Top-Down Demand Map

The best opportunities for useful data products come at the intersection of important and frequent decisions and situations in which better data leads directly to better decisions. Your organization may decide to enter a new market, but if this decision is guided by executive experience and happens infrequently, then there is little a data product can add. In contrast, how a sales team decides to discount for new customers is an area that may happen often and can be informed by an understanding of the long-term value of the new customer.

The following prioritization process helps identify where to start:

1. Build a comprehensive list of processes, functions, and decision-points that happen in your organization. You may want to start from the first point you interact with a customer and trace all the interactions directly or indirectly related to the customer. Consider functional areas such as human resources, finance, supply chain, operations, and sales.
2. For each element, rate it on a scale of one to five for importance to the success of your organizational goals. One means the element has little to no impact on organizational success; five means it is a core and critical decision.
3. For each element, rate it on a scale of one to five for the value better data can bring to decisions. One means data brings little to no value to the decision process; five means data always makes the difference between good and bad decisions.
4. Plot all the elements in a chart, as shown in Figure 7-3.

   

In this chart, the further an element is from the lower-left corner, the more value a data product will bring.

Grassroots Needs

The alternative to a top-down mapping exercise is to let the demand for data needs surface from the people who want it most. As a leader, how do you hear what the “market” needs? Encourage people to ask difficult questions and create an environment where these questions are embraced.

Warren Berger, author of A More Beautiful Question (Bloomsbury, 2014), suggests that encouraging questioning in your organization can start with the leader but is a grassroots effort:

1. “A culture of inquiry starts at the top—with leaders who question.”
2. “Questioning should be rewarded (or at least, not punished).”
3. “Give people the time and space to question deeply.”
4. “Provide the tools to question well.”

The last point is of particular practical use. Approaches like “The Five Whys” provide a proven model for getting people to think more deeply about their circumstances. When they do, they are likely to wonder why decisions are made the way they are made and whether better information would improve those decisions.

Where to Begin

Understanding where you need better visibility within your organization can be complex for a large university, a growing public company, or for a small start-up. The answer starts with a simple question: What goal—if achieved—would cause your organization to celebrate in the next 12 months?

With that answer in mind, you can define what you most need to measure to track progress toward this goal. Perhaps there are a number of factors that must come together for the goal. Data products focused on this goal will give you confidence that you are moving forward on the set of crucial factors—or where you need to make adjustments.

Design

Less than 30 percent of the potential users of organizations’ standard business intelligence (BI) tools use the technology today. This low take-up is due to the fact that longstanding tools and approaches to BI are often too difficult to use, slow to respond or deliver content of limited relevance.

—Gartner⁵

The three reasons cited by Gartner for this problem are:

1. Ease of use (“is hard to work with”)
2. Performance (“users are frustrated by delays”)
3. Relevance (“does not express content in line with their frame of reference”)

The first and last reasons link directly to issues of poor data product design.

In our role as dashboard and analytical application designers, this is an area that is close to home. We see it all the time: reports and dashboards that lack focus and a message that targets their audience, which is often undefined. We see poor choices in data visualization that distract from the important elements in the data and put the burden of deciphering meaning on the readers. We see data products that lack an obvious starting point and logical flow to conclusions.

Poor design is wasteful. It results in solutions that users don’t want to use, as noted by Gartner. It wastes the audience’s valuable time as it struggles to comprehend the data. And it wastes the development and distribution efforts necessary to deliver the data product.

Overcoming these design shortcomings requires systematic processes and capabilities within your organization. Chapter 5, “Data Authors,” discusses the skills involved with designing effective data products. But how do you incorporate these skills into your processes and create repeatable ways to ensure quality design?

Objective

Better information design in your organization will come from a combination of skilled people and processes that ensure design is a priority.

The Apple App Store has strict guidelines for applications that are made available to customers. Apple publishes design guidelines, enforces a rigorous review process, and provides tools that help app developers get their designs right. As a result, consumers of Apple applications get a consistent, high-quality experience. The design focus allows novice users to quickly learn a new app because they have come to understand common conventions and design practices.

You should consider four areas when building an environment that infuses good design into your data products:

1. Train your data authors on the design skills for communicating data. Chapter 5 describes many of the areas of content that are foundational knowledge. The best information designers are discriminating in what data they choose to present and thoughtful about choosing metrics that encourage appropriate behaviors. They have the ability to create a logical flow through a series of attractive and clear data visualizations. There are also practical design skills to teach. For example, we always create wireframes—rough mock-ups of user interfaces—before building the actual data presentations. In this way, we can experiment with layout and content using rapid, lightweight iterations.
2. Incorporate designers early in the development process. Many of the organizations we work with go from concept to development of data products without considering the best way to communicate the data. When an executive wants a dashboard with a set of required metrics, the data analyst gets right to work on putting all the content together. The analyst might meet the specific requirements, but not in a form that anyone wants to consume. One client referred to these types of reports and dashboards as “ransom notes” because of the great variation in styles on the page. Designers should be included early in the requirements gathering process—not at the last moment to try to polish graphics.
3. Document guidance to ensure consistent designs. Many of the best practices and information design advice shared in Chapter 5 can be gathered into documents for designers. We have had success working with clients to create style guides for information design and communication. In addition, templates for reports, presentations, spreadsheet models, and dashboards can give designers a consistent starting point for their work. We also make sure to capture design patterns and principles as a way to establish models for approaching similar problems.
4. Review process to maintain quality control. Your organization should set expectations for the quality of data products that are used to communicate your valuable data and share insights. The costs—confusion, frustration, wasted time, and misleading information—are not worth tolerating poor design. The skilled designers, armed with the documented guidance, may form a review board to evaluate and provide feedback on the design of data products in your organization.

Start with a Style Guide

Good design builds on consistent, fundamentally sound choices and learns from the quality work of others. This is where your organization can start its efforts to improve data presentation design.

In our experience, an effective first step is to create a style guide for the presentation of data. It can be modeled after the branding style guides that many large organizations create to ensure a consistent presentation of content in print and online. Your style guide should provide data product authors with direction on choice of colors, typography, chart selection, and number formatting. The document provides examples of well-designed data visualizations and sets the rules for legends and labeling. In Appendix B, we have included an outline for such a style guide along with examples of the content.

“Imitation is the sincerest form of flattery,” said Charles Caleb Colton. Your data product designers should have a standard style guide and set of illustrative data products they can reference to understand quality norms and the expectations of organizational leaders. To do so, select a few examples of dashboards or reports that set a high bar. Find opportunities to publicly present these data products—setting a standard for clarity and content.

Develop

There are dozens of technology solutions for developing data products. Microsoft offers a variety of options: PowerPoint for data-rich presentations, Excel for models and ad hoc analysis, Power BI for self-service analytics, SQL Server Reporting Services for reporting, SharePoint with dashboards, and undoubtedly many more. The fact that one organization offers numerous tools for communicating data says a lot about the complexity of choosing the right tool for your unique situation.

Ideally, you want to have a small set of data tools that support the variety of types of data products your organization needs. A single solution is unlikely to offer the breadth of capabilities necessary. In our experience, four to five tools for data presentation are usually sufficient for most organizations. The following table describes a set of tool categories and the specific solution we’ve seen work most successfully.

^*There is a natural cringe-reaction for many people when they think of PowerPoint or Excel as tools for communicating data. While these tools have their flaws (most notably the encouragement of poor chart choices), the critiques should focus more on how individual people have misused the tools rather than the general capabilities of the tools themselves.

There are many forms your data products may take. And for every form, there are many technology options. However, here are some common features that are worth evaluating in almost every case:

• End-user customization—Some presentations may target a single audience. This is the exception to the rule. Most often, a data product goes out into the world alone and is used by many people, each of whom comes from a unique perspective. Whether it is their department, region, or products, all audience members will want to see data that is customized and scoped to their situation. Many interactive applications can support this ability to filter to the relevant data.
• Sharing support—Data should spur conversation. However, some solutions for data products create an isolating environment. The data product should make it easy to share, discuss, and capture insights—whether the discussion happens online, offline, on a desktop, or on mobile devices.
• Quality visualization—It matters how data is visualized. Clean, clear charts can make it easy for readers to quickly understand the data. The default settings for data visualizations should adhere to the fundamentals shared by well-known data visualization authors like Stephen Few and Edward Tufte.
• Fit workflows—Finally, we believe it is important that data products integrate into how people do their jobs. If the consumer of data is constantly on the run, bombarded by information of all types, an effective data product will deliver simple, narrow content to this person. If the consumer wants to drill deeply into the data to understand underlying assumptions, the functionality should exist to allow for this need.

“It’s a Poor Craftsman Who Blames His Tools”

The tools you have today probably are a suitable start for developing data products. Better to start with the investments you’ve already made and the technologies your team is comfortable with. Communicating data is far more about the people creating the data products than the technology tools they use.

Start by refining the skills of the individuals using these tools—even if they are simply Microsoft Office products like PowerPoint and Excel. The design skills and capabilities described earlier in this chapter and in Chapter 5 have little to do with the specific technical solution. Furthermore, many tools—Excel in particular—can be extremely capable with practice and creativity. For example, we built the visualization in Figure 7-4 using Excel, inspired by a hand-crafted visual shown on The New York Times website.

Discover

Knowledge is of two kinds: we know a subject ourselves, or we know where we can find information upon it.

—Samuel Johnson, long before Google

We’ve never seen an organization that is short on data. Nor have we seen one without volumes of reports, analytical tools, data models, and dashboards. It isn’t that the right data doesn’t exist; it is that individual people may not know where to find it or how to access it.

This is the data discovery gap. The gap between supply and demand for data often is the result of, ironically, incomplete information. Just like imperfect information leads to price distortions in the economic marketplace, imperfect information within an organization leads to organizational data inefficiencies. The person who could most benefit from an understanding of customer lifetime value may not know that a well-researched analysis already exists. Better data discovery also has the added benefit of revealing duplicate, competing, or conflicting data products.

Objective

Data product discovery should mirror the capabilities of online content subscription services. Podcasts, blogs, or Twitter, all have established features for ensuring an audience can find and access the latest content. These include:

1. Searching of metadata about the content, including title, author, and description
2. Browsing of content sorted into categories and ranked by popularity or ratings
3. Surfacing of related content based on the consumer’s expressed areas of interest
4. Subscribing to allow consumers to sign up to receive updates to content
5. Automatedpushing that allows consumers to receive updated content automatically rather than having to remember to return to the source
6. Using user permissions to control who has access to applications and content

Organizations are starting to recognize the need to provide a centralized place to find and access applications such as data products. Gartner expects that by 2017, 25 percent of enterprises will have an enterprise app store for managing corporate-sanctioned apps on PCs and mobile devices.

An app store can be a natural way to share new applications within the enterprise, recognize great applications, provide feedback to development teams, and even create a bit of competition between them—all to drive the development of better solutions.”⁶

—Gartner

Where to Begin: A Centralized Inventory of Data Products

Before building an enterprise app store, there are easy, low-tech ways to deliver discoverability to your audiences. For example, a centralized inventory of all the data products in your organization is a start. The inventory should catalog the product author, update schedule, content area, source data, and directions for subscribing or accessing the data product. A shared drive and common naming conventions will go a long way to making it easier for data consumers to find what they need.

Discuss

In a vacuum, analysis and insight amounts to little. But when those results are discussed and debated, they have the potential to be acted upon. Data must get mixed with people.

The best data ecosystems don’t simply assume discussions will occur. They encourage discussions through mechanisms for sharing, capturing, and saving information and insights. The discoveries found in the data are treated as precious assets—after all, they are the purpose of all the effort put into creating data products. Finally, the ecosystem should encourage people to take action when the discussion is complete.

Some organizations consider data products a one-way information broadcast. They implicitly assume that a dashboard is intended to deliver an information result, not drive an action.

Consider the evolution of television and its audiences as an analogy. Before the pervasiveness of social media, television was a powerful medium for telling a story, but there was little room for engaging the audience beyond the broadcast. “Water cooler” conversations sometimes occurred after a big event with spiked TV rankings—when J.R. was shot or after the last episode of “M.A.S.H.” While there are some examples of early fan engagement via letter writing and call-in voting, generally the focus was always on the one-way delivery of a show to a broad audience.

Now things are different. TV shows have to build stronger connections with smaller audiences. To do so, the broadcast is only the beginning of the conversation. Loyal fans and TV critics write detailed reviews of every episode. Theories and predictions are hashed out on fan discussion boards. The analysis and collective synthesis of the TV show deepens the audience’s understanding and engagement with the content. It took many years for television executives to appreciate and want to foment audience discussion. Now it is a crucial element in a TV show’s success.

Objective

An ecosystem that encourages discussion first needs the capability to capture insights found in data products. No user is an island; they need to take useful information they find and share it, with the context in which they found it. We call this concept snapshots. If an effective data product is a guided safari through the data, then a snapshot is a picture of a trip highlight by the data tourist at a point in time. And like tourists, they will want to save these pictures, send them to relatives at home, and organize them into albums that can help recall the experience.

The technical mechanism for capturing snapshots will vary. It may be something as simple as taking screen shots of a dashboard or saving a graph from a report as a separate file. More important, these snapshots are best when accompanied by a note that describes what makes the information interesting.

Beyond capturing insights, a data fluent organization supports and encourages the discussion that is spurred by data. A monthly report on customer attrition may raise eyebrows when the trend line shifts up, but what happens next? What follow-on discussions occur to dig deeper and start to create a plan for action?

The content of the data products can help encourage their audience to discuss and take action. Often there may be direct and well-understood links between certain data or charts and suggested actions. In the case of increasing customer attrition, the organization may have learned that there are common underlying factors that contribute to this issue (for example, competitors are running promotional campaigns, recent service problems, or outages). This information should be integrated into the data product.

Information and insights are just the beginning of the journey. The DIKW Pyramid (Data-Information-Knowledge-Wisdom) in Figure 7-5 is a framework from the realm of information science. It emphasizes the hierarchy of understanding, the narrowing of volume, and the elevation in value as data is eventually transformed into wisdom.

The true objective for facilitating discussion should be to gradually synthesize the information extracted from data products into greater organizational knowledge—shared expertise and intuition—and eventually into wisdom—the ability to apply that knowledge effectively to bring more value to the organization.

Where to Begin: Create a Place to Capture Insights

Discussions on data—like most of data fluency—are more a social and human problem than a technology problem. The technology approaches may be simple. For each data product, create a document or folder for capturing insights. The document may simply be screen shots of the relevant part of the content along with an annotation explaining why it is interesting. As a historical artifact, this document will reveal patterns of common issues and best practice approaches for responding to those issues. More important than a complex technology solution is an organizational culture that encourages dialogue and action after the insights are first found.

Distill

The scourge of data in most organizations is the ever-growing collection of reports that get generated month after month. New reports are created but seldom killed. The mass of data products quickly becomes difficult to navigate and the right information is hard to find. Even for the data that has found a suitable audience, there is seldom a feedback loop. The direction and needs of the organization may change, yet the content does not change to fit evolving demands.

Data products should be living documents. They should improve over time or be removed if they are no longer relevant. It is a matter of survival of the fittest.

Data fluent organizations recognize that too much content—particularly data content—will clog up the channels of communication. The data products must be distilled to the essential information. You want to clean out distractions and emphasize the most useful remaining parts.

To filter and clean your data product ecosystem, you need processes in place to gather feedback from your users. The feedback needs to impact how data products are designed and produced.

Learning from Wikipedia

Possibly the best example of an online living “document” is Wikipedia. Although many people benefit from the front end of Wikipedia, there is a whole ecosystem undergirding the popular website. The data that is viewed by all is created and managed by a network of administrators, bureaucrats, stewards, reviewers, and other specialty groups. The ecosystem governs the content and makes public all edits on any page identifying the author. Each page and piece of data is vetted and improved by the power of collective knowledge. Wikipedia, a nonprofit organization, is fueled by voluntary effort. Yet, few large organizations have figured out how to harness the collective knowledge of their employees or associates.

Objective

The goal for a data fluent organization should be to have multiple mechanisms to provide feedback on data products to ensure they evolve as needs evolve. Those data products that don’t continually change to fit audience needs may slip into obscurity and be discontinued. Following are three basic approaches for distilling your best content:

1. Create a lightweight feedback mechanism. This should be a simple process that allows users to rate data products. Did you find this data product valuable? What information was most useful? What decisions did it influence? The results from a few questions would provide the authors with valuable guidance on improving what they deliver.
2. Track usage. For online products, you may find it straightforward to find out what content people are engaging with. Which reports are downloaded most often? How many people viewed the operational dashboard every day? Like content on a website, these usage metrics are a direct way to understand what is popular or not.
3. Conduct content reviews. From time to time, gather the audience for a data product and have an open discussion about what they find valuable and what features or data they would like to see added. To avoid the inevitable slide toward more scope, it is useful to create an artificial constraint (for example, number of metrics or pages) to force users into hard choices about what they need versus what they want.

What Can You Do Without?

With an objective to bring greater focus to your organization, consider the following simple plan to narrow the volume of data products being produced. First, for each functional area build a list of all the data products currently being used. Second, create a simple survey based on this list. The survey can have a single question: “What three pieces of information would you recommend to a new employee in your role?” Third, determine what data products received no votes. These are the data products that need to prove their value. Would anyone miss them if they were gone? In the final step, stop distributing these data products and see if anyone complains. If the data is being presented but not used, it is a good bet your resources are better focused elsewhere.

“Only Connect”

Because knowledges are so specialized we need also a methodology, a discipline, a process to turn this potential into performance. Otherwise, most of the available knowledge will not become productive; it will remain mere information. To make knowledge productive, we will have to learn to connect.

—Peter Drucker, Pro-Capitalist Society

Drucker is referencing E.M. Forster’s belief in the imperative of human connection. This chapter is about connection and how to create the conditions that facilitate connections. The six Ds are pillars that ensure data gets turned into productive action, connecting people to the information that leads to smarter decisions:

• Demand is the connection of your organization’s needs with the people who will produce data products.
• Design ensures quality outputs that will communicate to the audience.
• Develop is the mechanism to create data products, the connective glue that will deliver useful data to the people who can use it.
• Discover is connecting supply and demand after data products have been created.
• Discuss is the collaboration between people to understand and digest the meaning of the data.
• Distill is the feedback loop that ensures organizational needs are heard after a data product is created.

An ecosystem with these elements in place will allow data to flow freely, ensure quality communication, and improve data products over time.

The conditions for a data fluent ecosystem are far from trivial. In fact, in our decades of experience working with organizations that value data, we have never seen an organization that has put all these elements together. To help on the journey, we have included suggestions for manageable places to start.

This chapter completes our tour of the four quadrants of the framework for data fluent organizations. Next this question is answered: Where are you on your journey to data fluency?