Great Lakes & MMA hosted the Second Dr. Bala V. Balachandran Memorial Lecture with a talk on Fusion Strategy: Data & AI for Strategic Advantage by Prof. Vijay Govindarajan (Coxe Prof., Tuck School of Business, Dartmouth CollegeTuck/Dartmouth).
Pro f. Vijay Govindrajan
Co xeDistinguishedProfessorof Management
at Tuck School of Business, Dartmouth College
It’s a great honor to deliver this memorial lecture for Bala, who was my best friend, mentor and well-wisher. Let me start by laying out a fundamental framework I have for innovation—the three box solution. Whenever I work with organizations, I tell them to think about everything they do and put them in three boxes.
Three-Box Solution: A Framework for Innovation
Box one is about managing the present. It’s about improving the performance of your organization the way it is constructed today. This is about optimizing your system as it exists. If you’re General Motors, you have a job to do in box one—improving the efficiency of internal combustion-driven gasoline-powered automobiles because that’s your core business.
Box two is about selectively forgetting the past. And box three is about creating the future. What I find when I work with organizations is they over-focus on box one and then think they’re doing strategy. Is box one important? Absolutely critical. But strategy must include box two and box three.
If you want to create your future in the year 2035, you have a job to do in box two—you have to selectively forget. Let me clarify: strategy for any organization is about becoming a leader in 2035. But strategy is not about what you have to do in 2035. It’s very much about the projects you’re executing in 2026 across the three boxes so that you intersect with the year 2035. How are you allocating resources today? How is the organization’s energy focused today across the three boxes so that you stay relevant in the next decade?
The reason this is a challenge is that the thinking process, the people, the capabilities, the metrics you need to excel in box one are fundamentally different than what it takes to excel in box two and box three. Yet in 2026, you’ve got to do both. Another way we can define box one is that it’s about competition for the present—efficiency. Box two and box three are about competition for the future—innovation.
From IT Services to Gen AI
If you want an example of a box three idea with which we in India created enormous value, then that would be in the mid-1990s when somebody came up with global service delivery. Indian companies could do 75% of software-related work in India where talent is much cheaper but equally capable, and only 25% near the global client’s location. That idea created the whole IT services boom—Infosys, Wipro, Tata Consulting Services. We’ve seen enormous value unleashed in the last 30 years, but that’s going to be nothing compared to the value we can create with the next inflection point, which is Gen AI.
When I started my academic career in the early 1980s at Harvard Business School, the faculty was asked to research how information technology would change business and management. Here we are, February 2026, asking exactly the same question. Of course, the technology infrastructure we had in the early 1980s was primitive—IBM PC with a green screen, Lotus 123. We didn’t know about internet, mobile, or Gen AI. We have lot more digital tools today, but the question remains: how would digital technologies change the world of business and management in the next decades?
Three Waves of Digital Transformation
In the last 45 years, digital technologies have changed business in three phases. Transformation 1.0 improved efficiency but didn’t change strategy—SAP, automating supply chain, automating accounting. This gave rise to the big boom in India for IT services, all about efficiency.
Transformation 2.0 is where digital giants used digital technologies to introduce new strategies. This is box three. Google didn’t make advertising more efficient—it changed the way we do advertising. Netflix didn’t make watching movies more efficient—it introduced a new business model. Digital giants essentially destroyed the B2C consumer sector.
Transformation 3.0 is just starting. Digital technologies are going to create box three strategies in the B2B asset-heavy industrial sector. The fundamental difference is that in transformation 2.0 we dealt with pure information goods like Google, or physical products which disappeared like cameras. In transformation 3.0, the physical product will not disappear. We’re talking about tractors, buildings, power plants. A tractor will never disappear, but the value is going to migrate from the physical product to data and AI.
Transformation 3.0: The India Opportunity
I believe transformation 3.0 is the India opportunity. In the 1980s and 1990s, information technology had a humongous impact on organizations but practically no impact on individuals. The way we lived, worked, watched movies, transacted was still analog. You came to work, transacted with paper currency, went to movie theaters. Fast forward to February 2026—digital technologies are having a humongous impact on individuals. The way we live, work, transact, watch movies is in the digital world.
This happened because three laws of computing are converging: compute, connect, and cloud. Anything that can be computed will be computed. Anything that can be digitized will be digitized. I see many of you wearing eyeglasses. In the next 10 years, could eyeglasses become a digital product with sensors so your eyesight automatically adjusts as it deteriorates?
Think about the value. With analog eyeglasses today, I go to an ophthalmologist once a year, get new lenses fitted. All that cost and time disappears if you digitize the product. Not only that, if eyeglasses become digital, my eyesight is automatically adjusted in real time. I challenge every organization—if you’re making a physical product, ask yourself what it will take to digitize it. Perhaps the technology isn’t there yet, but the power of technology keeps improving and the price keeps dropping.
When compute, connect, and cloud converge, you have big data. What we’ve done so far is automate processes. When you automate processes, you improve efficiency but don’t change strategy. But today we have the opportunity to digitize the products themselves. When you digitize the product, the process of value creation changes, and strategy is nothing more than the process of value creation.
How Digital Giants Redefined Strategy
This inflection point happened when Steve Jobs introduced the iPhone. I looked at the top 10 market-capitalized companies in January 2007, before iPhones. There was only one tech company—Microsoft. Everybody else was a physical product company. Prior to 2007, when I taught strategy, I focused on competitive advantage based on products—the classic Mike Porter. Either you make the product cheaper or better.
I looked at the top 10 market-capitalized companies on February 2, 2026. These are by and large digital giants. The 2007 Microsoft is very different than 2026 Microsoft. These digital giants are not the IT giants of the 1980s and 1990s—not Cisco, IBM, Dell, Infosys, TCS, Wipro. Those companies were passively delivering products. These digital giants are actively shaping strategies and have created enormous value in the last 18 years.
How did these digital giants fundamentally change the definition of strategy? For us to understand that, we need to go back to how we build competitive advantage. I say there are only two ways: scale and scope. What these digital giants did was fundamentally change the very definition of scale and scope.
Product-as-Consumed and Smart Data
Let me give you a history lesson. In the 20th century, we listened to music through Sony Music, Columbia Music, Virgin Music. I used to teach a case on Sony Music 25 years ago, comparing their competitive advantage because they all thought of themselves as product companies. How many albums do we produce and sell? Today we consume music through Spotify, Apple Music, Amazon Music.
The most important lesson: Sony Music is a B2C brand, but their processes were B2B. All that Sony Music kept track of was number of albums and CDs produced and sold. Whereas Spotify started to track the music you are listening to in real time. When you track what an individual is listening to in real time, the process of value creation significantly changes.
To capture this phenomenon, I’ve come up with a concept called data graphs. Netflix has a movie graph. LinkedIn has a professional graph. Amazon has a purchase graph. Facebook has a social graph. Google has a search graph. Spotify has a music graph. Google was the first company to introduce this concept.
This word has two components: data and graph. Data doesn’t mean big data or any data. It means smart data. Companies don’t suffer from lack of data—we have the wrong kind of data. We’re drowning in data lakes because we’re not collecting smart data. Google collects smart data, which is product-as-consumed data. Sony Music tracked product-as-sold, whereas Spotify tracks product-as-consumed.
Data Graphs and Network Effects
Let me contrast Google with a physical library. If you walk into a library, look at 10 history books and withdraw one, the librarian only knows the book you withdrew. They don’t know you looked at 10 history books. Go back a month later, look at 15 travel books and withdraw one—the librarian only knows you withdrew that book. Certainly they have no clue you came earlier looking at history books. This is what Google does. Every search on Google is product-as-consumed. This is not big data. This is smart data.
The second word is graph—a relationship between variables. We learned this in high school algebra. Google has 550 million variables. Jaguar as an animal is a variable, Jaguar as a car is a separate variable. When you have 550 million variables, you can’t use paper and pencil. You need neural networks, deep learning, machine learning.
We all knew about direct network effects before 2007—when a new member joins, it adds value to existing members. A telephone network is a classic example. But digital giants introduced data network effects. This has nothing to do with new members. It’s about how existing members engage.
Netflix has 200 million subscribers. Suppose no new subscriber joins—that means no direct network effects. But they still benefit from data network effects. As long as the 200 current subscribers keep watching Netflix movies, every time you watch a movie, you’re making the scale, scope, and speed of the data graph grow exponentially.
Three important principles digital giants used in transformation 2.0: First, they tracked product-as-consumed and constructed signature data graphs. Second, they leveraged data network effects—Netflix is learning from 200 million subscribers to give solutions to a single individual. Third, they use sophisticated AI analysis: descriptive (what happened?), diagnostic (why did it happen?), predictive (what will happen?), and most importantly, prescriptive (what should happen?).
Most industrial companies stop with descriptive analysis. Digital giants did four types of analysis in an integrated way, giving you recommendations in real time based on what they’ve learned across customers.
Why Generative AI Changes the Game
These digital giants have also given us generative AI. Why Gen AI is a big deal: before Gen AI, we had AI models that could only analyze structured, quantitative data. Gen AI can analyze qualitative data—images, sounds. That’s important in the B2B industrial sector where you have the sound of machines, images, videos. Gen AI is going to put data graphs on steroids.
This takes me to transformation 3.0, which is just starting. This is the India opportunity. Today the world’s GDP is $100 trillion. Of that, only 25% is in the B2C consumer sector. That’s all we’ve digitized so far—$25 trillion. When we digitized that, think about the value we unleashed. If you add up the market cap of the magnificent seven tech companies, it’s bigger than the GDP of the US. That all got created in the last 18 years on the back of iPhone and digitizing one sector. Think about the value we can create if we digitize the remaining $75 trillion in the B2B industrial sector.
Fusion Strategy in Practice: John Deere
Let me give you an example—John Deere. In the 20th century, John Deere was a powerhouse making big tractors. But today they’ve become a fusion strategy company. Imagine a farmer in the US has 2,000 acres. When you plant, there will also be weeds. How do you get rid of weeds? In the 20th century, you could recruit hundreds of thousands of workers to pluck weeds row by row—expensive and error-prone. Or blanket spray with a helicopter across 2,000 acres with pesticides. But when you blanket spray, you can inadvertently kill plants.
John Deere recently introduced See and Spray. It moves at 25 miles per hour, taking millions of photographs. It’s equipped with sensors, IoT, and computer vision. As it takes photographs, it processes them in real time using machine learning, deep learning, neural networks. When it spots a weed, it kills just the weed and keeps moving. Pesticide use drops by 90%. You only kill weeds without killing plants. And you’re environmentally friendly.
This game is just starting. The game for consumer AI is already won by Silicon Valley—OpenAI, Anthropic, Google. One large language model can handle all consumer-related data. In the industrial sector, we’ll have multiple large language models. A model that analyzes data for automobiles has to be different than one for buildings.
The other big difference: in the consumer sector, AI hallucinates and gives you 80% accurate recommendations. That’s okay in B2C. If Amazon recommends 10 books and you like eight but not two, you’re irritated but can live with it. But in B2B asset-heavy sectors, British Airways cannot get 80% accurate recommendations from Rolls-Royce. When you’re flying at 50,000 feet, 80% accuracy is fatal. Large language models for industries have to be 100% accurate, and they haven’t yet been developed. That is an India opportunity.
Building Industrial AI Leadership for India
We in India have the software experts. We’ve invested in digital infrastructure. We talk about Make in India. Let’s combine the three and create large language models for industry. We can’t create for every industry—let us choose maybe 10 industries where India has a chance to become a global leader.
What would be one sector? Two-wheelers. Why? Because India has maximum market share. Between Hero, Bajaj, TVS and others, we produce 35% of the two-wheelers in the world today. We have bulk, market share leadership. Let us make our two-wheelers a digital industrial product. Equip them with sensors, computer vision, IoT and collect product-as-consumed data. Then leverage data network effects and create a large language model for mobility.
When you create a large language model for mobility, you can apply it beyond two-wheelers to automobiles, trucks—not just in India but globally. Just like we led the IT services boom, we can lead the world in industrial AI. This is a mega opportunity knocking on the doors of India.
Ancient Wisdom, Modern Strategy
This whole notion of three box solution is not my original idea. This is a philosophy written in Hindu scriptures 5,000 years ago. Hinduism tried to explain why life has survived for centuries by the role of three lords. Lord Vishnu is the god of preservation—that’s box one, manage the present. Lord Shiva is the god of destruction—that’s box two, destroy the past. Lord Brahma is the god of creation—that’s box three, create the future.
None of these gods are more important than the other two. Only if these three gods do their job in a balanced way will humanity survive. Hinduism draws life as a circle. There’s no beginning, no end. Every life form born will be preserved, ultimately destroyed, and everything destroyed will be recreated. I’ve simply taken something written 5,000 years ago and repackaged it. If your corporation has to survive forever, it’s the same principle—preservation, destruction, recreation as a rhythmic cycle.
Applying the Three-Box Solution Personally
This three box solution is not only to transform India, not only to transform corporations—you can use it to transform your own personal life. Everyone in this room must create a three box plan for yourself. What is your true north? What capabilities you have today, what new capabilities you need to build?
As individuals, the reason we don’t spend enough time in box three every day is, because if you don’t spend time in box three today, it doesn’t hurt you today. It only hurts you in the future. Think about exercise—if you do exercise every day, you’re assured of future health. But you wake up and say you don’t have time. If you don’t exercise today, your health doesn’t decline today. It only declines in 2030. That’s why you postpone that investment.
By the way, if you don’t exercise today, your health actually declines today. The decline is so small you don’t notice. Future comes in daily doses. Future never arrives all of a sudden. If India wants to be a leader in industrial AI by 2050, it’s not about what India has to do in 2050. It’s about what are you doing today to create that future. Future is a function of actions you take in the present. The last 60 minutes, all of you put your box one on hold. My talk is going to end, but don’t go back to box one 100%. This is not mission accomplished. This is not journey completed. Thank you very much.
