For a long stretch the business sold one thing: reach. You bought volume, you got volume, and the conversation with a buyer was almost entirely about price per thousand. Revenue grew the way that kind of business grows — slowly, and in proportion to how hard the sales team worked.
Meanwhile the platform was recording, for every message it delivered, whether a person engaged with it. Who opened. Who clicked. Which of them did it repeatedly. Sixty million records, more than fifty behavioural attributes each, updated continuously.
We used that data for reporting. We showed it to clients in a dashboard at the end of a campaign so they could see how it went. That was the whole of it.
The question nobody had asked
A person who engaged with a message last week is not the same asset as a person who did not. Everybody in the industry knows that in the abstract. The thing that is genuinely hard to see — and the reason it stayed unexploited for years across an entire sector — is that we were the ones holding the list, and we were selling both kinds at the same price.
The product that came out of it was not complicated: define segments by engagement behaviour, let a buyer target them specifically, and price them for what they are worth rather than what they cost to deliver. The engineering was a few months. The data had been there for years.
Why an obvious idea stays invisible
Not for a technical reason. The idea lives in the gap between two functions: nobody whose job is the data is thinking about the price list, and nobody whose job is the price list knows what the data contains. An idea that requires both views at once has no natural owner, which is precisely why it survives untouched.
That gap exists in almost every company and it is structural rather than anyone's failing. Engineering optimises what it is asked to optimise. Commercial optimises within the products it has been handed. The question that unlocks a business — what would a buyer pay more for that we can already produce — sits between the two, and organisations rarely have a seat for the person who asks it. Occupying that seat deliberately is the highest-leverage thing I have ever done.
Over the eight years that followed, revenue went from ₹1.8 crore to ₹20 crore. Other things contributed. This was the largest single one, and it was the cheapest to build.
How to look for yours
The test I have used since is uncomfortable and quick. Take every field your system records that is not required for the system to function. For each one, ask: if a competitor had this and we did not, what could they sell that we could not?
Most of the answers are nothing. The valuable part is that the exercise is cheap and the payoff is asymmetric — you are looking for one item on a list of forty, and you already own all forty.
The second test is simpler still: find the dashboard nobody acts on. Data that is presented but never decided upon is usually data that has been mistaken for a report when it was a product.
Written by Manmohan Singh, who builds the systems that move the money. About.
The build log, by mail
New notes, when I publish them.
One mail per note: what broke, the theory I held before I understood it, and the rule I kept afterwards. Irregular on purpose — I write these when something happens, not to a schedule, and I would rather send you nothing than send you filler.
Also in the log
Thirty per cent of our volume stopped earning money
Yield was falling and every instinct in the room said send more. We sent less of what paid, and yield rose about eighty per cent.
The ad you cannot change after you send it
An image URL inside a delivered email is immutable, and Google caches it. Both facts are fixed. The ad still had to change on every open.