Restaurant app cost: ordering, click & collect and loyalty
10 min read
Key takeaways
- A fixed price does not remove the risk: it makes you pay for it up front, in the form of a safety margin.
- Time and materials is not more expensive: it is more honest about uncertainty, and requires real steering on your part.
- A capped budget with an adjustable scope is the best compromise in the majority of projects.
- A fixed price signed on a fuzzy scope produces change orders, not savings.
- Expect a 10 to 25% safety margin built into any serious fixed price.
The debate between fixed price and time and materials is often treated as a question of trust. It is actually a question of risk allocation: someone has to bear the uncertainty of a project, and the contract simply decides who.
Here are the three models, what they really cost, and the one we recommend depending on the situation.
The three models
Fixed price
A fixed price for a written scope. The provider bears the risk of overrun.
What you do not see: every serious fixed price includes a 10 to 25% safety margin. You pay for it whether or not the risk materialises. And anything that was not written down becomes a change order, negotiated from a position of weakness since the project is already under way.
When it is the right choice: genuinely frozen scope, short project, detailed and stable specifications, tender with several bidders.
Time and materials
You pay for the days actually worked. You bear the risk, and you keep full control of the scope.
What it requires: real availability on your side. A time-and-materials project without steering becomes a project without end, and that is the only justified criticism of this model.
When it is the right choice: a product that evolves with user feedback, an in-house team to be reinforced, a deliberately open scope, a long relationship.
Capped budget with adjustable scope
A commitment not to exceed, billing for time spent, and a prioritised scope that is adjusted as you go.
How it works: we estimate the whole thing, set a cap, and rank the features on three levels. At mid-point, we look at what has been consumed and decide what remains. You pay only what is real, you are protected on the upside, and nothing is ever blocked by a change order.
Our default recommendation: capped budget for a development project, fixed price for a well-delimited scoping or design phase, time and materials for an ongoing maintenance and evolution relationship. Most projects combine all three, phase by phase.
The honest comparison
| Fixed price | Time and materials | Capped budget | |
|---|---|---|---|
| Who bears the risk | The provider | The client | Shared |
| Cash-flow predictability | High | Low | High on the upside |
| Real cost for an equal scope | + 10 to 25% | Lowest | Close to time and materials |
| Flexibility during the project | Low | Total | High |
| Involvement required on your side | Low in appearance | High | Medium |
| Risk of change orders | High | None | Low |
| Suited to a fuzzy scope | No | Yes | Yes |
The "involvement required" row deserves a comment. A fixed price gives the impression that you can delegate entirely. It is a costly illusion: a project without an available point of contact produces a deliverable that matches the document and is unsuited to the real need. You will then have paid the agreed price for something that will have to be redone.
What really costs a lot: the change order
The change order is the economic tipping point of any fixed-price project. A two-day request billed mid-project structurally costs more than the same two days planned from the start: it has to be qualified, estimated, validated, contracted, then rescheduled into an already tight calendar.
Our rule, whatever the contractual model: every new request during the project replaces an item of equivalent effort, unless there is an explicit decision to increase the budget.
This rule has an immediate effect in meetings. What was "essential" becomes debatable again as soon as you have to name what you are sacrificing, and the scope stays under control without anyone having to say no.
10–25%Safety margin built into a serious fixed price
ProfileWhat first makes the daily rate of a French agency vary in 2026
2–4 hMinimum weekly availability expected on the client side
The clauses that really matter
Beyond the model, five points determine the quality of the relationship.
Code ownership. You must own the code produced and the associated accounts, from the moment of payment. Unambiguously, without a limited licence of use.
Acceptance arrangements. Who tests, within what time, against what criteria, and what happens in case of disagreement. Acceptance testing with no defined deadline blocks payments and poisons relationships.
The warranty. Three to six months after delivery, covering the correction of defects, not enhancements. Beyond that, it is maintenance, which is budgeted separately.
The people assigned. The quote mentions profiles; the contract should mention the principle of continuity. A change of team mid-project costs two to three weeks of handover.
Reversibility. Documentation, access, deployment procedure, transfer of accounts. To be planned at signing, never at the moment of break-up.
The right contract is not the one that protects you best on paper. It is the one that makes trade-offs visible at the moment they arise, rather than six months later in front of an invoice.
In summary
- Frozen scope, short project: fixed price, while demanding the written list of exclusions.
- Evolving product, user feedback: capped budget with a prioritised scope.
- Long relationship, team to be reinforced: time and materials, with a weekly steering point.
And in every case, the question that reveals the quality of a proposal remains the same: what is not included? The corresponding budget tiers are detailed in our article on the cost of a mobile app.
Sources
- Article L131-3 - Code de la propriété intellectuelle (in French) — Légifrance
