When people talk about companies that use data well, the usual examples are Google, Amazon, Netflix or Meta. These businesses are often treated as natural data companies because their products are digital and their customer interactions generate information continuously.
Rarely does a paint company make the list. Asian Paints should.
Behind one of India's best-known consumer brands sits a business that has spent decades using data to make better decisions about demand, inventory, manufacturing and distribution. Its advantage did not come from adopting a fashionable technology overnight. It came from gradually building the systems, relationships and operating discipline that allowed information to improve decisions across the business.
That is what makes the Asian Paints story interesting.
It is not really a story about AI. It is a story about building a business that was ready for AI long before AI became a boardroom priority.
A paint business has a surprisingly difficult data problem
Paint looks simple from the outside. You choose a colour, buy a can and put it on a wall. The business behind that transaction, however, is much more complicated than the final product suggests.
A paint manufacturer has to determine which products people will want, where they will want them and when demand is likely to increase or decline. Those patterns can vary significantly between cities, neighbourhoods and seasons. A colour that sells quickly in one market may move slowly in another, while demand can rise sharply during periods associated with home renovation, festivals or new construction.
The inventory problem is equally difficult. A retailer cannot afford to keep unlimited stock, particularly when thousands of products and shades may be available. At the same time, the manufacturer cannot produce everything in advance without tying up capital in inventory that may not sell. If a customer wants a product and it is unavailable, the sale may simply go to a competitor.
So the question becomes:
How do you get the right product to the right place before the customer needs it?
That is where Asian Paints built an advantage. Rather than relying only on orders placed after demand had already appeared, the company worked towards understanding demand early enough to anticipate it.
The data was hiding in the distribution network
Asian Paints built a large and direct distribution network while developing an enormous amount of information about purchasing patterns. The company could observe what products were selling, in what quantities, through which retailers and at different points in the year.
Over time, those transactions became more than sales records. They became signals that could reveal how demand was changing across the market.
The data contained information about geography, seasonality, product movement and inventory requirements. It could show where a particular product was gaining traction, where sales were slowing and which dealers might need replenishment before they placed an order.
Eventually, those signals could be used to predict what was likely to happen next.
The HBS case study on Asian Paints describes how the company used predictive AI and machine learning tools to forecast demand for specific products in specific locations. That is a very different use of data from simply reporting what happened last month. Instead of treating data as a record of the past, the company used it to influence future production, inventory and distribution decisions.
From reporting to prediction
There are roughly three levels of data maturity.
The first is descriptive. It answers the question:
What happened?
Sales were down 4%. A particular product sold 12,000 units. A region missed its target. These figures are useful because they provide visibility into performance, but they do not necessarily explain the causes or indicate what the business should do next.
The second level is diagnostic. It asks:
Why did it happen?
Perhaps demand dropped in one geography because a product was unavailable. Maybe a competitor changed its pricing, or a seasonal pattern was different from expectations. Diagnostic analysis gives managers a better understanding of the factors behind a result.
Asian Paints' advantage comes from pushing further.
The third level is predictive. It asks:
What is likely to happen next?
Which product will sell? Where will it sell? How much will be required? When will the inventory be needed?
Once a company can answer those questions with reasonable accuracy, data begins to influence the physical movement of goods. Forecasting is no longer an exercise performed after the fact. It becomes part of the operating system, shaping what gets produced, where it is sent and when it arrives.
That is where the economics become interesting.
The supply chain becomes smarter
Imagine two companies selling similar products. Company A waits for a retailer to place an order before responding. Company B can anticipate the order because it has a much better understanding of local demand.
Company B has a structural advantage. It can plan production more intelligently, move inventory earlier and reduce the likelihood of stockouts. It may also use logistics more efficiently and reduce the amount of working capital tied up in unnecessary inventory.
The HBS case study reports that Asian Paints was supplying more than 70,000 registered dealers multiple times a day, with pre-emptively supplied inventory selling through rapidly. The exact numbers are less important than the underlying mechanism.
Predict demand → position inventory → improve availability → capture demand.
The better the prediction becomes, the more effectively the company can coordinate manufacturing, distribution and sales. Over time, that coordination can improve both customer service and operating efficiency.
The flywheel is the real advantage
This is where the Asian Paints story becomes particularly interesting. Every sale creates data, but the value of that data depends on whether the company can use it to improve the next decision.
More data can improve forecasting. Better forecasting can improve inventory decisions. Better inventory decisions can improve product availability, and better availability can generate more sales. Those additional sales then create more data, allowing the cycle to continue.
That is a flywheel.
A competitor can buy the same software, but it cannot instantly recreate decades of transaction history, dealer relationships, distribution infrastructure and operational knowledge. It would need to build the surrounding system as well, including the processes that capture reliable information and the teams capable of acting on it.
That is why data advantages can become difficult to copy. The advantage is not the database by itself. It is the system around it.
Asian Paints has continued expanding its technology layer
This is not simply an old case study about demand forecasting. Asian Paints' more recent reporting shows that its use of technology has expanded across the business.
The company describes analytics in supply chain operations, manufacturing, material planning, formulation optimisation, quality assurance and automated demand fulfilment. It has also reported the use of video analytics and GenAI applications.
Its 2025-26 annual report highlights AI and technology as part of its services strategy, alongside continued investment in digital capabilities. This suggests that data is not being treated as a specialised tool for one forecasting team. It is becoming part of how the company manages a wider range of operational and customer-facing activities.
The direction is clear. Data is increasingly connected to the decisions that determine how the business produces, moves and sells its products.
This is where many companies get AI wrong
There is a common temptation in the current AI cycle. A company buys a new tool, creates an AI team, runs a few pilots and builds a chatbot. Then, after the initial excitement fades, it wonders why the business impact is difficult to measure.
The Asian Paints example points to a different approach.
The technology works because it sits on top of a deeply connected operating system. There is data generated through real transactions, processes that turn that data into usable information, people who make decisions based on the information and physical operations capable of acting on those decisions. There is also a feedback loop that allows the company to learn from the results.
Without those pieces, AI has very little to work with. A model may produce a prediction, but the prediction will not create much value unless the organisation can trust it, act on it and measure what happened afterwards.
The biggest lesson isn't about AI
It is tempting to look at Asian Paints and say:
"Great example of AI in the paint industry."
That misses the bigger point.
The company was building its data advantage long before generative AI became mainstream. Its progress came from years of investment in distribution, information systems, dealer relationships and operational processes.
The lesson is about patience. Build the distribution network. Capture the information. Understand the customer. Connect the data to real decisions, measure the outcomes and keep improving the system.
Then, when better technology arrives, you already have something useful to apply it to.
That is much harder than simply adopting a new tool, but it is also much harder for competitors to copy. Technology can accelerate an advantage, but it rarely creates a durable one without the underlying business infrastructure.
The paint is only what the customer sees
When a customer walks into a store and picks a shade, they see a product. Behind that decision is a much larger system involving demand forecasting, inventory planning, manufacturing, logistics, retail relationships and customer behaviour.
Data and technology connect those activities, while thousands of operational decisions determine whether the right product is available at the right time.
That is why Asian Paints is such an interesting business case. The company sells paint, but its competitive advantage has increasingly come from understanding the market around that paint and responding to demand with unusual precision.
Perhaps that is the more useful way to think about AI in business.
The question isn't:
"Where can we use AI?"
The better question is:
"What decisions would become dramatically better if we could predict what happens next?"
Once that question is clear, the next step is to build the data, processes and systems that make the prediction possible.
Asian Paints offers a very good example of what happens when a company does exactly that.