The Value of Your River Guides
When and How to Price Design Partners in the era of Agentic AI
Design partners aren’t “early customers.” They’re a temporary extension of your product team. They are co-builders who validate what works, expose edge cases, and (if you get it right) become your first evangelists. The role of a design partner in the era of agentic AI, especially within vertical AI workflows, is more important than it’s ever been. The key to discovering the right wedge and creating continuous feedback loops hinges on being deeply embedded into the right workflows. Design partners serve as your river guides in this exploration.
But then how do you charge them? Should you charge them? Pre-agentic AI, the typical advice was “never give away your product for free!”. But I’m not so sure that holds up anymore, at least not right now when agentic AI is still so novel and competition is fierce.
I’ve had a few conversations with early stage founders around this recently so this is my attempt to break down how I see the role of a design partner in early stage development and where/how to charge.
Tl;dr I see 3 distinct phases along the design partner journey:
Product Discovery: learn and validate which agents and workflows matter
Product Validation: introduce nominal pricing to increase skin in the game and prove some amount of willingness to pay
Conversion & Pricing Optimization: convert to customers and optimize for ROI
In agentic AI, pricing is now about measurable outcomes. Your design partners, first and foremost, help you discover those outcomes. Then, they can help you price them.
Phase I: Product Discovery (what to build & where it fits)
In the first phase with a design partner, the priority is discovery. Pricing takes a back seat. Your goal is to uncover a repeatable pain point and the agentic workflows that matter most. Giving away your product, which at this stage are actually your services, is perfectly acceptable and in some ways a strong positive if structured well.
Part of that structure requires a defined scope, which I often see as a time-boxed engagement (8–12 weeks, sometimes up to 6–12 months) to avoid “pilot purgatory.” If your idea is well-formed, you can set written success criteria, but early on this is likely too hard to pin down.
Ideal design partners mirror the workflows and data of the market you plan to serve. It’s fine if your ICP is still sharpening (narrower is better), the right partners will help you pinpoint which sub-sectors and company profiles feel the pain most acutely. I generally think it’s stronger if you can get design partners from cold outbound. This shows there is something in your pitch that is resonating.
This stage is hands-on from both sides. You give white-glove support, working alongside the customer’s team on a regular cadence, in exchange for feedback and data. Therefore, it’s important to have a sponsor who can unblock access and keep you top of mind.
If a design partner isn’t paying in dollars, they should be paying in time and social capital. That means real, recurring involvement: weekly check-ins, pulling in live workflows and data, unfiltered feedback, and thoughtful introductions to other potential partners. This level of engagement forces ownership, surfaces whether the problem truly matters to them, and accelerates your learning. If they won’t commit that time, they’re not the right partner to help shape the product.
The output you’re trying to achieve in Phase I is to have identified a wedge workflow, baseline metrics, and hopefully at least a couple early wins (e.g., measurable time savings, improved accuracy) that you can then start to point to with other prospective customers.
Once you have this, the role of a design partner evolves. Rather than identifying what workflows matter, you’re now looking to them to help you understand what commitments could look like.
Phase II: Product Validation (introduce nominal pricing → “skin in the game”)
You now know the workflow and data; the next phase is to prove willingness to pay. Pricing comes to the foreground. This phase is about replacing some (not all) of the time the design partners spent with you in Phase I with money. It’s not about getting to the perfect price.
The ideal partners span your ICP by size, sub-vertical, and data shape. Usually this ranges between 3-7 partners. Keep bespoke asks limited, although some is healthy as training for broader adoption. Just as in Phase I, having a sponsor is critical. In this stage, that sponsor ideally has some discretion over budget.
Because you have baseline metrics, and hopefully at least a couple early wins from Phase I, you have a clearer sense of ROI math (baseline cost, time saved, defects avoided, revenue lift, etc…). You are still getting clarity around interventions, and human overrides.
In this phase, the goal is to prove that your wedge case (identified in Phase I) is in fact a true pain point that your customers are willing to put money towards. How much money can still be an open question. Therefore, price for commitment, not revenue maximization.
Ideally the structure of how you price can mimic the long-term structure you think you’ll have. For example, you could charge per workflow automated or case completed; or you could use a hybrid platform such as minimum plus usage tied to business outcomes (throughput, time-to-resolution, error rate). If you’re still unclear what shape the pricing structure will eventually take, then you can simply ask for some nominal platform fee to signal seriousness.
Avoid endless free pilots, fuzzy list prices, value metrics you can’t meter, and bespoke builds that won’t generalize. You graduate when partners show repeat usage and outcome deltas and agree in principle to convert on the published list price.
Phase III: Conversion & Pricing Optimization (turn partners into customers)
In this phase, the objective is finally to optimize price-market fit. In other words, how can you capture the most value without slowing close cycles. You now have clear metrics that prove the value of your product. You have logos and case studies from your design partners that you can reference.
You are now signing-up your early customers, focusing on true ICP accounts.
Keep pricing simple to learn fast. By simple, I mean things like offering only two tiers at most with monthly or annual terms. Map price to how customers win. For example, a platform fee plus workflow/agent modules (the platform fee can double as implementation); or workflow tiers by volume, controls, and governance; or a hybrid of platform minimums with outcome-aligned usage.
Contrary perhaps to popular belief, I think it’s best to avoid multi-year commitments this early. The unknowns are still high, and long contracts mask dissatisfaction and raise churn risk at renewal. Not to mention, they may leave money on the table as the ROI grows more compelling over time.
Pricing is a living test of cost versus value. You’re in the sweet spot when usage is habitual, outcomes improve, NRR expands by workflow or site, margins clear model/inference costs, and churn drivers are understood and managed.
Defining Value
The value design partners bring is much greater than revenue. They are your river guides, helping you scout currents and eddies, calling the clean line you can run again and again. In the beginning, you’re buying learning. It’s an investment into your product and technology. If you can create undeniable value for 10 customers even at a significant loss, you will be in a position to scale it profitably for 1,000. That is the real win.