How to Choose a Web Development Agency (Without Getting Burned)

Search "best web development agency" and you'll get a hundred listicles telling you to check the portfolio, read the reviews, and "make sure they understand your vision." All true. All useless. None of it predicts whether the thing you pay for actually ships, loads fast, and survives the first time you need to change it.

We've inherited enough half-finished projects from other agencies to know what the real predictors are. They have less to do with the showreel and more to do with how a team answers a few specific questions. Here's what to look for — and what to walk away from.

Watch how they run discovery

The first call tells you most of what you need to know. A good agency spends it interrogating your business: who buys from you, what the site is supposed to do, where the current one leaks, what "success" looks like in actual numbers. A weaker one spends it asking which competitor sites you like and whether you prefer blue or green.

If a team reaches for mockups before it understands your funnel, you're buying a brochure, not a business asset. The questions should feel a little uncomfortable — they should know more about how you win customers by the end of the call than you planned to share.

Find out who actually does the work

This is the one almost nobody checks, and it's the one that predicts the most.

Plenty of agencies sell you a polished senior team and then hand the build to whoever's free — often a rotating cast of juniors and subcontractors. The work degrades quietly: context gets lost between people, decisions get re-litigated, and three months in nobody can tell you why a particular thing was built the way it was. Frequent developer turnover inside an agency is a quiet red flag that shows up later as a codebase no one fully owns.

So ask it directly: who specifically will design and build this, and how many other projects are they on at the same time? The answer tells you the model you're actually buying — regardless of what the pitch deck says. (Ours is deliberately small for exactly this reason: the people on the call are the people who do the work.)

Read the pricing as a signal, not just a number

How an agency prices tells you how it thinks. The teams worth hiring are upfront about the model — fixed scope, retainer, or time-and-materials — and they tell you what's not included in the same breath. Vagueness here ("it depends, let's see how it goes") usually means the scope creep is priced into your stress rather than their quote.

Be just as wary of the bottom of the market. A quote that's a fraction of everyone else's isn't a bargain; it's a forecast. The corners get cut somewhere you can't see on day one — performance, security, accessibility — and they resurface as expensive fixes the week after launch. The cheapest builds are the ones we get called in to rescue most often.

Make the portfolio prove something

Pretty screenshots are table stakes. What you're actually looking for is evidence the work shipped and worked: live links you can open, load times you can measure, case studies that name an outcome instead of an adjective. "We redesigned their site" is a screenshot. "We cut their main page load under two seconds and the bounce rate followed" is a result.

If an agency can't point you at live sites or talk specifically about Core Web Vitals, accessibility, and how it measures success after launch, that isn't modesty. It's a gap.

Treat the timeline as a character test

Anyone who promises a real, custom site in a week is either misjudging the work or counting on your urgency to get a signature before you've done your homework. Most honest redesigns land in the six-to-fourteen-week range, and the variance is usually driven by your side — content readiness, decision-making, scope discipline — not theirs. (We broke down the phase-by-phase version in how long a website redesign really takes.)

A team that gives you an honest, slightly inconvenient timeline is showing you how it'll behave when something gets hard mid-project. That's worth more than a fast yes.

Ask what happens after launch

Launch is the middle of the story, not the end. Before you sign, get straight answers on three things:

• Ownership. Do you own the code, the accounts, and the domain outright? You should never have to ask permission to leave.
• Handover. Will someone show your team how to run the thing — or are you locked into a maintenance contract by design?
• Support. What's the actual plan when something breaks at 9pm on a Friday? "We'll figure it out" is not a plan.

An agency that goes quiet the moment the site is live has told you exactly how much it valued the relationship.

The shortlist worth asking everyone

If you take one thing from this, take these. Ask every agency the same set and compare the answers side by side:

1. Who exactly will design and build this, and what else are they working on?
2. What's your pricing model, and what's explicitly out of scope?
3. Can I see two live projects and the measurable outcome of each?
4. What's a realistic timeline, and what most often delays it?
5. Do I own the code and accounts when we're done?
6. What does support look like after launch?

The right team answers all six without flinching. The wrong one gets vague somewhere around question two.