Early customers as a knowledge source

14 design partners drove 70% of our 2025 roadmap. Not the marketing team, not the CEO's vision, not the competitive matrix. 14 people we call monthly for 45 minutes and ask. Here is how we structured it, what pattern we see, and which requests we IGNORE.

The structure

With every design partner: one 45-minute call per month. The structure is fixed:

• 0–5 min: What changed on their side since we last spoke
• 5–25 min: What they used heavily this month, what became important
• 25–40 min: What does not work, what they miss, what causes frustration
• 40–45 min: New idea — one concrete suggestion, however wild

The whole call is transcribed (by our own Nortinia Engine assistant, naturally), and the key takeaways land in a shared backlog the partner can see — so they know their ask is registered.

The shared backlog

This was the best product decision we made inside the partner program. We did not open a "public roadmap". Each design partner runs as a separate engagement in the Nortinia Consulting App, where a shared backlog view shows their requests, their status (queued / in-flight / shipped / declined), and which feature shipped because of their ask.

When a feature goes out, the release notes name the partner (with their permission). It is not a marketing trick. Emotionally it matters to partners: they know their talking to us was not wasted.

The 70% breakdown

47 major features shipped in 2025. Of those:

• 33 (~70%) came from at least 2 design partners explicitly asking
• 9 (~19%) were internal team pain points
• 5 (~11%) were leadership calls based on competitive/market factors

Of the 33, 12 were requested by a single partner alone, and we only built them once a third independently asked.

The pattern: "every 5th customer says the same thing → it ships"

This is our most useful heuristic. If 1 in 5 design partners says something, interesting. If 3 say it, we start paying attention. If 3-4 of 5 mention it independently — say across three consecutive months — it lands on the next quarter's roadmap.

Example: the entire Content Studio publisher pipeline idea came out because 4 of 5 partners independently said within one month "we would love to make content, we just don't have the capacity". After the fourth one we started designing. Six months later, 2,847 articles ship automatically.

The pattern we IGNORE: "we'd buy if you had X"

This is the easiest trap. If a partner says "we'd buy X if you had it", IT MEANS NOTHING. Buying intent is told by past action, not by future promise.

Concretely: in 2025 we received 11 "we'd buy if you had X" requests. Only one led to building X (a specific workflow module), and the requester hesitated for 4 months on the purchase, then left.

What we learned: "we'd buy if you had X" 95% of the time really means the partner does not believe they need it either. It is just a polite way to say "not interested".

What the design partner program is useless for

• UX detail decisions. A partner cannot tell you where the "Save" button should go. That needs observability and heatmaps.
• Pricing. Partners will always say it is too expensive. Pricing decisions must come from your own P&L.
• Website pull-quotes. We do not ask partners for testimonials. Their health is being a partner, not a marketing asset.

What we are building in this now

In H2 2026 the partner program expands from 14 to 25. The hard part: we do not need more calls. The new partners are different (mid-market, not startup), and a different pattern will emerge. We are not changing the monthly call structure. We are starting the engineering work behind the shared backlog (Consulting App engagement-backlog → ticket-system sync) now.

The lesson: early customers are the most credible knowledge source you have, if you let them talk and do not tell them what they should want.