The Challenge of Data

If you're this far into this book, you won't be surprised if I say the future of every company depends on customers, not technology. You also shouldn't be surprised if I say you can never know too much about your customer. Now I'm going to add something new: we have entered an era of unlimited data, and the volume is only going to keep growing as more so as more data sources and types of data get added to the mix, arriving faster than ever thought possible. Is your organization prepared to respond in real time to a customer comment made on Facebook from halfway around the world? You could be facing that challenge right now, today, as you read this. To reply by saying, “Good point; let me get back to you tomorrow” probably won't satisfy the customer who posted the comment. This is why you need not only to know why you want this data and how you plan to use it but also to be prepared to use it when the opportunity arrives. You need to understand how to effectively manage and use this increasingly rapid stream of unlimited data and be prepared to respond in real time.

The big data revolution gained significant momentum in 2014, amassing an incomprehensible amount of data: a zettabyte in just the past two years and significantly more by now. Yet the revolution currently under way is not in the unprecedented quantity of accumulated data, but the ways in which intelligence is derived from that wealth of information. Salesforce CEO Marc Benioff stated that we are “in the early stages of a data science revolution.”

A data science revolution sounds scary, exciting, and mind boggling all at once. The real question for you is what this means for enterprises, particularly your enterprise, and how it will impact your business. My hope, actually my expectation, is that it comes down to achieving predictability, which is the antithesis of scary or mind boggling.

In an ideal world, perfect data science takes every customer interaction and identifies patterns that can be repeated and proactively acted upon. Humans and artificial intelligence systems collaborate in an amalgamation of contextual, timely insight to better understand each customer and, ultimately, to predict their future behaviors. The results won't be flawless—data science is not yet perfect, and scenarios are not necessarily ideal—but this data-based guidance will be accurate enough to give us results that we can almost always depend upon to make the best decisions.

Don't get me wrong; I'm not conjuring up The Matrix here; this isn't about that kind of world. Rather, data science will increasingly enable an organization—your organization—to get ahead of a customer's actions, to know what your customers want before they do. This powerful ability is the focal point of business intelligence and customer engagement strategies, since it allows businesses to become more competitive and service oriented, getting them closer to their customers than ever before. In the past, we could only make thoughtful guesses at this and pat ourselves on the back when we got close. Now we can skip the guessing and hit the target almost every single time.

This goes beyond customer service and marketing. The next generation of successful businesses will look to data science to revolutionize their entire organization. The commitment of today's analytics and business intelligence is that data will no longer be isolated to databases, pulled by queries in response to an ask, but rather will be integrated throughout an organization's business process layer to predict and push prescriptive actions upon employees to take the most relevant and timely next step.

Yet data science is not confined to empowering just the service and customer-facing employees; it cannot exist in a siloed environment. It must be integrated across the entire organization, prompting businesses to realign executives, employees, and processes around the customer and absorb insight-driven value propositions throughout. Design and data science must converge to deliver seamless, relevant, and hyper-integrated customer experiences—across channels, devices, and touch points—to reimagine businesses and their entire workforce. In fact, it will also help organizations to more efficiently manage their workforces, assigning people to tasks they do particularly well but envisioning an entire career path for each individual throughout the organization.

The conversation in today's global market pivots to revolve around not how much data an organization captures, but how it leverages that data to get closer to its customers while maintaining scalability across the enterprise. Data science is the next generation of artificial intelligence and human capability that powers connections between a business and its customers. This isn't black magic, but solid science that can be grasped, tamed, and applied to a wide range of problems managers face all the time. Those who embrace this data science revolution will lead the way to becoming a customer-first business. Those who don't will scramble to play catch-up and wonder how the others did it.

Data Integration/Migration/Quality—A Salesforce Recap

The following section is primarily Salesforce specific, although the general principles apply to any operation. If you are not a Salesforce shop, feel free to jump ahead to the next section.

Clean data enables better decision making and improves the effectiveness of marketing and sales. It's also a huge commitment to keep your data clean—if not monitored properly, your data will become dirty and less valuable again. So once you've achieved clean data, how do you keep it that way? Here are four strategies to keep your Salesforce data clean.

Maintaining data quality is paramount to drive ongoing efficiency in your organization, and these strategies will help you reap the benefits of clean data.

Data Silos/Democratization

Data used to be concentrated in the hands of a few analysts. If you wanted a report pulled, you had to go through at least one person, if not a series of people, to get the information you needed. They had the required codes and keys. Then, if it needed to be tweaked based on feedback, you had to go through that same process all over again.

Today, data should be accessible by any member of any team, but if you don't have the structures, policies, and practices in place to take advantage of it, the information will remain stuck in silos. Cross-collaboration is imperative to take full advantage of data, so you have to break down the silos to achieve its maximum potential.

This is where the democratization of data comes into play. The democratization of data has been a hot topic for a while—and it's finally here for real thanks to the cloud. Business intelligence (BI) no longer rests solely in the hands (and minds) of IT and database administrators. Today all end users, from salespeople to marketers, can have easy access to data and the tools that improve their daily decision making. Cloud BI tools promise an unprecedented level of self-sufficiency to more or less an entire organization.

The promise of easy access to data is certainly achievable today. However, despite the democratization of BI, many organizations have yet to realize the benefits. Organizations are not leveraging BI tools to unlock the easy access to data that is technically very possible. In fact, Gartner surveys reveal that 70 percent of potential BI users in an organization fail to adopt CIO-sponsored analytics tools.³ With 71 percent of companies planning to increase their analytics budgets in 2016, it is imperative that users adopt new analytics tools and make sure that investment doesn't go to waste.

If you're not ready to implement cloud BI tools like Salesforce Wave Analytics across your entire organization, at least consider a proof-of-concept trial to prove value. Democratization of data is a result of tools and capabilities that, due to the masking of the previous complexity, are incredibly flexible and readily available. Many even use natural language, not code. Organizations are no longer locked into massive requirements gathering before undertaking an analytics project—cloud BI tools allow you to conduct proof-of-concepts with your own data and just a few plug-and-play BI tools. Looking for predictive capabilities? Buy an app. Need an extract, transform, and load (ETL) tool for data migration? Just plug it into your existing CRM. From data enrichment to report creation and data visualization, third-party apps and tools will allow you to empower your teams with self-service BI tools now.

Probably the best example of the democratization of data and its corresponding analytics is IBM's Watson. Remember Watson, the natural language intelligent computer that beat the best human champions of Jeopardy? Back then, Watson was physically large (two huge IBM Power Systems racks loaded with encyclopedias of data). Today, Watson is packaged as a cloud service. You can feed it your data, ask it questions in normal language, and watch it pull correlations and insights you didn't think were there, seemingly right before your eyes. You'll scratch your head saying “How the hell could we have missed that?” But you did. It's easy to miss the great gems hidden in your data.

Behind Watson is cognitive science. You communicate with Watson in human language, not code, and it figures out what you are talking about and puts your question or statement into the correct context. A little scary, huh?

IBM has also made Watson almost idiot proof. It's delivered as a cloud-based service. You upload your data and start asking Watson questions in plain English, and Watson will get back to you with correct answers in plain English. If your data isn't sufficient, Watson has access to volumes of information it can tap to enrich your own data. You don't need to own powerful servers, you don't need to build massive databases, code queries that resemble rocket science, or hire armies of PhDs. Watson knows what to do and will walk you through what you need to do as it delivers insights culled from your data.

In short, you use Watson as you would any service in the cloud. IBM is even offering Watson for free to those who want to sample the capabilities for real. These are the real services and you'll get Watson's best shot. When you want to use it seriously with production volumes of data, then you'll pay. But even then, you only pay for what you use. You can, in effect, buy Watson's services by the drink. These paying users can analyze higher volumes of data and tap more data sources, including live links to sources such as data warehouses and cloud sources. Watson is amazing, but don't take my word for it; go to YouTube and watch Watson take on the best human Jeopardy champions and win convincingly. And that was Watson several years ago; since then, IBM has enhanced Watson in so many ways that what you encounter in the IBM Watson cloud is light years beyond that initial machine.

What You Can Do Now

While data democratization may seem like a tall order, it will be a key differentiator for organizations moving forward. When you begin to plan how your organization will make data more accessible, consider how the following criteria could (and should) come into play:

Make data-driven decisions become the norm for end users and their processes. Data should be so abundant across your organization and embedded in business processes that consumption is easy and automatic. Think specifically about how business intelligence (BI) tools can revamp mobile through embedded analytics, rich visualizations, and data-driven workflow integration for your company.

Encourage and enable end users to directly influence how data is presented and consumed. If it's easy for them, they'll adopt the tools. If not, your organization will most likely struggle with poor adoption. But there is no reason it shouldn't be straightforward and easy. The tools are out there.

Align your data visualizations to the organization's key business outcomes. BI tools today are so powerful that they're able to make large data sets immaterial. Only the most relevant information is served up at a given juncture in the workflow process, despite capturing every meaningful customer touch point, relevant structured or unstructured attribute, or new data source like the Internet of Things (IoT). Invest in the art of dashboarding to serve up key metrics that address issues and prompt a next step.

Dashboarding represents an effort to make your data as intuitively clear as the information displayed on the dashboard of your car. You don't want to puzzle over how much gas is in your car. You want to know at a glance when your gas is running low and how far to empty. Same with the temperature gauge; dashboarding tries to clarify complex data and present it in such a way that any manager can get to the point fast.

At Bluewolf we worked with GoodData to come up with an approach to dashboard creation in five simple steps⁴:

1. Understand what attribute dimensions are most appropriate for your business or role, and which chart types will best visualize those dimensions.
2. The best charts show comparative data. This can be historical, year over year, month over month, groupings, or date progressions.
3. Dashboards should be organized so that trends and key performance indicators (KPIs) stand out and are easily identifiable. (The top left chart is usually the chart most people look at first in countries where written languages are left to right.) Also, think about what charts will be seen when the dashboard is first clicked on. The most important charts should be seen in that first screen.
4. Too many colors in a chart will be overwhelming to the reader and will obscure the message. Try another grouping instead. Keep background colors to a minimum.
5. Titles should be descriptive and any assumptions clearly explained. Attribute filters, such as date ranges or territories, should be applied at the dashboard level rather than individual report level whenever possible.

Data can be misleading, and you don't want to be led to the wrong conclusions. Proper visualization of data insights can be just as important as the data itself.

Enable your business with self-service data preparation, data enrichment, and data integration capabilities. Yes, there will be a learning curve and even some resistance (until your employees start seeing what their peers are achieving by way of data preparation). Consider who the users are and what they want out of BI, then develop training for them that will improve adoption. Also, don't forget gamification, one of my favorite strategies for cajoling cooperation and participation. Some modest gamification may be all it takes to get many of your employees to take the data plunge themselves.

IT and the business need to develop partnerships to make this new comingling of capabilities run more smoothly. Business needs to look to IT for insights surrounding data governance and security, while IT needs to be more involved in the business's strategic goals and understand users' needs for data visibility. The path toward data democratization can be disruptive, and for many there may be a turbulent road ahead, but the impact it will have on business outcomes like revenue growth, efficiency, and customer retention will be well worth the investment.

Data management and particularly data governance have become important with the emergence of cloud computing as a primary way organizations work. According to our annual report, The State of Salesforce, 64 percent of companies are now releasing new functionality to their Salesforce users at least monthly, up 20 percent from the previous year (Figure 9.1). Companies can achieve this speed of innovation through cloud governance—brief commercial break—especially our Cloud Governance practice, which focuses on building policies, creating governance boards, and deploying application life-cycle technology to fuel ongoing innovation. Unlike traditional governance practices, our approach to cloud governance is specifically intended to align IT and executive strategies, drive short development cycles, and engage the user.

Finally, a last word on master data management and the Salesforce Wave. Let's start with master data management (MDM), a strategy, structure, and practice for managing all data across an organization. Its goal is to ensure that all data in the organization is accurate and consistent, or to put it another way, that there is just one version of the truth. That's a big part of the Bluewolf Customer Covenant from the previous chapter. MDM ensures that every person who interacts with the customer has the same data, and that it is current and accurate.

This is not easy to do. In many cases, a company might have data in different forms, in different systems, for different purposes, all concerning the same customer. This data represents the details and history of the customer relationship, products the customer has purchased, all billing and invoices for the customer, and, to add a little complexity, the vendors used to service a customer or your customer's customers. As if that wasn't complicated enough, your enterprise might have several different business units with different product lines, each with its own departments and its own relationships to the same customers. And it gets worse; some customers use a middle initial with their name, but not necessarily all the time. So, are Joe Smith and Joe M. Smith the same customer or not? Confused yet? In short, MDM recognizes data inconsistencies across customers, multiple departments and systems, and more. It knows that Joe M. Smith is the same Joe Smith whether it is accounting, support, or sales.

This takes us to Wave. Seventy-six percent of companies still struggle with data integration and quality (Figure 9.2), but Salesforce Wave (or Salesforce Analytics Cloud) promises to change that. MDM really gets critical when you start to apply analytics to the data, which the majority of companies are planning to do. When you do that, you have to be confident that your data is good, meaning consistent, accurate, and current. Ever heard the acronym GIGO (garbage in, garbage out)? Even a little garbage data will pollute or, more accurately, poison your best analytics effort. Suddenly your people cannot be confident in the analytical results. And they are going to be relying on these results to make better business decisions? They might just as well as be wagering on a hunch. If you're struggling with unclean or siloed data, you must understand this problem and fix it.

While I've used Salesforce as my primary example for analytics and statistical measurements, these ideas and concepts can be applied to any cloud-based analytics tools, so don't feel constrained if you are currently using a different CRM platform.

Mobile and IOT

You can't think about data, data quality, and master data management without thinking about mobile and, at some point soon, about the IoT. Study after study points to mobile emerging as the dominant way people will access, use, and generate information and initiate transactions now and going forward. Like the PC was 25 years ago and the laptop 15 years ago, the mobile device is the vehicle all of us will use to access and work with information. Of course, it will be combined with and enhanced by cloud computing, but the mobile device will be dominant. You will use it to get a can of soda from a vending machine; to conduct a transaction with your bank, whether checking your balance or closing a commercial loan; book travel; watch a movie; consult with a doctor; or anything else. We will conduct our lives through the mobile device. In fact, many people already do. That means your business systems have to deal with mobile as a primary input-output device. It will be your wallet, calendar, organizer, address book, house keys, car keys, identification—it will become everything you carry in your pocket, purse, and various satchels (but without the Kleenex).

Now, try this: use your smartphone to access Watson in the IBM cloud and ask it some questions relevant to your business. Go ahead and try it; I'll wait. Back? You're now talking to the future.

Wave is not Watson (nothing except Watson is Watson). Wave now focuses entirely on mobile innovation with features designed specifically for smartphones that allow users to access, integrate, and visualize data generated by Salesforce and partner applications. There is a clearly documented ROI for analytics. In a 2015 interview with CRM Magazine, Stephanie Buscemi, Salesforce's senior vice president of analytics, cited two studies that show the growing influence of analytics and mobile, where IDC found a 78 percent higher ROI for companies that invest in analytics and another subsequent study found that 60 percent of online activity now comes from mobile devices.⁵ Wave Mobile Connector allows Salesforce customers to start exploring analytics from their phone by importing raw data and transforming it into visual charts or graphs. You can also use Wave Mobile Dashboard Designer to build multifaceted dashboards for visualizing dynamic information from multiple sources. Just follow the five steps mentioned earlier in the chapter.

In short, Adam Bataran, senior director of analytics for Bluewolf, believes the new mobile tools that can be used from any device are able to unleash our clients' data and increase customer intimacy. He further observed that analytics enables companies to know their customers and to make every customer interaction relevant, engaging, and personal. In the end, success with analytics is contingent upon capturing the right data at the right moment and transforming it into valuable insights that improve employees' actions both across an organization and with customers. That has always been the case and I have long harped on that; now it is just a lot easier to do.

Coming back to our discussion about the current onslaught of data, the IoT is ramping up fast. This is where smart devices communicate with other devices, servers, platforms, and whatever else you can think of. The device can be a smartphone or any variety of sensor, appliance, or electronic device with enough intelligence to communicate over an Internet protocol (IP) network. As adoption grows, it will generate enough data to drown you, if you are not drowning in data already. In fact, it can generate so much data that human staff will not be able to keep up. Instead, automation will be required and predictive analytics will steer staff to what is important and needs to be acted on. Remote management and monitoring, alone or combined with predictive analytics, have already emerged as the initial big payback use cases. This is an area that is just emerging, and already the valuations being projected for it—well into the trillions and tens of trillions of dollars—are staggering. Once things can talk to other things, especially over the cloud, almost anything is possible.

For example, what if your product could tell you how your customers were actually using it? Not only marketing folks, but product design and development teams could use this insight to design better, more useful products, products that would work the way customers actually used them. What if your product could automatically call technical support when it sensed a likely failure? It wouldn't take much for intelligent connected systems to connect the dots and have a replacement part ordered, schedule a technician to arrive when the ordered part does, and alert the customer every step of the way. IoT has the potential to change our world in ways we can barely begin to imagine: new ways to go to market, new business models, new capabilities intended to attract completely new market segments, and segments never previously addressed.

Executive Action

With the widespread availability of more and better data combined with easy analytics, we should anticipate what many describe as the arrival of a “new age” in almost religious terms. Overall, most executives will welcome the availability of data and analytics. Some will want to use this data to test new ideas and strategies and conceive new products and services. Heck, the data itself, particularly IoT data, might morph into a new value-added service. Would customers want to know how their people were or weren't using certain products and services, and would they pay a small fee for that insight? Maybe; at this point I don't know, but it's an exciting possibility.

However, those who continue to insist that their gut instincts are always right might not be so happy, especially when those gut decisions are contradicted by the data or proved wrong in practice. If they can't outright avoid data-based decision making, they will play dumb, make excuses, or ignore issues that may upset their gut. Sorry—keep a big bottle of Tums where it can be easily accessed.

Even scarier for this kind of manager will be the realization that there will be few or no places to hide and fewer excuses for wrong decisions or poor performance. There will be incontrovertible data that managers will have to explain, and results for which they will have to hold their teams and themselves accountable. Change is hard, and the arrival of highly accessible data with easy tools to access and understand it will be difficult for some old-school managers. The tools are so easy, however, that data shouldn't be hard to adopt. With Watson available as a free cloud service, there should be no excuse not to at least try data-based decision making.

For most modern managers the arrival of data-based decision making will be welcome, and it will benefit all, whether everyone acknowledges it or not. For example:

• CMOs can use data for nurturing along prospects not quite ready to buy or for leveraging data around more active prospects and customers to cultivate emotional connections with the company brand, product, and services.
• Heads of sales can use data to measure the effectiveness of the selling organization, its productivity, win/loss ratio, pipeline, quality of pipeline, and individual sales performance. In fact, they can show sales reps how to use particular data nuggets to help close a sale or upsell and cross-sell an existing customer.
• Product development can use the data to better understand what customers want, what they actually use, and how they use the products and services. That can't help but lead to better, more appealing products.
• C-suite executives should use the data to make sharper decisions, identify new areas of opportunity, and address C-level concerns like security and privacy, liability and risk.

In short, there is no area of the modern enterprise that can't benefit from a data-based approach. Facilities management, human resources, tech support—every aspect can be optimized in any number of ways. It's just a question of collecting the appropriate data and looking at it from a fresh perspective, which is what we have analytics for.

It all comes down to the democratization of data. Data is no longer a rare, precious commodity that must be rationed. Today, we have so much data we can drown in it if we aren't careful. Nor can we use it up or need to fear running out. Data is a valuable asset that continually renews itself. The more we share our data assets with the people who work with us and can take advantage of it, the richer we become. They will take that data to enhance the customer experience, employee experience, or partner experience, and the organization will be better off.

Data is never ending. Ninety percent of the world's data has been created in the past two years, and its growth isn't slowing down—the global economy creates 2.5 quintillion bytes of data each day! (FYI, a quintillion=1 followed by 18 zeros) Not only is this massive amount of data collected but it can be analyzed by employees in any department, whether it's sales, marketing, or leadership.

This is the democratization of data, and it will have a profound impact on the business, especially the office culture and specifically on office politics. Yet why aren't more businesses decisions driven by data?

Many organizations today are blind to metrics around cost, high-margin products, and employee productivity. They either don't measure it, or the data is collected and then buried in finance departments. When teams get together for budget or strategy sessions without that data in real time, conversations often devolve into conflict, sometimes described as dueling spreadsheets with conflicting numbers. Rather than focusing on what's objectively best for the business, politics dominates those discussions. I guess many just prefer to trust their gut.

On-demand access to market and internal data can add a layer of extreme objectivity to negotiations and strategic planning. With analytics tools like Salesforce Analytics Cloud or Watson, a team's decisions won't depend upon the loudest or seemingly most authoritative voice. The most persuasive arguments will be supported by clear, clean-cut numbers—without them, your points won't pass muster.

However, don't be naive; presenting this objective data won't automatically drive consensus or bring nirvana. Each executive has a level of accountability around their data. Each member of your team may interpret the data differently or apply it to another scenario. Not everyone will agree, nor should they. Data insights shouldn't replace debate and discussion in decision making, but instead enhance their effectiveness.

For that reason, invest in analytics to get closer to your customers than ever. In the next five years, we'll continue to see refinements and expansions of current analytics technology. Take the next step beyond just collecting data. Leave highly charged office politics behind and make informed, data-based business decisions to provide more value for your customers and your business.

As soon as you finish this chapter, start jotting down the problems you are wrestling with right now and what information would help you solve that problem if only you had it at your fingertips. Ironically, that information probably is sitting in your organization. It might be buried with some other information or need a little organization to make it apparent, but rest assured it is there and closer than you thought. Just look for it. And when you find it, remember: your data and analytics goals need to be driven by business outcomes, not the other way around. If you can't clearly define what it is you need from your data, then restate your need differently, dig deeper for the information, or design a system to provide it. All of the data in the world won't help you provide a better customer experience until you start asking the right questions or pose those questions in the right way. But when you do, you'll be amazed at how much better the customer experience and your business will be.