19,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Sprache: Englisch
Transform your company’s AI and data frameworks to unlock the true power of disruptive new tech
In From Data to Profit: How Businesses Leverage Data to Grow Their Top and Bottom Lines, accomplished entrepreneur and AI strategist Vineet Vashishta delivers an engaging and insightful new take on making the most of data, artificial intelligence, and technology at your company. You’ll learn to change the culture, strategy, structure, and operational framework of your company to take full advantage of disruptive advances in tech.
The author explores fascinating work being undertaken by firms in the real world, as well as high-value use cases and innovative projects and products made possible by realigning organizational frameworks using the capabilities of new technologies. He explains how to get everyone in your company on the same page, following a single framework, in a way that ensures individual departments get what they want and need.
You’ll learn to outline a comprehensive technical vision and purpose that respects departmental autonomy over their core competencies while guaranteeing that they all get the tools they need to make technology their partner. You’ll also discover why firms that have adopted a holistic strategy toward AI and data have enjoyed results far beyond those experienced by those that have taken a piecemeal approach.
From Data to Profit demonstrates the proper role of the CEO during an intensive transformation: one of maintaining culture during the change. It offers advice for organizational change, including the 3-Phase Data Organizational Development Framework, the Core <-> Rim 3 Main People Groups Framework, and the way to implement new roles for a Chief Digital Officer and Technical Strategist.
Perfect for data professionals, data organizational leaders, and data product and process owners, From Data to Profit will also benefit executives, managers, and other business leaders seeking hands-on advice for digital transformation at their firms.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 474
Veröffentlichungsjahr: 2023
Ähnliche
Table of Contents
Cover
Title Page
Introduction
A Novel Asset Class with a Greenfield of Opportunities
The Road from Laggard to Industry Leadership
Technical Strategy as a New Top-Level Construct
Playbook for the Enterprise
Systems, Models, and Frameworks
Introducing Data to the Enterprise
CHAPTER 1: Overview of the Frameworks
Continuous Transformation
Three Sources of Business Debt
Evolutionary Decision Culture
The Disruptor's Mindset
The Innovation Mix
Meet the Business Where It Is
The Technology Model
The Core-Rim Model
Transparency and Opacity
The Maturity Models
The Four Platforms
Top-Down and Bottom-Up Opportunity Discovery
Large Model Monetization
The Business Assessment Framework
The Data and AI Strategy Document
Data Organizational Development Framework
More to Come
CHAPTER 2: There Is No Finish Line
Where Do We Begin? With Reality
Defining a Transformation Vision and Strategy
Paying Off the Business's Digital Debt
Managing the Value Creation vs. the Technology
A Master Class in Continuous Transformation Strategy
Evaluating Trade-Offs
What Happens When the Business Loses Faith in Data and AI?
What’s Next?
CHAPTER 3: Why Is Transformation So Hard?
Cautionary Tales
Data-Driven Transparency
The Nature of Technology and FUD
The Business Has Been Lied to Before
Is It Sci-Fi or Reality?
The Coming Storms
Time Travel
Time Travel in the Real World
Data-Driven, Adaptive Strategy
What’s Next?
CHAPTER 4: Final vs. Evolutionary Decision Culture
Implementing Change and Taking Back Control
Paying Off Cultural and Strategic Debt
Playing Better Poker Means Folding Bad Hands
Fixing the Culture to Reward Data-Driven Decision-Making Behaviors
A Changing Incentivization Structure
What's Next?
CHAPTER 5: The Disruptor's Mindset
The Innovation Mix
Exploration vs. Exploitation
What Happens with Too Much or Too Little Innovation?
Innovate Before It's Too Late
EVs and Innovation Cycles
Putting the Structure in Place for Innovation
Building the Culture for Innovation
An Innovator's Way of Thinking
Managing Constant Change and Disruption
Preventing Data-Driven and Innovation from Spiraling Out of Control
What's Next?
CHAPTER 6: A Data-Driven Definition of Strategy
How Quickly the Innovators Became Laggards
Using Strategy to Balance the Scales
Redefining Strategy
Resistance and Autonomy
The Cost of Resisting Change
What's Next?
CHAPTER 7: The Monolith—Technical Strategy
The Business Model
A Few Examples of Business Models
The Need for Technical Strategists
The Operating Model
Scale to Infinity and Super Platforms
The Implications of an Automated Operating Model
The Technology Model
The Best Tool for the Job
Making the Connection to Value from the Start
What's Next?
CHAPTER 8: Who Survives Disruption?
Using Frameworks to Maintain Autonomy
Reducing Complexity While Maintaining Autonomy
Technology Cannot Solve All Our Problems
Making Decisions with Core-Rim and the Technology Model
Defining the Value Proposition
How Technology First-Businesses Scale
Can We Be Confident That Business Units Won't Be Completely Erased?
What's Next?
CHAPTER 9: Data—The Business's Hidden Giant
Does the Business Really Understand Itself?
Moving from Opaque to Transparent
Getting Deeper into Workflows and Experiments
Data Gathering and Business Transparency
Understanding the Workflow
Improving Workflows with Data
Designing a Better Framework
What's Next?
CHAPTER 10: The AI Maturity Model
Capabilities Maturity Model
Data Gathering, Serving, and Experimentation
Starting with Experts
A Race Against Complexity and Rising Costs
The Product Maturity Model
The Data Generation Maturity Model
What's Next?
CHAPTER 11: The Human-Machine Maturity Model
What Happens When Technology Adapts to Us?
The Human Machine Maturity Model
Hidden Changes as Models Take Over
Human-Machine Collaboration Is a New Paradigm
Holding Machines and Models to a Higher Standard
Understanding Reliability Requirements
What's Next?
CHAPTER 12: A Vision for AI Opportunities
The Zero-Sum Game: Winners and Losers
Near- and Mid-Term Opportunities
Best-in-Breed Solutions
Preparing Products for Transformation
Opportunity Discovery Gets the Business Off the Sidelines
Top-Down Opportunity Discovery
Monetization Assessment
Just Because It Can Be Built…
What's Next?
CHAPTER 13: Discovering AI Treasure
Bottom-Up Opportunity Discovery
Giving Frontline Teams a Framework to Leverage Data and AI
The AI Product Governance Framework
What Happens if No One Brings Opportunities Forward?
It May Be Bottom-Up, But It Still Starts at the Top
What's Next?
CHAPTER 14: Large Model Monetization Strategies—Quick Wins
AI Operating System Models
AI App Store
Quick-Win Opportunities
The Digital Monetization Paradigm
Understanding the Risks
What's Next?
CHAPTER 15: Large Model Monetization Strategies—The Bigger Picture
What Are the Costs?
How the Models Work
Flaws Are Opportunities
Disrupting College
Advanced Content Curation
How Microsoft Successfully Monetized Their $10 Billion Investment
Large Models Enabling Leapfrogging
Workflow Mapping Becomes Even More Critical
What’s Next?
CHAPTER 16: Assessing the Business's AI Maturity
Starting the Assessment
Culture
Leadership Commitment
Operations and Structure
Skills and Competencies
Analytics-Strategy Alignment
Proactive Market Orientation
Employee Empowerment
The Data Monetization Catalog
What's Next?
CHAPTER 17: Building the Data and AI Strategy
Defining the Data and AI Strategy
The Executive Summary
The Introduction
Strategy Implementation
Introducing the Data Organization
Next Steps
Needs, Budget, and Risks
What's Next?
CHAPTER 18: Building the Center of Excellence
The Need for an Executive or C-level Data Leader
Navigating Early Maturity Phases
The Data Organizational Arc
Benefits of the Center of Excellence Model
Connecting Hiring to the Infrastructure and Product Roadmaps
Getting Access to Talent
Common Roles for Each Maturity Phase
What's Next?
CHAPTER 19: Data and AI Product Strategy
The Need for a Single Vision
Defining Data and AI Products
The Business's Four Main Platforms
Leveraging Data and AI Strategy Frameworks
Workflow Mapping and Tracking
Assessing Product and Initiative Feasibility
Pricing Strategies for Data and AI Products
Problem, Data, and Solution Space Mapping
Managing the Research Process
The AI Evangelist: Community Building for Platform Success
What's Next?
Index
Copyright
Dedication
About the Author
End User License Agreement
Guide
Cover
Title Page
Copyright
Dedication
About the Author
Introduction
Table of Contents
Begin Reading
Index
End User License Agreement
Pages
i
vii
viii
ix
x
xi
xii
xiii
xiv
xv
xvi
xvii
xviii
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
325
326
327
328
329
330
331
332
333
334
ii
iii
iv
335
From Data to Profit
How Businesses Leverage Data to Grow Their Top & Bottom Lines
Vin Vashishta
Introduction
Businesses are thinking about technology backward. Many business leaders think that 2020–2023 brought about the fastest digital transformation in history, but in fact those three years were the slowest technology will ever change again. Technology and therefore business will never move this slowly again in our lifetimes.
Businesses today are in adapt-or-die territory. Apple CEO Tim Cook said, “We don't believe you can save your way to prosperity. We think you invest your way to it.” As such, Apple is looking for new markets to expand into, using innovation to drive growth. Amazon is slimming down and reallocating its war chest with the same goal. Both used smart investments during the Great Recession to become the companies they are today.
The last time Amazon snapped its fingers, half of retail disappeared. Amazon combined digital and web technologies to disrupt the retail industry and become a titan. Even retail giant Walmart seemed invincible until Amazon leveraged technology to reveal the weakness in Walmart's operating model.
Amazon is looking for ways to be a disruptor again. But Amazon won't be alone this time, and the disruption will touch every competitive industry, not just retail. Microsoft and OpenAI's partnership was the major technical story of 2022 and revealed the potential scope of AI applications. Nothing in recent memory has captivated people's attention in the same way.
ChatGPT reached 1 million registered users in just 5 days and soon after became the fastest application to reach 100 million registered users. GPT-4 has propelled OpenAI and Microsoft further into the public spotlight. Microsoft is looking to duplicate ChatGPT's growth trajectory by building Copilot into its suite of enterprise applications and regain its dominance in the space.
The fast-follower applications from tech giants such as Google, SAP, and Salesforce have created even more interest. With multimodal models, video, images, audio, and text are generated together seamlessly. The rate of technology change isn't slowing down; it's accelerating.
The more people use and share their results, the more front of mind the technology becomes for CEOs. AI was once viewed as speculative and often attracted indifference. Now data and AI are leading a wave of new solutions. New business models have emerged in the first half of 2023, and investment continues to flow into AI startups. Even in an economic downturn, AI is experiencing growth, with some speculating that it will drive a turnaround.
A Novel Asset Class with a Greenfield of Opportunities
In late 2022, I talked with a reporter and said cautiously, “This feels like the real thing.” Past false starts with data and AI have jaded the entire data science field. I was hit by the first hype cycle, the one most have forgotten, in the early 1990s. The data science career that I imagined took almost 20 years to materialize.
I am a skeptic, but even I must acknowledge how capable AI models have become. The first half of 2023 has proven cautious optimism to be warranted. However, the grandiose predictions will fail to materialize because this wave is focused on practical applications. But the time between a new tool being released, practical applications being delivered, and revenue materializing is getting shorter every year.
Data and AI are driving this wave of tools and products. The World Economic Forum classifies data as a completely new asset class, unlike any we've seen before. It's not the new oil. It's more potent than that. The rapidly improving product landscape is filled with examples of the new monetization paradigm.
Use a gallon of oil or an ounce of gold once and they're gone. A gigabyte of data can be reused to generate value across the firm. The models it trains are never depleted. They can return value for years with minimal overhead. Partnering with large model platforms opens even more opportunities.
A 2021 Accenture study discovered that companies with a holistic data and AI strategy generated 50 percent more returns than their peers. Over the last 2 years, dozens of surveys have been published with similar findings. However, it took real-world applications to turn skeptics into promoters.
C-level leaders spent 5+ years investing in data teams and infrastructure. There's a mountain of evidence that the ROI hasn't materialized for everyone. Inconsistent results have revealed that the problem doesn't have a technical or tactical solution. Hiring talent and buying tools support technology delivery, but something is missing.
Talk to data scientists, and the same story plays out repeatedly. The business doesn't get it, and leadership doesn't buy in. Data teams are developing reports and simplistic models. Their capabilities and value are trapped just like the value in the company's data. How do we unlock it?
It begins with technical strategy. This book focuses on data and AI strategy, but I will reveal the bigger picture that businesses need to survive in the modern competitive marketplace.
The Road from Laggard to Industry Leadership
When Satya Nadella took over Microsoft, the company was at a crossroads. The business thought of itself as the industry leader. Customers had the opposite opinion. This is where Google finds itself at now. Microsoft executed a brilliant turnaround, and Google needs to take a page from its book to navigate the challenges to its search dominance.
Microsoft adopted a challenger's mindset and abandoned hopes of regaining dominance in its traditional core business lines. It took a completely different approach to its development tools. Microsoft restructured the business around the next technology wave and went all in on the Azure cloud platform and building a services ecosystem on it.
Its partnership with OpenAI marks a new transformation to AI. Azure is the core infrastructure supporting the collaboration, and Microsoft has launched its GPT 3.5 and GPT 4–enabled services there. OpenAI leverages Azure to train some of the most complex models in the world.
Microsoft's cloud services began a long journey back to technical dominance, but it has always looked ahead with AI firmly in its sights. At the same time, transformation doesn't come in giant leaps. Digital yielded to the cloud, then data, and now AI.
Microsoft has realized that opportunities for revenue growth come in waves. However, a business built to monetize software and services would fail to monetize the cloud. A company built without best-in-class cloud capabilities would fail to deliver best-in-class AI products. Microsoft needed to transform again to monetize the models in development with OpenAI.
It has proven to be a successful technical strategy, and Microsoft will leverage what it has learned to transform in the future as new technological waves roll in. Microsoft has become AI-first this time, but the next wave is already on the horizon.
Azure is at the center of its OpenAI partnership. All its attention is funneled to products and services customers can buy today. In fact, Microsoft is teaching a master class on monetizing advanced models. I'm not talking about multiyear time horizons. Products hit the marketplace months after new models are developed.
Technical Strategy as a New Top-Level Construct
What is the key to Microsoft's growing track record of success? Satya Nadella is a technical strategist. VCs are now looking for the next generation of technical strategists. Investors have realized the connection between the hybrid capability set and business success. In 2022, at the World Economic Forum in Davos, Jim Breyer, CEO of Breyer Capital, said, “We invest in technical strategists.” We are at the front of a sea change in strategy and technology. Businesses need a new partnership between C-level leaders, strategists, and technical experts to unlock the value trapped in their data and organizations.
Supporting that partnership is a new top-level strategic construct, the technology model. Everything in the business that creates and delivers value to customers needs a strategy. Firms are transferring an increasing amount of their business and operating models from people to technology.
The transfer began with digital transformation and cloud adoption. Data and AI are the next stages. The Internet of Things, virtual reality, augmented reality, and 5G are potential amplifiers. This is the nature of the technology-enabled marketplace. Disruption isn't coming. It's here, and the impacts are playing out in rapid cycles.
Transformation now comes in waves. In the mid-to-late 1900s, the spaces between technological advances were more than 10-year spans. Since 2000 we have seen new software, the Web, mobile, cloud, data, IoT, analytics, AI, platforms, and more on the horizon. Each wave supports greater technical capabilities, so more parts of the business and operating models can be transferred into the technology model.
Businesses' reliance on technology will never be this low again. In this book, I will explain the climb from where enterprises are now to where they must be to survive the next 5 years. It is a techno-strategic process that blurs the lines between data scientists and strategists.
I will explain how strategy planning and implementation follow the data science life cycle. The two domains have more in common than most realize.
Strategy is prescriptive and forward-looking. It informs decision-making across the enterprise. Strategy reveals multiple paths for achieving business objectives, allowing the business to adapt to changing conditions. Optimization and efficiency are core objectives.
Reread the previous paragraph but replace “strategy” with “machine learning” or “AI.” The sentences still work.
Reliable machine learning models manage complexity and reduce uncertainty. They have predictive, prescriptive, and diagnostic capabilities, which increase a business's learning rate and accelerate continuous improvement cycles. Revenue and efficiency gains compound and make the company more productive and profitable with each cycle.
I will explain the new user-technology relationship paradigm. People view software as a tool like a hammer or a screwdriver; this is human-machine tooling. They look at models and AI as team members; this is human-machine teaming. Stable Diffusion and GPT-4 are the leading edge of the third transformation, which is human-machine collaboration. I'll explain the opportunities and how to seize them in this book.
The software paradigm handles logical processes with easily defined, stable steps. Models manage intelligent processes that require the synthesis of knowledge to novel scenarios. Today, people are handing over autonomy to models for some intelligent workflows. The number of intelligent processes that data and AI can manage and that people trust the technology to handle will multiply in the next 2 years.
This book explains how to go from a legacy business with early data maturity to competitive maintenance, advantage, and industry leadership. I will detail what needs to happen for early investments in data and AI initiatives to break even, generate revenue, and become the primary source of new growth.
The book will cover three pillars that are critical for success: frameworks, road maps, and technical implementations. This book's frameworks are lightweight and designed to generalize across businesses. They are flexible and adaptable. Road maps connect strategy to execution and value creation. Technical implementations and real-world examples will drive concepts home, making this all actionable instead of aspirational.
This book is written for data scientists, leaders, product managers, and project managers. Businesses must operate from a common playbook or transformation moves in competing directions. Business units must understand their role as part of a connected enterprise. Technology is just a tool. Companies must be rebuilt to monetize technology in partnership with people.
Playbook for the Enterprise
The book begins with the path to alignment. Continuous transformation requires a new business culture that must be rapidly implemented. Goals and incentives must align with it. As I'll explain, talent must be developed internally because external sourcing is unsustainably expensive and time-consuming.
Technical progress cannot happen without strategy informing and aligning decision-making across the enterprise. Internal adoption does not occur without operating model transformation and internal training. Value cannot be delivered to customers without business model transformation.
All three top-level strategic constructs must move together, or data and AI will fail to thrive. C-level leaders face pressure to quickly pivot from compelling transformation stories to capable execution. Investors are looking for tangible progress, a track record of successful deployments, and measurable returns.
Pressure is growing for data teams too. Budgets receive greater scrutiny, and data professionals are no longer untouchable. The risks are reinforced daily by the latest round of layoff announcements. Any group that isn't profitable now finds itself on the chopping block. Last year, Amazon, Microsoft, Alphabet, and Meta wound down unprofitable business units. The rest of the business world has followed.
The decade of bottomless budgets for innovation without expectations is over. Activist investors pose challenges to CEOs who haven't executed. They prefer companies that invest responsibly and show returns quickly. Investors expect businesses to navigate the current uncertainty by leveraging data effectively. It's not enough to keep repeating a data-driven story.
When companies like Nike are forced to confess that they don't have visibility into their inventory, stock prices plummet. Intel could not execute supply chain derisking and continues to struggle with execution. Its share price has been punished.
Companies such as Apple have used data to minimize supply chain disruptions and avoid excess inventory traps. This is the standard that other businesses are being held to. Leaders know the value that mature data capabilities deliver. Now business leaders are looking for people in the data organization to step up.
New roles are emerging. The emphasis on data and AI products has pushed a new type of product manager into the hiring plan. Data strategists are increasingly common, and C-level data leaders are seen as critical. Demand for people who can monetize the technology is growing faster than for technical practitioners.
This book is a training manual for these emerging roles. I explain how they work together and ensure the execution that CEOs need. Now more than ever, businesses must align and work together.
Systems, Models, and Frameworks
This book teaches systems, frameworks, and models. What is a system? Systems are anything that we interact with. A marketplace is a vast, complex system with multiple players and objectives. The business itself can be viewed as a system. Supply chains are systems.
This book's frameworks are designed to help readers navigate these systems. With systems, models, and frameworks, I will provide you with domain knowledge and expertise I've developed over the last 11 years of working in data science and more than 25 years in technology.
I had to figure out how to give readers a full view of what's going on across the business and make it work for several different types of companies. It took me a long time to develop a different way of explaining these constructs that goes beyond memorizing facts and definitions.
There's no step-by-step sequence of events that holds across businesses. Intuitively we all know that rigid frameworks and monolithic approaches are flawed. The most talented people can synthesize their frameworks and experience to novel scenarios. How?
Every business has domain knowledge. In this book, I'm passing on the domain knowledge that my company (V Squared) and I have accumulated. Strategy is, in a way, a representation of the business's domain knowledge.
When I tried to boil domain knowledge down to its most basic form, I came up with this: the set of biases, assumptions, and heuristics that the business uses to explain its marketplace view, processes to create value, and decision-making process.
Developing and continuously improving the business's domain knowledge is exceptionally important. There's so much value in that institutional knowledge. In the modern competitive landscape, it's more critical than ever.
The business is continuously transforming. We can introduce data into the business and influence domain knowledge. By leveraging data, we're trying to determine what's right and what can be improved. The purpose of data is to introduce new knowledge into the business.
I had to figure out a way to deliver this information knowing that you will be working with different types of domain knowledge and expertise. That's why I developed systems, models, and frameworks.
I'll explain systems and give you models and frameworks to adapt to those systems. Everything is designed to adapt and improve.
Introducing Data to the Enterprise
Unfortunately, strategy planners experimented in production in the past, which is obviously a worst practice. Without data and models, what else were they supposed to do? Business strategy resulted from best guesses, because it's all we had.
Now we have data, so I must explain what happens when we introduce data into the strategy planning process. Continuous transformation is the constant theme, and the technology model sits underneath it to help the business manage technology's value creation.
These technical strategies inform decision-making, and each framework will focus on decisions. The purpose of the strategies is to help people without all the information they need to understand the enormously complex systems they are making decisions about. We are giving them models and frameworks so that they can adapt to make decisions and be more successful.
The frameworks in this book are the same ones you'll teach to everyone else. The technology model supports decisions about what should transition from the operating model into the technology model. Technology-supported process or workflow automation initiatives can be planned to implement that transition.
Next comes the core-rim model, which I'll explain in depth. Technology lives at the core, and people handle the irreducible complexity at the rim. One layer deeper sits the value stream, which connects value creation to workflows and explains how individual tasks create value. Using that, the business can decide what should be transferred into the technology model. These changes enable continuous transformation to take advantage of each technology wave.
How do we measure it? I explain AI strategy key performance indicators (KPIs) at the end of this book. One of the KPIs you can use at the lowest level of granularity is the percentage of workflows generating data, leveraging data, and leveraging models.
These frameworks are like layers of an onion. As we get deeper into each framework, we get closer and closer to implementation—the KPIs track execution from the value perspective. We can't drag senior leaders into managing the technology or workflows. Frameworks allow them to continue to manage value creation.
Frameworks go from very high-level abstract constructs all the way down to managing the value creation of each one of these initiatives. Everything is a system, and we need a model to understand the system and a framework to interact with that system. Those help the rest of the business do their parts too.
That's where alignment comes from. Transformation becomes more efficient when the business works together from a single playbook. When the goal is to adapt and improve from day 1, business culture changes to support the data and AI-driven firm. The business takes back control, so value creation drives transformation.
Without these frameworks, businesses risk losing their autonomy to technology. When technology drives strategy, the results are catastrophic. These frameworks allow people to maintain autonomy while reaping the benefits of data and AI. Everything this big needs a justification. In Chapter 2, “There Is No Finish Line,” I'll explain what's changed to force this level of transformation.
CHAPTER 1Overview of the Frameworks
Profiting from data, analytics, and more advanced models is complex without using frameworks. Costs rise too quickly for more than the simplest of initiatives to be feasible. Most businesses are just coming to terms with the big picture of what it takes to monetize data and AI, but I've worked with it for more than 11 years.
I developed the frameworks covered in this chapter to optimize the process of technical strategy and monetizing data and AI. There are layers to the challenges, and there are layers to the frameworks. Introducing them up front in Chapter 1 will help you to keep track of them and answer some of your questions before you ask them.
This chapter serves as a quick reference for each framework. As you get further along in the book, revisiting earlier frameworks will be useful. For example, continuous transformation will take on an expansive meaning the further you read. The technology model and opportunity discovery will grow increasingly critical over the course of the book. The need for business culture changes gains urgency once I explain the big picture and maturity models.
As you work to monetize data, analytics, and more advanced models, you'll encounter a range of challenges. That's when you should come back to the frameworks in this chapter. Where in the process is the challenge emerging? The answer to that question will point you to the appropriate framework. How can you implement a framework to overcome the challenge? That will point you to the next steps to overcome the challenge.
Continuous Transformation
The continuous transformation framework explains what has changed and what businesses are adapting to. Technology is advancing faster than ever, and each technology implemented could deliver a competitive advantage. The business can't build data and AI products with software engineers. It must develop new internal capabilities to build with new technology. New products will be possible, and existing products will be refreshed with new functionality.
Each technology wave could transform the business in small or large ways. How SHOULD the business transform? That must be a strategic decision because technology threatens to take over if senior leaders don't have frameworks that help them retain autonomy.
Each wave will impose a cost on the business, so each wave must also deliver significant returns. When transformation happened slowly, companies could recoup their costs over several years. Waves are currently 3 to 5 years apart, but that span is reducing rapidly. Think of how quickly generative AI has moved forward, and you'll get a sense of the new speed limit.
Continuous transformation is part of each framework because it drives the modern competitive business. This is the root cause of emerging opportunities and threats.
Three Sources of Business Debt
Transformation should be easy, but it isn't. Adopting new technologies should be straightforward, but it's the opposite. Three sources of business debt are to blame, and businesses need frameworks to manage each.
Technical debt is accrued by making technical decisions without the continuous transformation framework. Each technological wave is inevitable. It's coming, and while we can't see all the impacts with 20/20 vision, there are obvious implications that every business should be preparing for. Building today with a high-level plan for how subsequent technology waves will impact what we're building lowers the cost and time to transform with the next wave.
Technical decisions today must amplify the value of future waves, and those subsequent waves must amplify the value of current products. Technical debt creates interference waves, while continuous transformation creates amplification waves and cycles.
Strategic debt is accrued by building a monolithic strategy that ignores technology's increasing role in the business and products. Leadership can turn to technology and the technical teams for help overcoming challenges and achieving strategic goals. Most companies don't integrate technology into strategy planning or opportunity discovery. As a result, a critical source of growth and competitive advantages sits on the sidelines instead of being elevated to partner status.
Strategy must continuously transform with each technological wave, and C-level leaders need frameworks to make technology decisions strategically. Enterprise-wide alignment on transformation and technical maturity is possible. These frameworks create the connection between technology and value creation.
Cultural debt is accrued by the business's evaluation, accountability, and incentivization structures failing to transform to support new technology waves. Technology must be integrated into the business's workflows, but most cultural artifacts are not designed to support integration. Some artifacts punish people throughout the company for leveraging new technologies. Others make people accountable for workflow outcomes that result from workflows that technology now controls.
Cultural artifacts must continuously transform to reinforce technology adoption and monetization. New frameworks support data-driven and increasingly technology-dependent business and operating models. Decisions must be made with transparency into the cultural artifacts they disrupt.
Evolutionary Decision Culture
Data and the business's technical capabilities continuously transform, which improves the quality of data available to support critical decisions. The business gathers new data and develops higher-quality models. Both introduce new information to the business. As a result, strategy planning must transform.
Strategy must be built with the best available data to avoid analysis paralysis. It must also be reevaluated when data and models improve. No decision should be considered final because there may be good reasons to change in the future.
People must be incentivized throughout the business to make decisions with their best available data and improve them as data improves. In some cases, that will lead to the business walking away from significant investments and firmly held beliefs. Transformation often requires quitting to improve.
The Disruptor's Mindset
Disruptors are critical for transformation. They challenge firmly held beliefs and what the business is working toward today. This mindset supports a culture of experimentation and rapid improvement. The big innovations come from the disruptor's ability to see flawed assumptions and the implications of replacing those assumptions with new knowledge.
Cultural artifacts must be built to support disruptors. People must be incentivized to find and replace flawed assumptions. Experimental frameworks must be implemented to enable disruptors to prove or refute their ideas rapidly. Value-centric frameworks must be implemented to prevent disruptors from derailing the business with experiments that cannot be monetized.
The Innovation Mix
Businesses cannot be innovating all the time, and a lack of innovation puts the business's future at risk. C-level leaders need a framework to decide how much innovation the company should be working on. Too much innovation and the business will fail to exploit its current opportunities. Too little innovation and margins will drop until the business's current opportunities run dry.
An innovation mix helps the business to continuously transform the resources it dedicates to innovation initiatives. The mix swings like a pendulum to meet the business needs of today and the next few years.
Meet the Business Where It Is
Technical strategy is often aspirational. Initiatives are built for an ideal business environment. Every business is at a different maturity level. Each has strengths and challenges. Not every company will choose to become AI-first or build massive internal capabilities.
We must meet the business where it is and plan technical strategy with a connection to the business's strategy and goals. We must assess the business's current maturity levels and make an opportunity-centric decision about how far it will mature to seize those opportunities.
The Technology Model
C-level leaders need a framework to make strategic decisions about how the business uses technology to create and deliver value to customers. Technology makes many things possible, but that's not how decisions should be made. Technical feasibility doesn't always lead to monetization.
The technology model framework introduces a top-level strategic construct that sits with the business and operating models. It makes technology a core pillar of business strategy. C-level leaders can make decisions based on business value without getting dragged into technical or implementation details.
They choose which parts of the business and operating models to transfer into the technology model based on the opportunities they see. These align technical initiatives with business value and strategic goals.
Individual technical strategies inform decision-making about the value-generating advantages each technology wave delivers better than alternatives. Digital, cloud, data, AI, IoT, and other technologies have unique strengths supporting different monetization strategies. Technical strategy is a thesis of value creation that explains why the business uses an individual technology to create and deliver value to customers.
The Core-Rim Model
Data and AI are novel asset classes, and a monetization strategy must be a core driver for technical initiative selection. There is a second aspect to the strategic decision process. Data and models can take over an increasing number of internal and customer workflows. But can they handle the complexity?
The logical steps performed in each workflow are only a fraction of the process. Dig deeper into workflows, and sometimes we find hidden complexity. Intelligent processes don't follow easily defined steps, and decisions are made based on expert knowledge. Are data and models reliable enough to manage the complexity?
C-level leaders need a framework to assess what workflows should be transferred to the technology model and which ones the business doesn't trust technology to manage. Core-rim provides support for those decisions.
Transparency and Opacity
Businesses have a hidden giant of data-generating processes. Where do the processes of data gathering and model development begin? The mistake most companies make is to try to gather everything. There's simply too much, and most data does not deliver much value to the business.
Transparency and opacity is a framework that aligns data gathering with high-value workflows. These workflows can be internal user or customer workflows. Data gathering must have a framework for prioritization; otherwise, costs spiral out of control, and value creation is not guaranteed.
This framework supports opportunity discovery and estimation, which lay the foundation for prioritization.
The Maturity Models
Continuous transformation must reach down into strategy implementation and execution. Without that connection, strategy is aspirational. The maturity models support implementing continuous transformation and technical strategy across four key pillars. This is the most challenging connection to make because it spans products, business units, customer segments, and technology waves. The big picture of continuous transformation is massive, and the maturity models create alignment with a lightweight framework set.
The business's data and AI capabilities continuously transform to deliver higher-value products. The
capabilities maturity model
keeps that development aligned with products and initiatives. Capabilities should be built in line with business needs so talent and infrastructure maximize their returns to the business. This framework supports value versus technology-focused decisions about capabilities development.
Products continuously transform to seize opportunities with each technological wave. The
data product maturity model
explains how digital products evolve into data, analytics, and advanced model-supported products. This framework supports decision-making about initiative planning. Products can be built today with space for the inevitable transformations coming next. They can be monetized at each phase, so incremental delivery creates incremental returns for the business. Transformation is sustainable, and costs are minimized.
Data gathering capabilities transform to seize opportunities as product lines grow into platforms. The
data maturity model
supports moving the business from opaque to transparent while delivering value at each phase. It's the connection between data engineering and monetization. Initiatives support bringing new information and domain knowledge into the business for high-value workflows.
Data and model development capabilities transform to deliver increasingly reliable models. Use cases have levels of reliability requirements depending upon how much of the workflow is turned over to technology. The way people work with technology, especially data and advanced models, must be taken into account. Breaking initiatives down to support internal users' and customers' reliability needs is supported by the
human-machine maturity model
.
The Four Platforms
The business must create a vision and product strategy to prevent product lines from delivering inconsistent experiences. Disconnected products transform in different directions and create expensive redundancies. Internal users and customers adopt these products at lower rates because they lack a coherent alignment.
The maturity models enable initiatives to be designed in alignment with business value, capabilities, and how people engage with technology. The platforms connect products as they progress through the maturity models.
The
product platform
handles customer workflows.
The
operations platform
handles the tasks performed as part of internal user workflows.
The
decision support platform
provides data to improve the decision chains associated with internal user workflows.
The
operating system model and AI app store platform
support partnering with large third-party models to deliver value to customers and internal users.
Each platform needs a vision to articulate where it is today, what's being built this year, and what the platform will be in three to five years. The vision explains how each transformation supports technical strategy and top-level business strategic goals.
Top-Down and Bottom-Up Opportunity Discovery
The business needs frameworks to support nontechnical leaders and domain experts in discovering high-value opportunities. Most companies depend upon the data team to put forward opportunities, but they aren't the closest to strategic goals, internal business needs, and customer needs. Opportunity discovery is more successful when technology teams partner with the business.
Top-down opportunity discovery provides a framework for C-level leaders to surface large opportunities that align with strategic goals. It's an evaluation process that helps leaders separate hype from reality. Opportunity discovery is managed based on technical viability, proven value estimates, and technical feasibility.
Bottom-up opportunity discovery provides a framework for mid-level leaders and frontline workers to surface opportunities based on the needs they are closest to. It creates a prioritization framework that prevents data teams from being overwhelmed with low ROI work and ad hoc requests. That justifies delivering data literacy training and self-service tools to manage those requests. The highest-value opportunities are estimated and added to the product road map for rapid delivery.
Large Model Monetization
OpenAI's GPT presents new monetization opportunities, and businesses need a framework to support decision-making. The operating system AI and AI app store frameworks explain the new paradigm in nontechnical terms so C-level leaders can surface opportunities for near-term value creation. Familiar paradigms provide context for decision-making and opportunity discovery.
Maturity model leapfrogging explains the new competitive paradigm. Businesses can leverage operating system AI to deliver high-maturity products without developing those capabilities internally. They work for only some use cases, and leapfrogging helps C-level leaders decide which use cases to move forward with.
Data curation and operating system AI retraining explain mid- and long-term opportunities created by this paradigm. C-level leaders leverage this framework to direct data capabilities maturity and data engineering initiatives that align with high-value opportunities.
The Business Assessment Framework
Meeting the business where it is requires an initial assessment. Nothing works until the business moves forward with a common understanding. The business assessment framework defines the current data and AI maturity level on seven assessment points.
The information gathered as part of the assessment feeds into the data and AI strategy document. The information helps develop the product, infrastructure, internal training, and hiring road maps. The first rounds of opportunity discovery are completed as part of the initial assessment process.
The assessment marks the boundary between planning and implementation. It's where the work begins. One of the critical artifacts is the data monetization catalog. It defines data as a novel asset class by connecting the business's data to the use cases it supports. Companies can put a value next to each data set and see where it derives the greatest value from.
The data monetization catalog enforces the need for value and use cases to drive data curation instead of gathering everything and finding value later. This avoids data swamps and high-cost, low-value data engineering initiatives.
The Data and AI Strategy Document
This framework outlines the key components of the data and AI strategy document. The purpose of the data and AI strategy is to inform decision-making across the enterprise about the technology. It should create alignment and continuity.
The data and AI strategy document must explain why the business uses data and AI to create and deliver value to customers. It must provide actionable information about how external teams will be impacted and their motivation to transform. It introduces the data team and describes its place in the business.
The document creates certainty about what comes next and why C-level leaders have chosen this direction.
Data Organizational Development Framework
The data organization must build in alignment with the maturity models. Meeting the business where it is often means there are existing resources in place today. Those serve as the starting point, but scattered resources and disparate teams don't support rapid maturity.
The data organization goes through a three-phase transformation. The initial assessment identifies where resources are today and what they are accountable for. The transformation strategy and capabilities maturity model defines where the team must go to deliver. In this first phase, the data team does not own everything it needs to deliver data, analytics, and model-supported products.
The second phase centralizes talent and infrastructure. Functional and project deliver improvement cycles optimize product quality and delivery speed.
The third phase distributes resources to the business units and product teams they support. The closer they are to the workflows and needs, the higher-value products become.
The data organizational development framework provides support for the three phases and rapid capabilities maturity.
More to Come
Throughout this book, I will define smaller supporting pieces and frameworks. I support each framework with implementation cases and examples pulled from the real world. The goal is to deliver a common playbook for the enterprise to follow.
Each framework supports and amplifies the others. They are lightweight and adaptable. Everything is meant to transform because, as I'll explain in the next chapter, there is no finish line for transformation.
CHAPTER 2There Is No Finish Line
This chapter and framework is step 1 of a much longer journey. I don't expect you to follow my lead blindly. Instead, I will explain the drivers and help you understand why taking these steps is crucial. The modern competitive landscape is fundamentally different than ever before and at its core is technology.
The business world understands technology has had an impact on the modern competitive landscape, but few understand the extent. Technology is a critical part of the business. The modern firm relies on it to create and deliver value to customers. Technology is not a static entity, and companies cannot be static, either.
Google provides an excellent example of what happens when even the most innovative company stands still. Google developed advanced data science and machine learning capabilities. However, it isn't enough to innovate and be technically competent. Monetization remains elusive if the business doesn't transform along with technical capabilities. That's how Google got caught flat-footed.
As technology evolves, the business must adapt to that new technology. In this chapter, I start by explaining what that means. Continuous transformation is a framework to manage the evolutionary nature of technology. It enables businesses to control technology's value creation. There's a massive difference between having data, analytics, and advanced models and actually monetizing those technologies.
By the end of this chapter, you'll understand what has changed and have the first tool in the technical strategy toolkit: continuous transformation. Each chapter adds a new capability and framework to help you navigate the modern competitive landscape.
Data and AI strategy paints a grand vision for data science in the enterprise, but the journey there is incremental. Transformation has no finish line because conditions continuously change as technology advances. The business's capabilities must grow to integrate new technologies and adapt. However, monetization requires the rest of the company to transform as well.
Where Do We Begin? With Reality
Today the business could be deep into digital transformation or cloud adoption, but that's not the end. Data, analytics, and AI come next. Behind them are other technology waves. However, the pervasive mentality is that transformation will come to an end. Technology will stabilize, and progress will return to where it was in the 1980s.
That's not going to happen, and businesses must adapt or be swept aside by rivals that have. Not everyone in the business understands the magnitude of the risks and opportunities. Transformation would happen faster if they did.
It's not just the technology that must change, so we can't ignore where the business and its people are today. They won't magically catch up after the data team puts the technology in place. The process is more efficient when the business transforms on an aligned road map. When technology leads the way, monetization is elusive, and adoption rarely materializes.
So much of what data scientists do fails because we refuse to meet the business where it is and let it dictate the pace of transformation. The current maturity level is rarely perfect, but this is the reality we face. The first guiding principle I will introduce about data and AI strategy is this one: “We must meet the business where it is, not where we want it to be or think it should be. We must transform at the business's speed, not the technology's.”
Transformation is more successful when it meets people where they are now and accepts how they think. This book would work in a company like Meta or Google, but most companies aren't technology-first and have no aspirations of becoming a Meta rival. Most companies' challenges start at an earlier maturity phase, which is fine.
Few initiatives require advanced machine learning and deep learning methods. The majority of businesses will not develop high-end models in-house. They will partner and purchase. The value most businesses will generate and benefit from comes from straightforward approaches, simple descriptive models, and advanced analytics.
The work required to transform the business to that maturity level is often underestimated. The focus and investment are skewed toward the technology and data team. We're in a time of “adapt or die” for companies across industries. The urgency is high, but to succeed in continuous transformation, we must meet businesses where they are. We must acknowledge that transformation is two-sided.
We must also support the business through imperfect stages of transformation. I would love to live in a world where people read this book and suddenly became data-driven mental powerhouses. I have been working with businesses in various phases of their data journeys for more than a decade, and believe me, we don't live in that world.
These frameworks apply to real people at actual companies. I wrote this book for humans, whereas most books in the data field seem designed for the ideal business with perfect people. Transformation begins humbly and imperfectly. It never stops because if it does, the company will fail.
It's the same with people. People never stop learning, growing, and improving their capabilities in technical fields. There's always a new programming language and model architecture to learn. Continuous transformation is a framework for helping the business come to terms with a reality that's been with people in the technology field since the beginning.
Defining a Transformation Vision and Strategy
The purpose of these frameworks is to create a company that's better tomorrow than today. Technical strategists must provide a future state and big picture, the main components of the vision. At the same time, there's no finish line to this journey. The vision will help the business see where they're going, but getting to that point isn't the end.
We need to be forward-looking and prescriptive to support senior leaders. Just like everything else that has been disrupted by continuous transformation, the vision will evolve and improve. Anytime I get better data or a more accurate model, my vision for the future may change or become more certain. The central theme of this book is continuous transformation.
Continuous transformation supports the most granular initiative level and enterprise level so each phase can deliver incremental returns. The business can structure transformation to align with its business goals. C-level leadership should decide how fast and far it wants to go with each technology. To achieve that objective, they need a framework to manage value creation versus technology.
Continuous transformation helps the business make timing decisions about adopting data, analytics, and AI. Longer term, there are more technology waves on the horizon. When will a company be ready to move to platforms and, potentially, even evaluate quantum computing?
These technological waves are inevitable. No matter what the business's transformation rate, each phase is still unavoidable. Companies will leverage each technology wave, so decisions should support technologies that are further out on the transformation timeline.
In the early 2000s, we made a huge mistake, and I was part of it. We built applications for the old paradigm, scaling by incrementally adding on-premises hardware and infrastructure. We should have built incrementally, knowing we were going to the cloud eventually.
When the inevitable cloud migration happened, it was harder to get buy-in because it was so expensive and time-consuming. We were forced to rebuild core components. It would have been faster and more cost-efficient to have taken a long-term view. Even in a technology organization, we fought the inevitable transformation.
Taking a continuous transformation view would have been the better choice. The entire business would have been on board with cloud adoption because the changes could have been deployed incrementally. Most of the pushback we received was rooted in the disruptions that rapid adoption and years of poor decision-making had led to.
The cost per quarter would have been more manageable, and the impacts on the product release schedule could have been minimized. There wouldn't have been a scramble to hire, select infrastructure, and retrain technical staff.
We know the business will adopt data and AI. As we're building digital apps and adopting the cloud, we should gather data differently, with AI in mind, as part of creating digital solutions. Businesses have not taken that approach, and the lack of intentional, incremental transformation makes each future wave more expensive.
Paying Off the Business's Digital Debt
Data teams inherit a massive amount of technical debt that comes from early digital transformation initiatives. The business has been working to become a digital enterprise for the last 10 years or more. This transformation focused on moving from a manual to a digital business.
Digital transformation began for some companies in the late 1990s. The goal was to transform paper records into digital equivalents. The paper record's layout became the front end and the user interface. The information people put on those paper records was captured by an application for storage in a database. The application's back end was the logic that took data where it needed to go and delivered it when users needed it again.
