Short answer
Judge a development agency on how they handle ambiguity, not on their portfolio. Ask what is excluded from the quote, who owns the code, what happens when the project runs late, and what they would refuse to build. Start with a small paid project rather than committing the full budget on the strength of a pitch.
Why portfolios tell you almost nothing
Every agency shows their best work, and you cannot tell from the outside how much of it was the agency and how much was a strong client with clear requirements. You also cannot see the projects that went badly, ran late or ended in a dispute.
What actually predicts a good engagement is how a supplier behaves when things are uncertain or going wrong. That is what the meeting should be testing.
The questions worth asking
Ask what is excluded from the quote. Good suppliers answer immediately and specifically, because they have thought about it. Vague answers here reliably become invoices later.
Ask who owns the code and the accounts at the end. The only acceptable answer is that you do, unconditionally, on settlement.
Ask what happens when the project runs late — not if, when. You are listening for a process: how they tell you, how re-planning works, who absorbs the cost of their own estimation error.
Ask what they would refuse to build. A supplier who says yes to everything has no judgement, and you are paying for judgement as much as for hours.
Ask who will actually do the work, and whether the person in the meeting is one of them. Being sold by seniors and delivered by juniors is common and worth knowing about in advance.
Red flags that reliably predict trouble
A quote arriving without questions being asked is the clearest one. Anyone who can price your project from a paragraph has either built it before, in which case they should say so, or is guessing.
Pressure to sign quickly, discounts that expire, or a proposal you cannot take away and read are all signs of a sales process rather than a delivery business.
Reluctance to name what is excluded, evasiveness about code ownership, or hosting arrangements where they hold the domain in their own name are all mechanisms for making you hard to leave.
Finally, be wary of a supplier who never disagrees with you. If you have described something impractical and nobody says so, you are being sold to, not advised.
How to compare proposals fairly
Read the exclusions first, then the assumptions, then the price. A proposal with no stated assumptions has not been thought about.
Normalise for scope before comparing numbers. Two quotes for "a website" may differ by content, testing, integrations, revisions and post-launch support, and once you add the missing items the cheap one is frequently dearer.
Weigh how each supplier handled the pre-sales process, because that is a fair sample of how they will handle the project. A supplier who was slow to reply, vague about detail and reluctant to put things in writing before they had your money will not improve afterwards.
Structure the first project to limit risk
Whatever the eventual scope, start with something small, real and paid. A discovery phase, a single module, one page type — enough to see how they work without committing the full budget on the strength of a pitch.
Insist on staged payments tied to reviewable deliverables. Never pay the whole fee upfront, and be equally wary of a supplier willing to do substantial work with no deposit — that usually signals they need the work more than they should.
Key points
- Portfolios show best cases. Judge suppliers on how they handle ambiguity and bad news.
- Always ask what is excluded, who owns the code, and what happens when the project runs late.
- A quote given without questions is a guess. Treat it as one.
- Normalise proposals for scope before comparing prices.
- Start with a small paid project and stage payments against reviewable work.
Frequently asked
Should we choose the cheapest quote?
Only after normalising scope. Cheap quotes are frequently cheap because they exclude content, testing, revisions or support that the dearer one included.
Is it risky to use an offshore agency?
It carries different risks, mainly around communication and ambiguity. A small paid pilot, written scope and a fixed review rhythm reduce most of them.
How much should we pay upfront?
A deposit of 25–50% is normal, with the balance staged against milestones. Paying the full fee upfront removes your only real leverage.
Written by the Softech Team team. Last reviewed September 2026. This guide is general information, not advice for your specific circumstances.