How Do You Run a Design Partner Program?

How do you run a design partner program?

Pick three to five companies who already feel the problem, agree in writing what each side gives and gets, meet weekly for a fixed period, and decide up front how it ends. A design partner program is a search for a repeatable product, not a discount scheme with meetings attached.

We see the same failure repeatedly: fifteen enthusiastic logos, no written agreement, no cadence, and a roadmap that now belongs to whoever complains loudest.

Here is the structure we would run, and the decisions that determine whether it produces a product or a bespoke build.

What is a design partner, and what are they not?

A design partner is an early customer who trades access and feedback for influence and a better price. They are not a pilot customer, not a beta tester, and definitely not a consulting client. The difference is that you are building the general version, and they know it.

Steve Blank's framing is the right one to hold onto. In January 2010 he defined a startup as an organisation formed to search for a repeatable and scalable business model, noting that most of the time the customers do not behave as you predicted.

A design partner program is one of the cheapest ways to run that search with real stakes attached. Feedback from someone who has committed something is worth ten conversations with someone being polite.

How many partners should you have?

Three to five, and never more than you can call every week. Fewer than three and you build for one company's edge cases. More than five and you cannot hold the pattern in your head, which is the entire point of the exercise.

Diversity matters more than count. Two partners in the same segment with the same workflow will agree with each other and mislead you both.

If you have twenty interested companies, that is a good problem. Take five as partners and put the rest on a waiting list with a monthly update, because they are your first launch audience.

Who makes a good design partner?

Someone with the problem now, a budget line eventually, and a named person whose own targets improve if this works. That third condition is the one people skip, and it is the one that predicts whether you get the weekly hour.

Prior experience with the category helps too. 6sense's 2025 B2B Buyer Experience Report, drawing on nearly 4,000 responses across North America, APAC and EMEA, found buyers evaluate about 5 vendors and buy from their day one shortlist 95 percent of the time, nearly always with prior experience of the vendor.

Read that as a warning about enthusiasm without authority. A champion who has never bought anything in this category will struggle to get you paid. Our notes on defining your ICP cover how to describe the right company before you go looking.

Should they pay you?

Yes, something, and never for your time. Charging for the product itself, even at a heavy discount, is the cheapest test of whether the problem is real enough to carry a budget line. Charging for services turns the relationship into consulting, and the incentives change the same week.

Paul Graham's July 2013 essay puts the risk plainly: consulting is the canonical example of work that does not scale, and it is safe to do it only so long as you are not being paid to. Once somebody is paying for attentiveness, they expect a bespoke solution rather than gratitude.

A workable structure is a reduced annual price fixed for the program period, with the standard price agreed in advance for renewal. That keeps the conversation about the product. Our guide to pricing and packaging tiers covers how to set the number you are discounting from.

What goes in the agreement?

One page, six things. What they get, what they give, how long it runs, who the named people are, what happens to feedback, and what happens at the end. Put it in writing even with a friendly customer, because the friendliness is exactly why nobody remembers what was agreed.

The give side matters most: a weekly hour with a named person, access to real data or a real workflow, and permission to be named as a partner if it goes well.

Be explicit that feedback shapes the roadmap but does not own it. That single sentence prevents most of the difficult conversations in month three.

What does the weekly rhythm look like?

Thirty minutes of watching, thirty minutes of talking, and a written summary the same day. Watch them use the product on their real work rather than asking what they think, because what people report and what they do are different.

Keep a shared channel open between sessions for the small things. Those messages are usually more valuable than the scheduled call, and they are where you find the workarounds people invented rather than complaining.

Write the summary yourself and send it. It forces you to decide what you actually learned, and it gives the partner evidence that the hour they spent produced something.

How do you avoid building one company's product?

Only build what at least two partners ask for independently, unless you have a specific strategic reason to do otherwise. One request is an anecdote. Two separate requests are a signal worth investigating. Three across different segments is a feature you should probably ship.

Keep a visible list of what each partner asked for and how often. The pattern becomes obvious within six weeks, and it protects you from the persuasive partner who asks for everything.

When you do say no, say it clearly and say why. Partners handle a reasoned no far better than a vague maybe that quietly never ships, and the honesty is what makes their feedback keep flowing.

How long should the program run?

Eight to twelve weeks, stated up front. An open-ended program has no forcing function, so it drifts until the relationship becomes normal support. A fixed end date gives you a decision point and gives them an exit that is not a breakup.

Inside that window, aim for two or three product iterations that partners can actually feel. If nothing visibly changed from their input, the program was a research project wearing a partnership label.

Build the ending into the first conversation. Everyone behaves better in a relationship with a known finish line.

How does the program end?

With a conversion conversation and a case study request, in that order. Ask them to move to standard pricing at the agreed number, then ask whether you can publish what happened. Doing it in the other order makes the case study feel like the price of a discount.

Those stories matter more than most early teams expect. TrustRadius, surveying 1,862 buyers and 444 vendors for its 2026 report, found 74 percent of buyers use reviews to inform their decisions, and ranked demos, free trials, prior experience and user reviews as the most influential resources.

A partner who renews and lets you tell the story has given you two of those. Our piece on running a customer story program covers how to collect them without burning goodwill.

What would we do first?

Write the one-page agreement before you approach a single company. Drafting it forces you to decide what you are actually offering and what you need in return, and it makes that first conversation concrete rather than aspirational, which is what gets a serious company to say yes.

Then approach eight companies to land five, expect two to be quietly useless, and treat the three that engage as the real program. That ratio is normal and planning for it stops the early panic.

If you want help turning what you learn from partners into a site and story that sells to everyone else, we are happy to talk. You can see how we work at phoenix.studio.