Design partners: The pre-launch edge most AI founders ignore

Before writing (or generating) a single line of code, the hardest question for any early-stage AI startup to answer is figuring out what they should actually build, for whom, and what customers will potentially pay for.

Founders who answer that question well don't do it through theoretical market research alone; rather, they pair that research with design partners.

Design partners are a small, deliberately selected cohort of target customers who help shape the product from the very beginning, in exchange for early access and discounted pricing.

Done right, a design partner program validates your product, tells you what to build first, confirms what customers will pay, and has the potential to be a startup’s first batch of paying advocates before you've ever run a formal sales motion.

Two companies come to mind that successfully leveraged this product development strategy: Ada and Strella. Their early-stage stories illustrate why working with design partners isn’t just efficient, but also a strategic differentiator to getting early traction.

Takeaways for early stage founders

• Start before you have a product. Design partners aren't for testing a finished MVP. They're for shaping what gets built in the first place. The earlier you engage, the less you build in the wrong direction.
• Enthusiasm is noise. Conversion is clarity. Founders mistake positive feedback for validation. Structure your program with a hard deadline and a binary ask: go paid or don't. When all 12 of Strella's cold-recruited partners converted, that was PMF evidence. A great call is not.
• Pricing is a product decision. Ada and Strella both arrived at their pricing models through design partners, not internal modeling or spreadsheets. Ada discovered enterprise buyers needed predictability over outcome-based billing. Strella learned per-project pricing created friction. Pricing should reflect how your ICP actually buys.
• Your design partners become your first sales assets. Every partner who converts becomes a reference customer, a case study, and a proof point for the next deal. The design partner program is the beginning of your sales motion, not separate from it.
• Renewals are the real proof of concept. First sales tell you people were willing to bet on your promise. Renewals tell you you delivered. Strella hit 150% NDR on its first cohort. That traces directly back to building the right product for the right people from day one.

Ada: Becoming the customer to understand the customer

Before Ada was a product, it was a question: how do businesses scale customer service without sacrificing quality? To answer it, Co-founders Mike Murchison and David Hariri didn't commission a study. They became customer service agents themselves, working at the support desks of seven different companies before writing a single line of code.

Those seven companies gave Mike and David the qualitative research they needed to figure out their startup’s product strategy.

Each company eventually became Ada’s original “design partner,” allowing the first version of Ada to run inside their operations, which would serve as the proof-of-concept that the software could replicate the value Mike and David were creating manually.

That immersive, pre-product partnership also shaped Ada's pricing evolution: the team discovered early that enterprise customers didn't want outcome-based billing tied to individual resolved tickets. They wanted predictability: annual commitments they could forecast and budget around. It was the kind of signal you can only get from real customers with real procurement processes. [Read the full Ada case study →]

Strella: Cold outreach as a signal, not a last resort

Strella took an equally rigorous approach, but with a different structural twist. After validating their core behavioral hypothesis, that people are genuinely willing to talk openly to an AI interviewer, Co-founders Lydia Hylton and Priya Krishnan recruited 12 design partners entirely through cold LinkedIn outreach to build a fully-fledged AI-powered customer research platform. No warm introductions from their network of friends and family. That was deliberate because they wanted to see whether this product idea was enticing enough to get design partners interested in their prospective company.

"It's much harder to convince a stranger on the internet than someone doing you a favor," Lydia explained. "If you can get a stranger to become a design partner, you likely have something."

Each of those 12 partners met with the team biweekly, used early product versions, and gave structured feedback throughout. In exchange, they'd receive a discount at launch. The program ended with a hard conversion deadline: go paid, or don't.

All 12 converted. And from those partnerships, Strella refined not just the product but its pricing architecture, landing on a seat-based model rather than per-project billing, specifically to reduce organizational friction and encourage usage without requiring recurring budget approvals. Strella reached $1.6M ARR in its first year of monetization and 150% NDR on its first cohort of renewals. [Read the full Strella case study →]

What this means for early-stage AI founders

Three principles worth internalizing when considering a design partner program:

1. Conversion is the signal, not feedback. The goal of a design partner program isn't to elicit enthusiastic responses on a call; it's to measure prospective customers’ willingness to pay (WTP). Structure your program with a hard deadline and a clear conversion ask. When Strella's 12 cold-outreach partners all converted, that was evidence of product-market fit. Positive sentiment is encouraging, but not a sign of monetization.
2. Let pricing emerge from the partnership, not a spreadsheet. Both Ada and Strella arrived at their pricing models not through internal deliberation but through what they learned about how their customers actually buy. Ada's shift from outcome-based to conversation-based pricing stemmed from understanding enterprise procurement realities. Strella's seat model was the result of watching how organizations approved and controlled spending. Pricing is an integral part of a product and should be viewed through the lens of your ICP.
3. The harder to recruit, the stronger the signal. Conventional wisdom for early-stage founders is to tap into their networks to build their first cohort of customers. But Strella's cold outreach demonstrated that their idea of building an AI-powered customer platform was enticing enough to get their ICP interested in learning more. A design partner who had no prior relationship with the founders and no personal obligation to say yes is a fundamentally more honest signal than one who agreed as a favor.

Download this quick read here and share it with your team.

Frequently asked questions about design partner programs

What is a design partner program?

A design partner program is a structured, pre-launch arrangement between an early-stage startup and a small cohort of target customers. Design partners co-develop the product before it's built, providing structured feedback in exchange for early access and preferential pricing. Unlike beta testers, who evaluate a finished product, design partners actively shape product direction, pricing architecture, and go-to-market strategy, and are expected to convert to paying customers at the program's conclusion.

When should an early-stage startup start a design partner program?

Before product development begins. The right moment is when you have a clear problem hypothesis but no built solution. Waiting until you have an MVP means you've already made product decisions that design partners could have informed. Ada's founders became customer service agents at seven companies before writing a single line of code. Strella's co-founders recruited their 12 design partners before their AI moderator was fully built.

How many design partners should a startup have?

Most practitioners recommend between five and 12. Enough to surface patterns across your ICP and stress-test your assumptions — not so many that the feedback becomes unmanageable or the program loses its high-touch character. Strella ran a cohort of 12; Ada worked with seven companies. The number matters less than the quality and representativeness of who you recruit.

Should design partners pay?

Eventual conversion to paid is the whole point. Design partners typically receive discounted pricing or early access in exchange for structured participation, but the program should end with a hard conversion deadline and a binary ask: go paid or don't. Strella structured its program exactly this way, and all 12 partners converted. Indefinite free access is an advisory relationship, not a validation signal.

What's the difference between a design partner and a beta user?

Beta users test a product that already exists. Design partners shape a product that doesn't fully exist yet. The relationship is higher-touch, more structured, and oriented toward co-creation rather than quality assurance. Beta programs answer "does this work?" Design partner programs answer "what should we build, for whom, and at what price?"

How do you recruit design partners?

The strongest signal comes from cold outreach, not your existing network. A stranger who agrees to become a design partner, with no prior relationship and no social obligation to say yes, is a fundamentally more honest signal than a friend doing you a favor. Strella recruited all 12 of its design partners through cold LinkedIn outreach, deliberately avoiding warm introductions. If a stranger says yes, you likely have something worth building.

If you’re an AI founder or operator, join Bessemer's Building AI Differently series to get quick reads and vault drops on the latest AI insights.