India Has Talent. Where Are Its Customers?
Why the quality of demand matters as much as the supply of talent in building Indian technology companies
A technology company learns from its customers long before it becomes a technology leader.
The customer tells it which problem matters, what failure costs, how much reliability is enough, and whether a better solution is worth paying for. A demanding customer asks for a feature the company has never built. A patient customer gives it time to get that feature right. A customer willing to take a chance gives it something even more valuable: its first order.
This is an uncomfortable way to think about India’s technology ambitions. We spend an enormous amount of time asking whether we have enough engineers, enough founders, enough research, and enough funding. We should ask an equally important question: who is buying what these people are capable of building?
In my earlier piece on industrial policy, I argued that India has become better at paying for activity than paying for capability. But capability cannot be created by conditions in a subsidy scheme alone. It also needs a customer willing to use a product, find its limitations, pay for improvements, and place the next order.
You cannot build a sophisticated supplier base without a market that gives suppliers a reason to become sophisticated.
The customer becomes part of the company
Imagine two firms building industrial sensors. Both have good engineers and access to similar components.
The first sells to a customer who wants the cheapest device that passes a basic specification. The customer replaces it when it fails and treats maintenance as someone else’s problem. The supplier learns how to cut costs and win tenders. Those are useful skills, but they will not necessarily make it better at reliability, design or systems integration.
The second sells to a factory that measures downtime, shares failure data and is willing to pay for a product that survives difficult operating conditions. It demands an improvement, tests a prototype and buys more if the result works. Over several years, the supplier learns not just to make a sensor, but to understand the system around it.
The difference is not the intelligence of the engineers. It is the work the market asks them to do.
This is not just a thought experiment. Research on UK firms found that firms with past exporting experience were more likely to report learning from buyers, and firms reporting such learning were more likely to see subsequent productivity growth. The authors also note that the wider evidence on learning from exports is mixed. The result is suggestive, not proof that selling abroad automatically makes a firm better, or that every exporter deserves a subsidy.
The mechanism matters more than the geography. A buyer can transfer knowledge through specifications, testing, feedback and repeated orders. A foreign customer may do this. So can an Indian customer. And a demanding customer can also simply squeeze a supplier without sharing enough value for the supplier to invest.
Why the American market matters
One reason the United States produces so many successful technology companies is that it has an extraordinary market for new technology. It has large enterprises, government agencies, universities and households with the resources to try new products. It has customers who can place orders large enough to turn a promising idea into a viable company.
This is not the whole explanation. Universities, immigration, research funding, deep capital markets, military spending and a large integrated economy all matter. Some of these advantages also help create the customers themselves. A wealthy enterprise with access to capital can afford to experiment in ways a cash-constrained firm cannot.
That last point is important. It is easy to describe Indian customers as excessively cautious. Sometimes they are. But refusing to take a chance on an unproven supplier may also be rational when a failed installation can destroy a small business, financing is expensive, and replacement equipment is difficult to obtain.
What looks like a cultural aversion to risk can be a financial constraint wearing a cultural costume.
This is why telling Indian businesses to be more innovative is not enough. We have to make experimentation less likely to ruin the person paying for it.
India does have demanding customers
The argument is not that Indian consumers cannot support innovation. They already do, and they often impose constraints that are difficult to understand from outside India.
Two-wheelers are a useful example. Indian firms operate in a large, competitive market where price, fuel economy, maintenance, durability and increasingly design matter enormously. Consumers do not buy a good engineering story; they buy a vehicle that has to survive daily use at a price they can afford. That pressure has helped create firms capable of competing outside India too.
A demanding customer does not have to be rich. But the customer must be able to buy, and the supplier must have a reasonable chance of earning enough to improve. Competition without that possibility becomes a race to cut costs rather than a reason to invest in capability.
This also means we should be careful with easy explanations for successful Indian consumer companies. Ather and Atomberg operate in markets where customers exist and where product differences can matter. Their success does not prove that demand alone explains their performance. Founders, execution, capital, regulation and competitors matter too. The point is narrower: their engineers had a market in which the effort to build a better product could be rewarded.
For an industrial startup, a semiconductor design firm or a supplier of specialised equipment, that market may be much harder to find at home. The buyer may be abroad. The few domestic buyers may be large incumbents with established vendor relationships. The potential buyer may need a proven installation before it will take a risk, while the supplier needs its first customer to produce that proof.
That is the first-order problem.
The first customer problem
Government is an obvious candidate to help break this cycle. It buys transport, healthcare equipment, defence systems, energy infrastructure and software at a scale that could support serious domestic suppliers. But being a large buyer is not the same as being a good first customer.
A procurement system built around prior turnover, prior experience and the lowest upfront price will naturally favour firms that already have contracts. The new supplier is asked to prove that someone else has taken the risk first. In sectors where the state is itself the largest potential customer, this becomes circular.
India has recognised parts of the problem. The government’s startup procurement guidance describes exemptions from prior experience, turnover and other requirements in eligible cases, while retaining quality and technical standards. Some states have created additional procurement routes. So it would be wrong to say nothing has changed.
The question is whether the route works in practice. Can a new firm find an appropriate tender? Can it demonstrate technical capability without an unrelated turnover threshold? Can a department run a small trial, evaluate it honestly and buy more if it succeeds? Does the supplier get paid on time? An exemption on paper answers only the first part of a much longer journey.
A good public buyer should be able to place small, competitive first orders without pretending that every prototype is ready for nationwide deployment. The trial should have a budget, a defined problem, independent testing and a route to a larger order if the product passes. Failure should be documented rather than automatically treated as proof of corruption or incompetence. Payment terms should be credible enough that a small supplier can survive the experiment.
This is not an argument for giving contracts to anyone who calls themselves a startup. Public money needs scrutiny. A pilot that never publishes its results, a procurement process written around one favoured supplier, or a permanent domestic preference without performance requirements would create the opposite of capability.
The discipline has to apply to the buyer as well as the seller.
The capital constraint behind the customer
There is another reason this question cannot be separated from finance.
A customer who pays late transfers the cost of working capital to the supplier. A large company may survive that. A small one may have to borrow, delay salaries or refuse the next order. Even a successful sale can weaken it if the cash arrives too late.
The buyer has its own constraints. An expensive loan, thin margins and uncertainty about future demand make it harder to buy an untested machine or invest in a new supplier relationship. The supplier and the customer can both be acting rationally while the system produces too little experimentation.
That is why industrial policy needs to look beyond incentives for the producer. It should examine access to working capital, contract enforcement, timely payment and the ability of buyers to finance useful upgrades. None of these automatically creates great companies. They create the conditions in which more companies and customers can try.
What should we measure?
If the objective is capability, the number of startups registered or the value of subsidies announced is not enough. We should ask more uncomfortable questions:
- How many new suppliers receive a first meaningful order from government or a large enterprise?
- How long does payment take after delivery, and how often are disputes resolved on time?
- Do successful trials lead to larger orders, including from buyers other than the original sponsor?
- Are suppliers improving reliability, design and technical performance, or merely assembling more units?
- Can a firm win business without an existing personal relationship?
None of these measures is perfect. A first order can be political theatre. A larger order can create dependence on a single buyer. A local supplier can become complacent if protected indefinitely. But together they get closer to the question that matters: is the market helping firms learn, compete and grow?
India needs more excellent engineers and better education. It also needs more customers who can make use of that excellence. The two reinforce each other: capable people make better products possible, and demanding markets give people a reason and an opportunity to become more capable.
We keep asking why Indian companies do not spend more on research. Perhaps the better place to start is with the first customer who will pay for the result.