Design partners replace the free trial

The standard opening move for a software company used to be a signup form. Build something small, put it in front of everyone, watch which users stay. The companies being started now increasingly skip that step entirely and begin with a handful of contracts.

The arrangement is a design partnership: a small number of customers who pay something, commit real internal time, and get influence over what gets built. The startup trades away breadth deliberately, on the theory that ten committed users teach more than a thousand curious ones.

Part of the reason is that self-serve stopped being cheap. Acquiring users through content and search assumed a distribution surface that has narrowed, and paying for attention in a category where every competitor is funded produces cost per signup that never amortizes.

The larger reason is the shape of the problems being solved. Software that touches a regulated workflow or sits next to a system of record cannot be evaluated by an individual who signed up with a work email. Somebody has to grant it access to something that matters.

That access is the actual product decision. Founders describe the first partner as a negotiation over permissions rather than features, and the companies that win are the ones that arrive with an answer about logging and revocation before being asked.

The approach carries a failure mode that is easy to walk into. A startup with three design partners and a roadmap that serves all three has built a consultancy with a product's cap table, and the divergence usually shows up as a fourth prospect who wants none of it.

Experienced founders manage this by choosing partners for similarity rather than prestige. Three mid-sized companies in the same industry with the same workflow produce a specification. One bank, one hospital and one retailer produce three products.

Charging matters more than the amount. A partner paying a modest fee has an internal sponsor who defended the expense, and that sponsor is the difference between a pilot that gets used and one that is remembered fondly by a team that never deployed it.

The reference is the compounding asset. In enterprise categories, buyers do not evaluate a startup on its documentation. They call someone comparable, and a partner willing to take that call is worth more than the revenue that partner produces.

Investors have adapted their diligence accordingly. The metric that carries weight at this stage is not signups. It is whether the design partners renewed, expanded, and will speak to a stranger about it.

Self-serve has not disappeared, and for products an individual can adopt alone it remains the cheapest distribution ever built. For everything that requires permission from someone with a budget, the first ten customers are now recruited rather than acquired.