He Sold Consulting To Buy The Dataset He Couldn't Get

The Insight

Some products can't be built until someone hands you something first. Data. A logo. An integration. A reference customer. And the people who have it won't give it to a company that hasn't shipped anything.

Felix Hoffmann hit the hardest version of this. 7Learnings forecasts demand for every product at every price, then sets the price that hits a retailer's goal. To train that, he needed a large retailer's sales history. No large retailer gives its sales history to a company with no product.

He had the expertise. Six years as a pricing consultant at Kearney, then two years running price optimization at Zalando, Europe's largest fashion marketplace. None of it got him the data.

So he stopped trying to sell software. The first contract 7Learnings signed was a consulting project.

The decision rule: when the blocker is an input rather than a customer, sell whatever gets you the input.

How They Did It

1. They named the real constraint. Not "we need a customer." They needed volume. As Felix put it, "for this type of product you cannot work with like a small retailer from the corner of your street. That doesn't work. You need like really big data sets to develop such a product."

2. They sold the thing they could already deliver. A decade of pricing expertise, packaged as consulting. The retailer wanted to implement decision optimization in-house, and Felix knew how.

3. They negotiated the data rights as the actual deliverable. This is the part that makes it a strategy instead of an accident: "at the same time they agreed for us to use their data to develop a product on our own."

4. They got paid for it. "We got paid for it, for the consulting itself." The dataset was free. The work funding it was not.

Then they built. A forecast engine, an optimizer, and a front end, which Felix says took "like a year or so."

What Trips Up Founders

They wait for the ideal first customer. The founder who insists the first contract must be SaaS waits for a buyer who wants an unbuilt product. That buyer is rare, and the wait is unfunded.

They take consulting for the cash, not the asset. Most services-to-SaaS stories are about runway. This one isn't. The money was secondary; the dataset was the point. If your consulting work leaves nothing behind that compounds into the product, you bought time and nothing else.

They never ask for the rights. The retailer implemented its own version and Felix kept the data. Both sides got what they wanted, because he asked.

When This Doesn't Work

If your product doesn't need a scarce input, this is a detour. A founder who can build the whole thing from public data or their own usage should just build it.

It also fails when the asset isn't transferable. Data rights carry across customers; a bespoke integration for one company usually doesn't. Before signing, ask what specifically you keep, and whether it works for customer number two.

The signal you've got it wrong: six months in, the consulting is growing and the product isn't.

The Question

What do you need that nobody will give you until you already have a product?

Name it precisely. Then ask what you could sell today to a company that has it, and what you would need to keep from that engagement to be unblocked.

You'll know quickly whether anyone will pay for it. That answer is worth more than another month of building without the input.