Clients ask “what if I pay and get nothing?” and builders ask “what if I build it and never get paid?”. Both fears are real. Our answer is to split the risk with clear rules rather than let one side carry all of it. This article explains how.
Why is commissioning software risky?
- Paying upfront with nothing to see, only a promise.
- A verbal or vague description of the work, followed by arguments about “what was supposed to be there”.
- Code and publishing accounts stay with the builder, so the client depends on them; the app order checklist exists to prevent exactly this.
- Written brief
The goal, a fixed price and approval criteria go on paper.
- Live sample
In about 3 working days you see and try the sample.
- Approval
Within 7 days you say “approved” or “rejected, with the failing item”.
- Payment
Within 48 hours of approval. If you don’t like it, you owe nothing.
- Full handover
Source code and accounts move into your name.
What does “live sample first, then pay” mean?
- A call and a written brief: the goal, a fixed price and the acceptance criteria (a testable pass/fail list) go on paper.
- A live sample: on small projects, in about 3 working days you see and try a working sample, with no payment.
- Approval: within 7 days you say “approved” or “rejected, with the failing item” against those criteria.
- Payment: within 48 hours of approval. If you don’t like it, you owe nothing and the sample is removed.
- Full handover: source code, accounts and access in your name; ownership transfers on full payment.
What do you see before paying, per service?
| Service | Before payment | After payment |
|---|---|---|
| Android app | A sample APK limited in capacity or time, a video or a live walkthrough | Activation code, final build, source code, publishing accounts |
| Chatbot and automation | An online sample on a few of your own questions | Access, knowledge base and settings |
| Website design | The home page and one inner page on a temporary, unindexed address | Move to your domain and all files |
| AI model | A proof of concept on sample data with an evaluation measure | The full model after approval; larger projects in stages |
Our own apps show the same thinking: Chekyar’s trial lets you record 10 cheques before buying; see also the case study of four Persian apps.
What protects both sides?
- Written acceptance criteria: approval isn’t a matter of taste, and the builder can’t revise forever.
- The sample is for evaluation only: limited, watermarked or time-boxed; no commercial use.
- Deadlines for deciding and paying: nobody is left waiting.
- Ownership after payment: the client becomes the owner and the builder gets paid for the work.
Where doesn’t this apply?
For large projects (a custom model, a CRM, a complex app), building everything for free makes no sense. There it is a proof of concept (the core feature) first, then staged payments after you see each stage; see what a chatbot and model cost for estimating.
Test any other builder with these five questions
- What can I actually see before I pay?
- Where are the acceptance criteria written down?
- Whose name are the source code and publishing accounts in?
- What happens if I don’t like it or the work stalls?
- Do you have real, try-able work?
An example of a trade app built with a live sample and field testing: KiloChand.
For an example of pricing and a live sample in a clinic, read what an automatic reply system costs a clinic.
Frequently asked questions
Is “see first, then pay” for every project?
No. It is for small projects (a live sample in about 3 working days). Larger projects use a proof of concept and staged payments.
What if I don’t like the sample?
You pay nothing. The temporary sample and your data are deleted from the server, and the acceptance criteria are written in advance.
Can I use the sample commercially?
No. The sample is for evaluation only; commercial use is possible after payment and final delivery.
Who owns the code?
After full payment, the source code, publishing accounts and access are handed over in your name.