Assumption and evidence plan
A one-page statement of who the product is for, the problem it solves, what they do today instead, the assumption most likely to be wrong and the signal that would confirm or refute it.
Service · Product validation
Commission a first release whose job is to answer a question: will this group of people use, pay for or come back to this product? We write down the riskiest assumption with you, cut scope to the one journey that tests it, build that journey properly, and instrument it so the evidence is clear enough to decide what happens next.
Riskiest assumption written down before features are chosen
One complete journey built to production standard
Analytics and feedback capture in place on launch day
Deliverables
A one-page statement of who the product is for, the problem it solves, what they do today instead, the assumption most likely to be wrong and the signal that would confirm or refute it.
A map of the single path from first visit to the moment of value, with everything outside it listed as deferred, so the team can say no to good ideas without relitigating them.
A web or mobile app covering that journey with real sign-up, real data and, where the test depends on it, real payment, built on a codebase that can be extended rather than thrown away.
Named analytics events for each step of the journey, a funnel view, in-product feedback prompts and an export your team can analyse, all configured before the first user arrives.
Simple admin screens that let your team do by hand what will later be automated, such as matching, approvals or fulfilment, so the MVP tests demand without building every system first.
Early decisions
An MVP aimed at willingness to pay looks different from one testing retention or supply. We pick one primary question so scope decisions have something to be judged against.
Behind a polished front end, many first releases run on people. We decide which steps must be software for the test to be valid and which a team member can handle from a console.
A responsive web app reaches testers without store review; a native app may be essential if the idea relies on the camera, location or push notifications. We choose for the experiment, not for the long-term roadmap.
Small does not mean careless. Authentication, data protection, backups and basic accessibility are kept at production level, because an MVP still holds real people’s data.
Before launch you agree what result would make you pivot, persevere or stop. Writing this down protects the decision from being rationalised after the numbers arrive.
Engagement
A short series of working sessions produces the assumption statement, the target users you can actually reach and the evidence plan, along with a list of features that will not be in this release.
We design and test a clickable version of the core journey with a handful of target users, fixing confusion before it becomes code.
Engineering covers the journey end to end, including the manual console and analytics. You review working builds frequently and make scope calls quickly when trade-offs appear.
The release goes to a defined group of users. We monitor stability, fix defects, then sit down with you to read the funnel and feedback against the criteria agreed at the start.
Fit
You have a waiting list, a community, a sales pipeline or a partner willing to put the product in front of real people soon after launch.
There is a clear assumption that, if wrong, sinks the business, and you would rather learn that from a small release than a full build.
A founder or product owner is available to decide what stays in and what is cut, often within the same day a question comes up.
Alternatives
If you are adapting the mechanics of a successful app rather than testing whether demand exists at all, an app clone build gets to a fuller product sooner.
Open registerWhere clickable prototypes and interviews would answer the question, a design-only engagement is cheaper than writing code.
Open registerOnce customers pay and the next problem is tenancy, billing and plan management, a SaaS build is the better frame.
Open registerAn MVP is often misread as a cheap version of the full product. In practice it is a precise instrument for reducing one uncertainty. The minimum is in the number of things it tries to prove, not in how well it is built. A narrow product that works reliably gives clean evidence; a broad product held together with tape gives noise, because users leave for reasons unrelated to the idea.
Every new product rests on a stack of beliefs: that people have the problem, that they will try a new solution, that they will pay, that supply can be found, that the unit economics work. They are not equally uncertain or equally fatal. We rank them with you and design the release around the one that is both most likely to be wrong and most damaging if it is.
Sometimes the honest conclusion is that code is not the right test yet, and a landing page, a manual service or a prototype study would answer the question more cheaply. We will say so.
Many successful products began with people quietly doing work that later became software. An MVP can match supply to demand by hand, approve listings manually or send reports from a spreadsheet, as long as the user experience is genuine. We build a simple console for those tasks, which keeps engineering focused on the parts users touch and lets you learn which manual steps are worth automating.
Instrumentation is planned before the build, not added afterwards. Each step in the journey emits a named event, so you can see where people drop off, and feedback prompts appear at moments that matter. We do not set benchmarks borrowed from other products; you define what a meaningful result looks like for your market, and the data is read against that.
The release ends with a decision meeting. If the evidence supports continuing, the next step might be broadening the journey, automating the manual console or preparing for paid acquisition. If it does not, you will have spent far less than a full build to find out, and you keep the code, the data and the learning. Either way, the backlog of deferred features is waiting, already written down.
FAQ
Mainly the length of the core journey, whether real payments are involved, the number of platforms, integrations the test depends on and how much needs to be automated rather than handled manually from a console. Choosing a sharper question is the most effective way to reduce cost.
Because duration depends on the journey you choose to test, how quickly decisions are made on your side and external dependencies such as app store review or payment provider approval. We give a plan once the assumption and scope are written down, not before.
No. We keep the codebase small but structured so it can be extended if the evidence is positive. What we avoid is building infrastructure for scale or features for users who have not yet appeared.
Yes. Code, design files, analytics accounts and user data belong to you, and are set up in your accounts from the start so nothing needs to be transferred when we finish.
Acceptance covers the core journey working end to end on supported devices, the analytics events firing correctly, the admin console handling the manual steps, and no open defects that would distort the test. You sign off on staging before the pilot audience is invited.
Features outside the tested journey, scaling work, extensive admin automation, multiple languages and secondary platforms unless the test requires them. User acquisition and marketing are also yours, although we help you plan how pilot users will be recruited.
Related services
If you already built a first version with AI tools or a no-code platform and it is now unstable or stuck, rescue may come before rebuild.
Open registerIndependent testing for releases where trust matters early, such as products that take payment or handle personal data.
Open registerA support arrangement that keeps the pilot stable while you gather evidence and plan the next release.
Open registerPrimary sources
Dated official documentation, standards, and research that support the factual claims on this page.
How custom events are defined and sent, the basis of funnel measurement in a pilot.
Apple’s beta distribution service for putting iOS builds in front of pilot users before store release.
How to read this page
Scope, cost and timeline figures are planning guidance until confirmed in a written proposal. Diagrams and screens are illustrations, not client deliverables, and source-code rights and handover follow the signed agreement. See the delivery process and the full disclaimer.
Next decision
Define outcomes, constraints, evidence, rights and handover before delivery begins.
Commercial rights, repositories, environments, documentation, acceptance and handover remain contract-defined.