Insights · Working together · 12 August 2026
How to choose a development agency: eight questions
Choosing a build partner is the highest leverage decision in a project, made with the least information. These eight questions surface most of what matters in one call.
You will spend more decision effort choosing a phone than most companies spend choosing the agency that will build their product. The phone costs a thousand dollars and is replaceable in a day. The agency choice decides a year and a budget with another zero or two on it.
The information gap is the problem: every agency website says senior, agile, and passionate, and none of that is checkable. What is checkable is how they answer eight specific questions on the first call. We are an agency saying this, so hold us to every one of them.
The four questions about the work
Who exactly will work on this project, by name? The failure mode this catches is the bait and switch: impressive people pitch, different people build. You are not hiring a brand. You are hiring the four humans who will touch your product, so meet them before signing, not after.
Can we see something you built that is still live? Live matters more than beautiful. Anyone can screenshot a prototype; a product that has survived real users for a year is evidence of engineering, maintenance, and the unglamorous discipline in between. Then go one step further than most buyers: ask to speak to the client behind it.
How will we see progress? The only good answer is working software on a fixed schedule, weekly is healthy. Slide decks about progress are not progress. A team that will not commit to showing running software regularly is planning to have something to hide.
What happens when scope changes? Scope always changes, so the mature answer is a process: changes priced and agreed in writing before they are built. Beware both extremes, the rigid no and the cheerful unpriced yes. The second one is how a budget doubles politely.
The four questions about the terms
Who owns the code and designs, from what date? The only good answer is you, from day one, in your accounts. Ownership that arrives at final payment is leverage dressed as a milestone, and you will feel it at the worst possible moment.
What is in the quote, and what is billed extra? The surprises live in the second list: app store accounts, third party services, revision rounds, meetings beyond an allowance. Ask for the extras list in writing and watch how comfortable the answer is.
What happens after launch? A defined support window, then a clear choice between handover and a priced retainer. Vague promises of partnership are how you end up unable to ship a bug fix without an invoice negotiation.
How easily can we leave? Ask where the code lives, whose accounts things run in, and what leaving costs. A confident agency makes leaving easy, and that is exactly why its clients stay.
Reading the answers
No agency aces all eight, and the point is not a scorecard. What you are reading is comfort with directness. A partner who answers plainly before there is a contract will communicate plainly when something goes wrong after there is one, and something always eventually goes wrong. That moment, not the pitch, is what you are actually buying.
Watch for the tells on the way: hesitation on names, portfolios with nothing live, offence taken at the ownership question. Offence at that question is the most informative answer on the list.
Turning the answers into a decision
Take your best two candidates and give them the same small first phase, one to two weeks, clearly scoped, honestly paid. Watch how each one communicates, demos, and handles the first surprise. Two weeks of evidence beats any deck ever presented, and a good agency will respect a buyer who works this way, because the ones with nothing to hide have nothing to lose from being tested.