Building a new product is a chicken-and-egg problem: You need users to make the product better, but no one wants to use an early, unpolished solution. At Hex, we solved this challenge through a method we call commitment engineering.
Commitment engineering is the process of systematically building feedback loops with early users through small, reasonable commitments.
I've used this approach to build multiple tools over my career, and after sharing it with other founders, I've seen how effectively it helps teams validate their ideas and find their first customers.
Here's how commitment engineering works, and why it's essential for early-stage product development.Commitment engineering principles and processThe most effective way to start commitment engineering isn't with a product pitch — it's with questions about the problem you want to solve. Talk to potential users who have an intimate knowledge of the problem, because they experience it regularly.
Ask them about their pain points and the issues that get in their way. Then follow up with questions about the solutions they’ve tried before, and where those products or services have fallen short. The goal in these initial conversations is to understand your potential users’ challenges deeply, before presenting any solution.
A dialogue like this can kick off your first commitment loop. After a productive problem discovery conversation, you might say: "If we built a first version that addressed these challenges, would you spend 45 minutes trying it out and giving us feedback?" This creates your first test of genuine interest. If they decline, you've learned something valuable — either you're misunderstanding their problem, or the issue is not painful enough to warrant their time.
From there, each successful interaction can lead to bigger requests. You can ask the potential user to:
Schedule a follow-up feedback session with you.
Bring a colleague to the next demo.
Use the product for a specific task.
Present it in a team meeting.
Introduce you to the decision-maker who can purchase (this would be the last step in the process).
Each request is an exchange of value. You improve the product based on their feedback, and they get closer to a solution that fits their needs. The key is to keep your requests reasonable and proportional to the value you're providing.
You're not asking for a full deployment or significant workflow changes — you're requesting small investments of time and social capital that indicate real interest.Why commitment engineering worksTrying to sell an early product puts both founders and potential customers in an awkward position. The product isn't ready for real deployment, and customers feel pressured to make a yes/no decision before they understand the value. Commitment engineering removes this pressure by creating a collaborative environment instead of a sales dynamic.
This approach helps validate your product ideas through real evidence rather than polite interest. When someone says your product looks promising, it's easy to mistake that for validation. But actions speak louder than words. If they won't commit 30 minutes to a follow-up conversation, they probably won’t be willing to spend money on your product later.
The relationships you build through commitment engineering can be invaluable. Each interaction teaches you something new about your users' needs, workflow and constraints. These insights help you build a better product and find product-market fit (PMF). At Hex, our earliest users became our strongest advocates because they saw their feedback shape the product.
Most importantly, commitment engineering prevents many common startup pitfalls. Teams often spend months building features that seemed promising in sales conversations but never get used.
Small commitments test interest before you invest significant resources. When someone is willing to bring their boss to a demo or run a team meeting using your product, you know you're on the right track and you’re addressing important pain points.
Building feedback loops and finding PMF with commitment engineering
Have any ideas or suggestions to improve this article?
Building feedback loops and finding PMF with commitment engineering