Drucker Wrote the Sentence. Dawar Wrote the Instructions.

Cannes, February 2000. The GSM World Congress, and a Finnish company with a small booth among very large ones.

I had just moved to Stockholm to open NetHawk's Swedish subsidiary, selling protocol test equipment for networks that did not yet exist. A man stopped at the booth: senior, grey, expensive suit, cigarette, in a genuinely bad mood. He was out of options, not out to be sold to.

He was developing 3G base stations and needed a simulator that behaved like the network controller his base station would eventually talk to. He had been to the large vendors. They all had a 3G roadmap. None of them would tailor anything, because in 2000 the standard was not really a standard yet. Every vendor had implemented it with its own variations. A simulator built to specification was not a simulator built to his base station. Close is useless when you are hunting a bug at three in the morning.

The large vendors were not wrong to say no. A product manager's job is to build for the market, not for one account, and forking a codebase for a single customer breaks the roadmap. The reasoning is sound. It just happens to be built entirely around the seller's own cost.

I had no roadmap to protect, so the reasoning did not apply to me. I built him a version matched to his exact base station, not the general standard. It won us an initial order worth half a million euros; we delivered on the agreed lead time, and once the rollout started, more orders followed.

For thirteen years I told that story as a story about being small, the underdog outworking slow, large competitors. It is a good story. It is also the wrong one, and you cannot deliberately repeat the wrong story.

The correction came in 2013, on a flight, when I read a Harvard Business Review article by Niraj Dawar called "When Marketing Is Strategy." His argument: the center of competitive advantage has moved downstream, into the space where the customer lives with the purchase. The criterion of advantage is not your costs. It is the customer's costs and risks.

I recognized Cannes immediately. The man with the cigarette was not shopping for a simulator. His program was blocked, and every week without a test environment was a week his engineers could not close faults against a launch date already promised to an operator. His cost was schedule. His risk was arriving late to a market that would only be new once. The large vendors were competing on product. I had, without knowing it, competed on his risk.

Here Dawar goes further than Drucker. Drucker gave us the sentence everyone can quote: the purpose of a business is to create a customer. Magnificent, and it tells you nothing about Monday morning. Dawar turned it into instructions.

First, measure the advantage in the customer's costs and risks, not your own. Every internal objection to tailoring is about seller cost. Every reason a customer signs is about buyer risk. The money sits in the second conversation.

Second, whoever defines the purchase criteria defines the competitive set. The large vendors had defined the criterion as conformance to the standard, and on that criterion they were unbeatable. The moment the criterion became conformance to his implementation, their scale stopped being an asset. I did not beat them. I changed the question.

Third, ask whether the advantage accumulates. A special edition that ends with the project is a favor with an invoice attached. Mine bought a position inside a development program, and the follow-on orders came to us because we knew that base station better than anyone outside the customer's own building. The advantage compounded.

I have also seen this go badly, and product managers who resist tailoring are often right. Fork one account too many and you are left with duplicated support and no product to sell. What made Cannes work was a condition, not a principle: the variant problem was industry-wide and temporary, and the standard would converge. Building for one customer in 2000 was building for the market in 2002. When it is genuinely one customer's private problem, the orthodoxy wins and should be respected.

Here is an exercise worth running. Do not look at a deal you lost. Look at a deal you won that you did not expect to win. Write down the reason you gave at the time. Then find out what it actually cost that customer to have the problem, and what they risked by not solving it.

If the two answers differ, you are already running a strategy. You simply have not named it, which means you cannot teach it, budget for it, or repeat it on purpose.

Kauko Väinämö, CEO, Ginolis.