Both types of innovation use different types of knowledge. Incremental innovation is grounded on knowledge that is widely shared in the organization. People know what the problem at hand is; they know what potential solutions exist, what the current process intends to do, why it is in place, and how it works, and they can easily communicate ideas. Knowledge that is widely available and easy to communicate is called explicit knowledge. It can be stored and retrieved from knowledge repositories. Consulting firms such as Accenture encourage their project teams to write up the main features of each project—the problem, how the team approached the solution, and the final recommendation. The objective of this knowledge-management process is to make existing knowledge available throughout the company. Instead of starting from scratch, new projects can reuse existing knowledge, coded in an intranet, to provide better service to the client.

Radical innovation relies to a lesser extent on explicit knowledge. Because it dives into unexplored territory where knowledge is not well articulated, it is raw knowledge in the heads of people that crystallizes through their interactions. Management may know that it has a potentially great idea, but it is often unable to fully articulate it or find hard numbers to back it up. It is intuitive.¹² This type of knowledge is called tacit. Radical innovation is hard not only because of the novelty of the idea, but also because communicating it so that other people understand it is difficult.¹³ SpaceShipOne developers could not specify exactly how many and what types of people would buy their commercial space flights; however, it was intuitively obvious that a business opportunity existed.

Learning Systems for Innovation

Systems interact with the learning process of an organization at four different levels. The first two are more amenable to incremental innovation, while the latter two are more relevant to radical innovation:

• Systems for delivering value. These systems reflect what the organization knows and make this knowledge explicit in processes that can be controlled and acted upon if deviations happen. Learning is embedded in the design of the process and the responses to the deviations.

• Systems for refining the current model. These systems move the current business model into the future. They embed the Learning to Act cycle that forces constant improvement.

• Systems for building competencies. These systems facilitate the learning associated with new capabilities. Top management uses these systems to induce the organization to experiment and develop capabilities needed for future strategies. These systems guide the knowledge-creation process in the direction of the chosen capabilities.

• Systems for crafting strategy. These systems encourage and capture knowledge outside the current business model that emerges throughout the organization. Ideas happen all the time, and these systems make sure that they are not wasted and do not go into creating value in a different company.

Systems for Delivering Value

The first set of systems distills what the organization knows and applies it to a new event. For example, product-development roadmaps synthesize what the organization knows about the product-development process. Texas Instruments’ product-development process is described in a booklet. It details the various stages that a project goes through, from validation of the idea to technical specifications, execution, and market introduction. The booklet specifies the requirements at each stage, what is evaluated at the review points, and who gets involved.

These systems capture explicit knowledge coded into systems that govern the innovation process. Their purpose is to make sure that the innovation effort has the highest chance to deliver the value it is intended to generate. Learning happens through adapting the systems to the needs of each particular effort.

These systems give visibility into the future of the innovation project. This type of learning is anticipatory—it accrues from planning ahead of time, from examining the different alternatives before the team dives into the execution of a project, from outlining a path that, even if flexible, provides direction to the innovation effort. The learning involved in these systems is mostly incremental, in that it usually ensures that the project develops knowledge along the expected path.

Systems for Refining the Current Model

The second level at which systems interact with learning is through their own improvement and the improvement of organizational processes. During the execution of a particular project, learning about the process itself is captured. In other words, learning takes place not only about the particular innovation, but also about how the company can improve its innovation processes. Systems here have the purpose of refining processes. Semiconductor manufacturing equipment relies on very high technology. When products are introduced, they meet demanding technical specifications. However, product costs are not sustainable as competition erodes prices. The product development teams rightly focus their attention in solving electron-related problems. But over time, lowering costs is necessary. Applied Materials, the leader in the market, addresses this problem by having a department whose objective is to reduce product costs after product introduction. The department has a set of procedures to identify, prioritize, and address cost-reduction opportunities. In one case, the department redesigned a subsystem made of 17 different parts into 1 part, at one-ninth of the cost, and with better lead time and quality.

The learning here is not as much anticipatory as it is experiential. Team members draw on their experiences to identify problems and envision solutions. The learning process may often incorporate knowledge from other organizations, through visits, use of consultants, or external experts. Knowledge is tacit; it does not exist before the problem is solved and usually develops through the effort of a team. In a way, the knowledge dispersed in the heads of people and teamwork makes it into coherent knowledge that is made explicit.

We have identified the main learning purpose of each of these two systems. It does not mean that these systems cannot fulfill different roles. A system to deliver value may lead to review and refinement of the process itself. That may lead to identification of a radical innovation prospect. The framework presented here is meant to help in design and execution, but it is not meant to be a forced straitjacket.

Systems for Building Competencies

Certain systems interact with learning through their role in building future competencies. Top management often drives strategic renewal. It asks the questions and comes up with novel answers for the design of the company’s strategy. Top management senses the need for a new approach—for lack of current performance or anticipation of future threats to the current model.¹⁴

The process of building competencies is much more complex than that of evolving current knowledge. It requires both anticipatory and experiential knowledge. Anticipatory knowledge happens through strategic thinking and planning. Merrill Lynch had to design a new strategy to respond to the emergence of online brokers such as E*TRADE or Charles Schwab. The market forced Merrill Lynch’s top management to think creatively and respond to an important threat to its business. It could not wait for something in the organization to solve the problem; the company had to develop new competencies to adapt itself to the new market. But new things cannot be fully brewed in the laboratory. Developing new strategies requires experiential knowledge that happens through the experience in developing the competencies. Barnes & Noble, the successful book retail chain, quickly learned that responding to Amazon.com was not simply a matter of reproducing Amazon.com’s website. It required learning how to use the strength of its physical presence in the online world. Both types of knowledge constantly interact as new information requires a redesign of the plan.

This constant back-and-forth between vision and action benefits from periodic meetings, revised goals, and deadlines. In contrast to incremental innovation, in which systems to deliver value compare plans with progress to make sure that the project is on track, systems to build competencies use periodic deadlines to pace the organization and to bring together different players to exchange information and crystallize knowledge. When Sony began the development of its magnetoscope line (the earliest video recorders), the first goal that Masaru Ibuka, Sony’s founder, gave to Nobutoshi Kihara, the manager of the product line, was to produce a working product. The first video was the equivalent at the time of $55,000. Once the product was developed, Ibuka asked Kihara to develop one that would sell for $5,500. As the cheaper version was introduced to the market, Ibuka went back to Kihara; this time, he asked for a $550 color video for the Japanese market. The product was later known as Betamax.

These types of meetings are comparable to board meetings in startups. Board meetings pace the organization, force management to leave tactics and look at the strategy, and bring together people with different backgrounds to give a fresh new look at the company.

Systems for Crafting Strategy

Finally, systems interact with learning to craft the future strategy and business models. Innovation often emerges from unexpected places in the organization. The idea of the PlayStation did not emerge from top management, but came from a middle manager who did not find support and had to sell the idea as a way to sell more CDs. Intel’s transition to microprocessors happened because of middle managers’ decisions. Siebel, Salesforce.com, and Business Objects are all successful startups founded by Oracle managers.

These are unplanned discoveries that initially may grow outside top management’s span of attention. If the organization does not identify them, they will move and thrive as independent companies (and potential competitors). This type of radical innovation has attracted most of the attention because of its uniqueness, the appeal of the story when successful, and the attractiveness of making it happen even if odds were against its success. In contrast to the proactive process of building competencies for a new business model, this type of radical innovation has a larger reactive, improvisational component.¹⁵ Figures 8.1 and 8.2 summarize the four interactions with learning systems.

Figure 8.1. Incremental innovation

Figure 8.2. Radical innovation

How to Make Learning Work in Your Organization

Learning is captured through a proactive approach. Several tools have proven to be valuable in helping crystallize learning so that the organization can use it.

Knowledge and Ignorance Management

Managing innovation requires knowledge management (using what you know to the best effect) as well as ignorance management (being aware of what you do not know and the implications of that ignorance). Knowledge management is useful in incremental innovations. Ignorance management is most valuable in semiradical and radical innovations.

Knowledge Management

Knowledge-management systems are important elements to code data and give it a structure that makes it useful throughout the company. These systems rely heavily on information technology. They store particular executions of organizational processes that can help current projects be more efficient. Their value depends on their design—how easy it is to store and retrieve information, how is the database structured—and the discipline of the organization to code learning histories of the projects.

British Petroleum designed peer assists as a mechanism to leverage the knowledge dispersed throughout the company. Business units lent people in their organization certain knowledge for another business unit facing a specific problem that required this expertise.¹⁶ McKinsey & Co. developed its Practice Development Network, documenting the company’s experience with its clients, and published its Knowledge Resource Directory (with people’s expertise and other key documents), which was quickly adopted throughout the company. By 1996, the Practice Development Network had almost 12,000 documents, with 2,000 downloads per month.¹⁷

Ignorance Management

Incremental innovation builds on data about established technologies and existing markets, and uses the process of knowledge management and data mining to move forward. For radical innovation, ignorance management replaces knowledge management. Managers familiar with incremental innovation feel massively uncomfortable when ignorance dominates. Data are hard to find and must be generated from ignorance management.

When Salesforce.com first introduced its concept of sales force management delivered through the Internet, there was little knowledge to rely on. Little was known about how small and medium companies would use the system, which ones would adopt the system faster, what features were most useful, or how to approach its potential clients. Decisions were based on managed trial and error. During this early period, the company carefully collected data about its customers and how they used the product. As the company matured, analysis of this data became the source of improvements to the product and redesign of the strategy. One of the studies of the data uncovered that large companies were also using the product. The study concluded that complexity of the business rather than size was a better way of segmenting the market.¹⁸

Ignorance management is the process of identifying the most important things the team does not know and designing an approach to help reduce that ignorance to a level that allows forward movement. Experiments are great ignorance-management tools. They help in resolving technological decisions and also business model design. Approximation is another ignorance-management tool. Rapid prototyping puts in front of potential users the existing concept of a product and provides quick feedback about its design. Educated guesses are needed to advance through radical innovations. It is always better to have something “good enough” than to wait until the perfect product has been designed.

The Project Roadmap

Project roadmaps assist management in understanding how different innovation efforts reinforce each other. A roadmap visualizes how the learning in a particular project becomes the basis for a new project. Projects are not isolated efforts that compete against each other as in ranking systems. They form a whole, in which learning accumulates to make possible alternatives otherwise unfeasible.¹⁹ Motorola has effectively used project roadmaps to plan and develop its line of products. Robert Galvin, the company’s legendary CEO, was a big supporter of the process: “The fundamental purpose of Technology Reviews and Technology Roadmaps is to [ensure] that we put in motion today what is necessary in order to have the right technology, processes, components, and experience to meet the future needs for products and services.”²⁰

Figure 8.3 describes the product roadmap (a particular application of project roadmaps) for a medical device company. This particular map was developed at the strategic level and then detailed within product divisions. The roadmap projected the evolution of technologies, markets, and how products would be brought into the market as technologies and markets created opportunities. The technology roadmap includes owned technologies, as well as technologies that the company knows are being researched outside. Finally, the product roadmap is more detailed early on. As it moves into the future, it outlines expected products and synergies.

Figure 8.3. Product roadmap for a medical device company

Failures As Part of the Process

It is impossible to predict in the beginning of the creative process which ideas will be successful and which won’t. However, it’s important to realize that failure is part of the creative process. Viagra was born from an apparent failure to treat heart disease. A logistics software company found out who would buy its products through experimentation and failure. Initially, it targeted medium-size companies. This strategy seemed adequate because software firms did not have logistic products for this segment of the market. Despite the selling efforts, the software was moving slowly. The software was too expensive, and logistics was low on the priority list. Luckily, the vice president of marketing tried to see how the concept would come across in a division of a large firm that happened to be located in the neighborhood. The selling process went very smoothly. The software performance and price point perfectly suited the need of large companies.

Learning what does not work often leads to what does.²¹ If organizations do not recognize “failure value,” people will be discouraged from experimentation through the fear of failure.

Learning Histories

Learning histories is a term for stories that are specially constructed to uncover the truths about how an organization innovates. Learning histories review past projects, initiatives, and situations to identify in as unbiased a way as possible what really happened, what worked, what did not, and what the possible root causes were. In story form, they speak to individuals in the organization better than slide presentations. They use a historian’s perspective to gain crucial perspective and insights. Learning histories are not designed to highlight specific actions, people, or events, but instead identify recurring themes to answer the questions “What do we, as an organization, do repeatedly that impacts our performance positively and negatively?” and “What are the consequences, and why do we do it?” Frito Lay used learning histories to better understand the root causes of success and failure in its innovation efforts when it developed its strategic plan for growth; knowing what made its innovation engine run well was important before it could place its strategic investments.

The purpose of the learning history process as a whole is to spark new insights both in the people who took an active part in this experience and in others within the organization who could benefit from sharing in this learning. The purpose is not to assign blame or credit; it is to try to learn from what these people are willing to share with others about their experience.

The Dynamic Nature of Innovation Strategy

As an industry moves through its lifecycle, the learning that supports innovation varies. Figure 8.4 illustrates the level of change in different parts of the industry lifecycle. The y-axis indicates the level of turbulence. Turbulence starts at a low level, peaks, and then goes back to a low level. The highest threat to survival for an incumbent at a particular stage happens when turbulence is highest. A lower level of turbulence indicates a lower threat but not the absence of risk. Ice makers in the nineteenth century were improving the efficiency of ice harvesting to beat each other while a new technology was emerging that wiped them out in less than two decades. Coca-Cola was focusing on more effective marketing campaigns to beat Pepsi, while startups were coming with new beverage concepts for market segments that the two main players totally overlooked.

Figure 8.4. Learning lifecycle

The Technology Stage

Early in the lifecycle of an industry, technology innovation often dominates. This is a very fluid stage in which different companies bet on different technologies, a risky environment typically populated by startups. Learning focuses on exploring new technologies and generating new solutions. Learning systems focus on building the capabilities to develop the technologies that top managers have in mind or on crafting new ideas that may radically change the technology. In the late nineteenth century, car companies were numerous. Each one was trying different ways to put together a car. Some of them used gasoline engines, but lots of them were betting on steam engines similar to the ones used in railroads. The personal computer industry in the late 1970s consisted of myriad companies, each one understanding the PC as solving different needs. More recently, the customer relationship management (CRM) software market was populated with several hundred startups. Each one had a different solution to managing customer information. Today, several groups in large and venture-backed companies are competing to grab the Voice over Internet Protocol (VoIP) market.

The outcome of this first lifecycle stage is the emergence of a technology that dominates the market. In the car industry, it was the gasoline engine; in the PC market, the WinTel structure became the dominant solution; and in the CRM market, Siebel Systems emerged as the dominant software solution. The many companies that bet on the wrong technology disappear.

Large firms may create the technology and drive this first stage. Sony did that in developing the technology and business model for the Walkman. But often companies rely on learning from outside. Their corporate venture capital arms establish links with the startup community to scout the environment. Their objective is to spot promising new technological solutions. This learning strategy, in which the crafting of new innovations happens outside the firm, supports the acquisition of innovations that are complementary to the current strategy or that can be brought inside the company into a new business unit. Intel capital supports startup companies with technologies that complement Intel products. This support happens not only through funding, but also in helping the companies leverage the resources and contacts available at Intel and its partners.

The Performance Stage

After a dominant technology emerges, performance begins to improve quickly. Radical innovation in the underlying technology is still possible but much less likely than in the previous stage. A few companies may still bet on a new technological solution that would radically change the market. But most companies invest in improving the performance of the technology as quickly as possible. At this stage, performance is measured primarily on a single dimension—for example, image resolution in digital cameras or uptime for network providers. Competition often focuses solely on this dimension.

The competitive advantage goes to the company able to execute the learning cycles faster. Market share shifts quickly to the product that performs better. In the late 1980s and early 1990s, defibrillators, medical devices implanted in the chest of patients that stimulate the heart when they detect a problem, entered the performance stage. Guidant and Medtronic were the main players. The main limitation of defibrillators was their size. Bulky devices could not be implanted close to the heart (as pacemakers were) and had to be implanted in the belly. A belly implant had two main problems: One was the reliability of the long cables going to the heart, and the second was the patient’s comfort. Doctors and patients did not care much about performance dimensions (both companies delivered excellent performance) other than size. Each learning cycle by one of the companies meant a smaller device, and each time a new product entered the market, the market share moved to this product. Both companies were excellent at executing their learning cycles, and market share kept shifting from one to the other as new products were introduced. The performance cycle ended once both companies were able to get a small enough defibrillator that it could be implanted next to the heart. At that point, size became just one of the many performance dimensions relevant to customers. The industry moved to the market segmentation stage, with different customers valuing different considerations.

The Market Segmentation Stage

As product performance improves, certain customer segments become happy with the level achieved and start valuing different product dimensions: price, availability, cost of ownership, aesthetics, style, and so on. A new stage in the innovation process starts.

In the previous stages, the challenge from the business side was to design a business model to deliver the technology. Compared to the fluidity of the technology, the market was stable. Now the technology stabilizes and fluidity moves to the market. Customer needs evolve quickly, and new segments appear at a fast pace. The investment in learning moves to develop market knowledge. The winners are companies that are able to “read” the market and understand the differences across market segments.

The medical imaging division of a large European company suffered because it missed this transition. The division had been the market leader since its inception. The company had commanded the market through constant improvement of the quality of the image. The resolution of the image had been the main performance dimension for a significant period of time. As image quality improved, market segmentation began including the evolution of a price-sensitive segment that was willing to sacrifice performance for price. This segment saw the division’s products as offering too much performance and being too expensive. New entrants quickly captured that market segment, and others reversed the division’s dominant position.

The Efficiency Stage

As market segments stabilize, competition shifts to efficiently create more value to customers—whether in the supply chain, design, or marketing. At this stage, efficiency becomes critical, and the winner is the company that becomes more efficient. Winning can happen in this stage via superior learning on how to make a steady flow of incremental innovations. Toyota’s early lead in the post-1970s auto market came from its ability to constantly learn how to become better at quality. No single event, but rather the intersection of various management practices, led to its success.

At this stage, as in previous ones, most of the innovation game is played around a particular theme—but surprises can occur and disrupt the focus. A new technology, a quantum leap in performance, or a new segmentation of the market may redefine the industry and move it to a new lifecycle. Airlines in Europe were competing at how to be more efficient when the low-cost carriers such as Ryanair and easyJet entered the market with a radical proposal along the price dimension.

The Complementarities Stage

In the last stage, the focus shifts to managing complementarities. This capability comes from the ability to maximize the synergies among different products and businesses within a company. It also comes from establishing a network of partners that can substantially enhance the value proposition to the customer. Competition shifts from identifying the value proposition for each market segment to managing interactions and complexity. The success of FMC Corporation, a diversified chemical company, resides on its ability to manage the synergies among its three main businesses: agricultural, industrial, and consumer. Competition among game console manufacturers Sony, Microsoft, and Nintendo is mostly about creating and maintaining a network of software companies that come up with blockbuster games for their consoles.

As industries evolve incrementally through these lifecycles, radical innovations can move the industry to a new lifecycle at any time. Unless a company has a strategic imperative to invest part of its portfolio in radical innovations, the processes and culture solidify around improving the status quo and leave the company vulnerable to radical changes in any of these dimensions.

Learning and the Innovation Rules

In a healthy innovative company, leadership supports learning and puts in place the systems for it to happen. This includes quick-and-dirty diagnostics that are run to provide critical insights into problems and opportunities, as well as more complex learning systems that operate continually to provide feedback and guidance, such as planning tools. Driving innovation into the business mentality requires learning and change. Dell learned what was important to succeed in its innovation strategy and worked hard to ingrain the learning into the business mentality and culture.

Managing the balance between creativity and value capture requires learning systems. Otherwise, despite best intentions, one always becomes dominant over the other, and the correct balance is lost. Apple suffered from too much creativity and not enough value capture, despite a strong focus on innovation processes. Learning what went wrong with that balance and fixing it was one of the things Steve Jobs made happen when he returned as CEO.

In addition, innovation networks are fairly dynamic. Managing them requires information and learning to remodel and update the structure. Without learning, the networks become bureaucratic, cumbersome, and ineffective. Many companies that have established strong networks have failed to maintain the levels of learning and change required to keep them current, and the networks have become weak and fallen into disuse. A major global equipment company did not maintain its networks during a growth spurt that lasted for the better part of two years. It was distracted by the opportunities at hand, and it committed resources to other activities. Once the growth rate decreased, the innovation networks had deteriorated and were not in shape to support the level of innovation required to fuel the next round of growth; the company’s growth rate began to decline.

Finally, learning is one of the most important elements in combating organizational antibodies. Preventing the antibodies requires learning systems and activities that allow the organization to differentiate good change from bad change. Otherwise, the organizational antibodies become unselective and attack and disrupt all change. In that state, innovation is dead. Company A, described earlier, had severe learning disabilities. There was not enough learning to counteract the organizational antibodies, and the company was in a death spiral. Company B had similar learning problems but engaged in focused learning on its innovation approach and broke the spiral.

Innovation learning changes over time, as the business and industry evolve from an initial technological focus for innovation through to a mature stage where efficiency is the focus of innovation. However, the importance of learning does not change—it stays a high priority throughout an organization’s involvement with innovation.