Cover Page

The Sentient Enterprise

The Evolution of Business Decision Making

Oliver Ratzesberger
Mohan Sawhney

Wiley Logo

Foreword to The Sentient Enterprise

The Sentient Enterprise is both a good book and a good sign. Let me explain each component of that proclamation further.

It’s a good book because two very smart and experienced gentlemen collaborated to produce some excellent advice on data and analytics. I’ve known both authors for over a decade, and they are, individually and collectively, powerful thought leaders.

Individually, either of these fellows could write an excellent book (and Sawhney has already written several!). Oliver Ratzesberger is one of the most visionary thinkers today on IT architecture issues. I was dazzled by the data architecture he created at eBay, and he’s risen quickly to be the head of product at Teradata. That affiliation doesn’t mean that he only espouses Teradata solutions; he’s long been a strong advocate of open architectures and open-source tools like Hadoop.

Sawhney, trained as a marketing professor, is clearly much more than that. He’s been a leading thinker for many years in the areas of innovation, digitization and e-business, networked organizations, and many other topics. If you were looking for a coauthor to think strategically about technology, you couldn’t find anyone better than Mohan.

Collectively, the authors’ expertise yields a unique set of topics that appeal to both strategists and practitioners—each of whom have an important stake in the journey toward sentience. One moment you’re reading about how the massive explosion of data will reshape business and society. The next thing you know, you’re learning how to prevent uncontrolled proliferation of virtual data marts with time-limited lifespans, or the correlation between call duration and customer sentiment in call centers. No one could accuse these authors of dwelling solely in either the strategic or the tactical realm.

The Sentient Enterprise is also a good sign that analytics are becoming a mainstream, professionally-managed activity—at least if organizations practice the lessons in this book. Historically, the creation of analytics was a somewhat unstructured and “artisanal” activity, driven by individual human hypotheses and curiosity. The analytical outcomes might be implemented by a decision maker, or they might not. There was no vehicle for embedding them into business processes and systems, or for learning from them across the organization.

But with the five “platforms” described in this book, analytics can move well beyond the artisanal stage. They are a linked set of capabilities for making analytics operational, production-based, and shared across the enterprise—a high-volume manufacturing operation for data and analytics-based decisions. They suggest a future—and a present in a few companies—in which analytical decisions are produced automatically and embedded into decisions and actions without much human intervention. For anyone who believes as I do that analytical decisions tend to be more accurate and less biased than those based on “the gut,” this is great news.

These approaches won’t be adopted for all decisions, at least not for a while. They currently apply to repetitive and tactical decisions based on a lot of data—much of it Internet-based “behavioral data,” as Ratzesberger and Sawhney refer to it. In advertising, for example, versions of the platforms are already being used to support digital marketing decisions. But they won’t be applied to decisions about Super Bowl ads anytime soon.

It’s also important to remember, as the authors point out, that these platforms aren’t just based on technology. Platforms consist of both technical and human capabilities. If you want the technical capabilities to succeed, you need a set of human skills, behaviors and attitudes that will make them possible. These human factors come into play both at the executive level—supporting development of the technical capabilities and their fit with business processes—and at the front lines as well.

In health care, for example, we desperately need approaches to “precision medicine,” and they will require all of the five platforms that these authors describe. But part of those platforms are senior managers who are willing to sponsor the development of the platforms, and doctors and nurses who are willing to use them and work with their recommendations and decisions on a daily basis. It’s anybody’s guess whether developing the technological or the human capabilities will be more difficult, but both will be hard.

So get busy and start building these five platforms if you haven’t done so already. Absorb both the strategic and tactical advice in this excellent book. And start preparing the wetware in your organization for a future in which analytical decisions are produced like high-quality widgets.

Thomas H. Davenport

Distinguished Professor, Babson College

Fellow, MIT Initiative on the Digital Economy

Author of Competing on Analytics and Only Humans Need Apply


This is a book about business technology and business culture. Specifically, it’s about how the right combination of technology and culture can transform the use of data and analytics so that even the largest organizations achieve new found levels of agility, insight, and value from their information sources.

This book is also written for a very wide range of business professionals. By that we mean not just senior technology executives and data scientists, but also business users, anyone who might have “analyst” in the job title, and pretty much everyone whose role is impacted by how data is gathered, analyzed and applied in the organization.

Whether you’re establishing the next-generation digital strategy, setting up data experiments to explore deep neural networks, or establishing controls for access to your corporate KPI dashboard, this book is for you. Our goal is to build bridges across job functions and departmental silos to solve common challenges that most business professionals will recognize—challenges such as:

“How can we stop multiple teams from pulling information into their own data silos and then spending all our meeting time wondering why everyone’s data doesn’t match up?”

—Data scientist at a major auto manufacturer

“Just because we’re big doesn’t automatically mean we’re the best; what’s the best way to leverage our economy of scale while remaining agile?”

—Chief data officer for a telecommunications giant

“Why is it that my kids at home have self-service apps on their phones to build their own games, but I have to go through IT and a long requirements process every time I want to experiment with data?”

—Product testing analyst at an electronics manufacturer

“Given that our clients rely on us to be there tomorrow with the innovations they need, how can we get on a more predictive curve so any success we have today isn’t just on borrowed time?”

—Senior VP for a profitable global networking company

These are tough questions from the many business perspectives you’ll find across any company that relies on data (and in today’s information-driven economy, which means pretty much any company at all!). Furthermore, these questions are not hypotheticals. They happen to be actual challenges relayed to us by top executives—from Dell, Verizon, General Motors, Siemens, Wells Fargo, and nearly a dozen other organizations we interviewed for this book—about the challenges they and their colleagues face on a daily basis.

Fortunately, these companies came up with innovative and scalable analytic solutions to address these challenges. In the pages to come, we’ll examine these success stories and combine them with our own research and emerging best practices in big data and advanced analytics. In doing so, we’ll chart a journey through what amounts to a new model for analytic capability, maturity, and agility at scale—something we call the Sentient Enterprise.

At its core, the Sentient Enterprise will change the way everyone in business makes decisions—from small, tactical decisions to mission-critical strategic decisions. We’ll chart the path that technology and all of us who leverage it are taking to become more productive. The journey is as complex as it is valuable, so we’ve organized the Sentient Enterprise into a capability maturity model with five distinct stages:

  1. The Agile Data Platform as the technology backbone for analytics capabilities and processes. Here is where outmoded data warehouse (DW) structures and methodologies are shifted to a balanced and decentralized framework, incorporating new technologies like cloud and are built for agility. Virtual data marts, sandboxes, data labs, and related tools are used in this stage to create the foundational technology platform for agility moving forward.
    Diagram shows five stages in the Sentient Enterprise agile data platform, behavioral data platform, collaborative, ideation platform, analytical application platform, and autonomous decisioning platform.
  2. A Behavioral Data Platform that captures insights not just from transactions, but also from mapping complex interactions around the behavior of people, networks, and devices. Here is where enhanced job functions for the data scientist start to emerge. We also loop in CXOs and orient them to think in terms of behaviors and ultimately a customer-centric model. As we build this platform, Net Promoter Scores and other measures of customer sentiment and behavior get elevated to mission-critical importance for the enterprise.
  3. The Collaborative Ideation Platform to let enterprises keep pace with the data explosion by socializing insights across the community of analytics professionals. With this platform, democratized data, crowdsourced collaboration, incentive-based gamification, and social connections within the enterprise can be leveraged together to connect humans and data in a fast, self-service manner that outperforms traditional centralized metadata approaches. As part of this platform, we build a “LinkedIn for Analytics” environment to analyze how people both use and talk about data in the organization. This includes social media conventions to see which ideas, projects, and people get followed, liked, shared, and tagged.
  4. The Analytical Application Platform to leverage the simplicity of an exploding app economy for deployment of analytical capabilities across the broader business user community and to boost enterprise listening. In the process, we move away from static applications and extracting, transforming, and loading (ETL) in favor of self-service apps and self-awareness through enterprise listening. Visualizations now become more than just a pretty picture on an executive’s wall; we instead put these visualizations to work to drive change and act on insights.
  5. The Autonomous Decisioning Platform, where true sentience is achieved as the enterprise starts to act as an organism to make more and more tactical decisions on its own—without human intervention—so people can put more focus on strategic planning and major decisions. In this platform, we go beyond predictive technologies and increasingly deploy algorithms, machine learning, and even artificial intelligence (AI) at scale. This enables examination of all data to detect trends, patterns, and outliers as real-time context for human analysts and decision makers about shifts in behaviors. We take the bulk of data sifting and decisioning off people’s shoulders and save human intervention for critical junctures. This is where true sentience is achieved in the enterprise.

While Chapters 3 through 7 deals sequentially with each of these five stages, it’s important to remember that the journey is an ongoing one, and there is no single point of entry or completion. Think of the Sentient Enterprise as less a finish line than a North Star to guide your quest toward the strongest possible agility and value around data. The good news is that you don’t have to do it all—and you don’t have to do it all at once—in order to find plenty of big wins along the way.


Data is driving progress across all kinds of industries, but too many people—from analysts and business users to top C-suite decision makers—still don’t know enough about how to innovate with it.

A few decades ago, this was enough rationale for most information technology (IT) leaders to dole out resources to the rest of their company colleagues through a stately and slow requirements-driven process. Sometimes that approach is still necessary. But in a world where every home, pocket, and purse has countless real-time and self-service apps, many companies are embarrassingly behind the curve in making data and analytic muscle more accessible to the diverse workforce that needs these resources to innovate.

On an organizational level, failing to leverage data for innovation and decision support can put your whole business on a downward trajectory. Success today requires navigating a constantly expanding data universe, and companies that don’t fully embrace the data available to them are operating on borrowed time.

We’ll explore in this book how the five-stage Sentient Enterprise capability maturity model can put data in the hands of more business users, part of a broader revolution into how companies listen to data, conduct analysis, and make autonomous decisions at massive scale in real time. In the process, we’ll visit with top analytics professionals at some of today’s largest and most successful organizations—Verizon, Dell, Cisco, General Motors (GM), Wells Fargo, and Siemens, just to name a few—to see how this revolution has, in many ways, already begun.

“We remind ourselves every day—it’s even in our Company Credo—that being big is not the same as being the best,” said Grace Hwang, Executive Director for Business Intelligence and Advanced Analytics at Verizon Wireless, one of the top executives who gave extensive interviews for the writing of this book. “Our job is to leverage economy of scale—but at the same time to be nimble and proactive.”

In the pages to come, we’ve packed lots of real-world perspective from Verizon and other influential companies that have agreed to share their stories—their headaches and challenges, their insights and solutions—as they innovate their way to success. Throughout this book, in fact, we prioritize on-the-ground relevance and accessibility for a wide range of readers.

We’ve designed this book to be accessible and succinct for the lay business audience, with plenty of bread crumbs for more technophile information. While we are indeed talking about capabilities made possible by servers, nodes, data warehouses, and the skein of other infrastructure and software resources that go into any large analytics infrastructure, we do so from a perspective that’s not too wonky or overly technical.


Especially when working with many experts and massive infrastructure that might scale all the way to the global production level, it’s easy for collaboration to veer into chaos if you don’t have the proper platforms and hassle-free governance to help people stay in their lanes. But it’s important for people to still collaborate effectively with those in other parts of the business, so silos don’t develop as barriers to agility.

We’ll see in the chapters to come how that one word—agility—is key to getting the enterprise to the sentient point where it can analyze data and make autonomous decisions at massive scale in real time. Agile systems and processes enable this by loosening IT roadblocks, democratizing data access, breaking down silos, and avoiding costly inefficiencies like data duplication, error, and just plain chaos.

Merriam-Webster’s Collegiate Dictionary defines agile as “marked by ready ability to move with quick easy grace” or “having a quick resourceful and adaptable character.” In the corporate world, business agility is usually defined as a company’s ability to rapidly respond and adjust to change or adapt to meet customer demands. For our purposes, however, let’s entertain a more targeted definition:

Agility is the ability to decompose or break big problems and systems into smaller ones, so they’re easier to solve and collaborate around.

In our effort to build this new agile environment for analytics, we looked across many industries for other examples of agility. This cross-industry perspective can solve problems in one sector by looking to other kinds of business settings for challenges met and lessons learned. The context may be different, but the insights and solutions can be strikingly similar.

For instance, we can learn much about an agile decomposition approach to tomorrow’s data architectures by examining the Open Systems Interconnection (OSI) model that the telecommunications industry deployed as far back as the 1970s. OSI was developed to segment complicated infrastructure (wiring, relay circuits, software, etc.) into manageable chunks for better collaboration among various specialists.

By designing modular but interoperable parts of the system known as abstraction layers, OSI ensured that the work of software programmers, for instance, didn’t conflict with what engineers and line workers might be doing in the field—or vice versa. We like the OSI example because, even though it was developed four decades ago, the technique of segmenting big systems into overlapping but distinct and manageable elements is a powerful ingredient for agility—one that we continue to see in some cutting-edge settings today.

Check out a technology called Docker ( to see what we mean. Docker lets you break down the app-building process into a series of manageable steps. Through a simple Docker Engine and cloud-based Docker Hub, the company lets you assemble apps from modular components in a way that can reduce delays and friction between development, quality assurance (QA), and production environments. By breaking things down into smaller components, Docker aims to make the app-building process more manageable and reliable.

Another example is the entire “microservices” approach to building software architectures. Unlike traditional service-oriented architectures (SOAs) that integrate various business applications together, microservices architectures involve complex applications built from small, independent processes. These processes communicate with each other freely using application programming interfaces (APIs) that are language agnostic.

With microservices, you’re still building powerful architectures; but it happens more efficiently, with modular elements broken down to focus on discrete small tasks. As a result, microservices architectures can be tremendously agile. They facilitate continuous-delivery software development and let you easily update or improve services organized around distinct capabilities such as user interfacing, logistics, billing, and other tasks.


These examples show how we’re on a journey away from monolithic and nonagile IT applications. But a caveat along this journey—one we’ll emphasize often in the course of this book—is that you must fold in the right kind of governance, so your newly agile systems don’t create more problems than they’re solving. We’ll talk through the Wild West pitfalls of data anarchy and error that arise when we try to loosen old systems and rules without putting some kind of (seamless and hassle-free) governance in place to support our new and agile methodologies.

We’ll also see how most of the steps a company takes on the journey to sentience follow this definition of agility as decomposing problems into manageable components. The word is even embedded in the first of the Sentient Enterprise’s five stages—the Agile Data Platform—proof of how front and center agility needs to be for anyone looking to survive and compete in today’s data-driven marketplace.

The Layered Data Architecture is, in turn, a foundation for the five complementary platforms that make up the Sentient Enterprise:

  1. Agile Data Platform
  2. Behavioral Data Platform
  3. Collaborative Ideation Platform
  4. Analytical Application Platform
  5. Autonomous Decisioning Platform

By setting up an environment of five agile and closely linked platforms, we mature an organization’s capabilities around data. That’s why we refer to the Sentient Enterprise as a “capability maturity model”—not unlike the famous Six Sigma methodology for business processes and quality control—that many organizations can use as a yardstick for building capabilities and success.


As we’ll discuss in Chapter 1, the Sentient Enterprise is the result of two very distinct but complementary perspectives on big data and analytics. For one of us (Mohan), the understanding of data’s pivotal role in business has grown through many years as an academic researcher and corporate adviser. For the other (Oliver), the perspective comes from a long career as a technology executive and practitioner trying to manage the rewarding, often challenging, relationship a company keeps with its data and technology over time.

The two of us first compared notes in November 2013, and we have been iterating our way through this maturity model ever since. Our initial “aha” moment over a dinner meeting near Chicago was made all the more enticing by how fully our perspectives seemed to dovetail despite two very different careers. We aligned on how the right combination of technology, governance, and human engagement can launch enterprise analytics to new heights while preserving the coveted start-up-style agility that tends to atrophy at scale.

What was missing, however, was a framework for understanding and optimizing this alchemy so it can be replicated by any company with the vision and resources to try. After several years of joint collaboration and research, that framework now exists as the Sentient Enterprise, and this book is its manifesto.

Consider the Sentient Enterprise your road map for harmonizing analytic power, business practices, and human dynamics in ways that have already begun to transform and supercharge what’s possible for big data and the industries that leverage it well. Inspired by our own progressive understanding of these trends over our careers and by the collaborative alchemy that has brought that evolution to the point we’re at now, we’re able to present the Sentient Enterprise to you as a viable framework for analytic agility, ready for application to your own business.