Skip to content
Sohayb Baset

Partnerships & Pilots

Test Payment Before You Finish Building: How to Design a Commercial MVP

A technical MVP asks whether the product can be used. A commercial MVP asks whether someone will pay.

English edition of an issue of my Arabic newsletter. اقرأ النسخة العربية

Also published on EgyptInnovate.

Published
Last reviewed
Reading time
4 min

Many teams do not fail because they are not working. They fail because they work for a long time before testing the hardest commercial hypothesis: will the market pay for this value?

The product can improve, the interface can look better, the features can grow and the team can close sprint after sprint. None of that, on its own, proves that a customer will pay, a partner will deliver, a transaction will complete or money will reach the account.

Technical MVP versus commercial MVP

A technical MVP answers: can the product be used? A commercial MVP answers a different question: can a small system deliver real value and collect clear payment?

At the start, that system may be made of simple tools: a spreadsheet, WhatsApp, a sign-up form, follow-up calls and a person connecting the parties by hand. That does not make the model weak. It means the test is designed to check payment before the product gets complicated.

Why the payment test gets delayed

The product is controllable. It can be planned, split into tasks and tracked. The market does not respond that easily. The customer may not understand the offer, the partner may be slow, the price may meet objections, and the moment of payment may reveal that the value was not clear enough.

So some teams take the easier road: a new feature, a better interface, a broader dashboard, more integrations. The postponed question stays the same: who will pay, for which result, and at what moment does the value fall due?

What makes an MVP commercial

Not every prototype or demo is a commercial MVP. A commercial test meets four conditions:

  1. It serves one clearly defined segment. Do not test the whole market at once.
  2. It delivers an understandable result. Do not sell "a platform"; sell a result the customer or partner can describe.
  3. It defines the payment moment. On sign-up, on confirmation, on completion, on delivery, or when a specific result is achieved?
  4. It completes the journey to collection. Interest, a request or a meeting is not enough. The test is complete only when value reaches the counterparty and payment falls due or is collected.

These conditions stop the team from confusing interest with payment, usage with value, and activity with revenue.

The 5-5-100-5 pilot

In multi-sided platforms, a small commercial pilot can look like this:

Sohayb frameworkSaudi Expansion Notes

The 5-5-100-5 pilot

  1. 015 committed partnersOne successful transaction with one enthusiastic partner does not prove the model repeats.
  2. 02100 confirmed offersOffers or service units with confirmed availability, so the test measures the market and operations, not stale data.
  3. 035 completed transactionsAt least one successful transaction with each partner. One deal can be luck; several reveal a pattern.
  4. 045 payments collectedPayments or commissions actually collected, which tests the whole chain.

These numbers are not a general standard or a benchmark for scaling. They are a test design that spreads the evidence instead of relying on one successful case. Several completed transactions show where the customer stalls, where the partner is late, where price objections appear and where the founder has to step in.

The aim is not to maximise the number. It is to find out whether the chain works: a clear offer, the right customer, a responsive partner, delivered value and collected payment.

Do not use the pilot to prove you were right

The weakest pilot starts with a decision already made: we will continue whatever happens. Then it is not a test; it is a small campaign to justify the build.

Before starting, write down what will make you continue, adjust or stop:

  • Continue if customers complete transactions, partners deliver and willingness to pay is clear.
  • Adjust if the value is understood but the price, segment or channel is wrong.
  • Stop if every transaction needs founder intervention that cannot repeat, if no real willingness to pay appears, or if one loss reason keeps recurring and exposes a flaw in the offer.

Failing early in a small pilot costs far less than failing late after building a full product.

Manual operation is not a flaw

Manual operation does not scale. That is exactly why it should not be presented as the final operating model. Its job is to reveal what deserves automating.

If the team completes several transactions and finds the real bottleneck is confirming availability, the next feature becomes obvious. If the bottleneck is qualification, the problem is not the interface. If it is collection, a new dashboard will not fix it.

Automation after learning removes proven work. Automation before learning can speed up a process the market never wanted. The best MVP is not just the smallest technically; it is the smallest that completes a real value cycle.

Decision questions — check these inside your company

  • What is the payment moment in our current test, and has anyone reached it?
  • What result would make us stop, and have we written it down?

About the author

Riyadh-based operator: co-founded PH Store (acquired by Grinta) and HOMEi; former Grinta Sales Director; VP at Sahat Venture Studio; Founder & CEO of HADREEN.

Full backgroundLinkedIn profile

Last reviewed 31 Oct 2026.