How to choose a web design agency
Every proposal reads the same and every portfolio looks good. These are the questions that reveal the difference, and the signals worth taking seriously.
Choosing an agency is hard because you are buying something that does not exist yet. You cannot try it first, and the proposals you receive all use the same words: strategic, custom, results-driven. Portfolios do not help much either, because everyone shows their best work and nobody shows the project that went wrong.
What actually distinguishes them is process: how they work, what they ask before quoting, and what happens when something does not go to plan. This guide is the set of questions that reveals that on a first call.
First: decide whether you really need an agency
This is the part no agency writes, which is exactly why it belongs first. We are not always the right answer.
If your project is a simple page with content you already have, a good freelancer will cost considerably less and respond faster. If you are still validating whether the business works, a well-configured template lets you test without committing budget you may need elsewhere.
An agency starts to make sense when the project needs several disciplines at once —research, architecture, design, development—, when the site is your main commercial channel, or when it has to stay alive and grow for years. If none of those describe your case, hiring an agency means paying for structure you will not use.
The questions that reveal how they work
Ask them on the first call. They are not looking for one correct answer, they are looking for whether an answer exists at all.
- What do you need to know about my business before you can quote? If they need nothing, they are quoting a template.
- Who will work on my project and who will I talk to? If those are different people, ask how many layers sit between you and whoever designs.
- How many projects do you run at once? There is no correct number, but the answer tells you how much attention you will get.
- What happens if I still do not like it on the third review round? What matters is not the policy, it is that one exists.
- How do you measure whether the site worked after launch? If the answer is “it looks great”, there is no measurement.
- Can you show me a project that did not go as expected and what you learned? It is the most uncomfortable question and the most revealing.
Red flags
None of these disqualifies on its own, but two or three together do.
- They give you a price before asking anything substantial about your business.
- They guarantee first place on Google. Nobody can guarantee that, and whoever promises it either does not know or knows and says it anyway.
- The proposal does not state how many review rounds are included or what happens afterward.
- It is unclear who owns the code and the domain at the end.
- The portfolio has no project of your size.
- They push a limited-time discount to close. A project decided under urgency gets regretted for months.
How to actually read a portfolio
A portfolio shows aesthetic capability, which is the easy part to evaluate and the least decisive. What is worth looking for is something else: variety. If all ten projects resemble each other, you are probably looking at a mold that will be applied to you too.
Look for projects of similar complexity to yours as well. Someone who has only built five-section sites can build an excellent five-section site, but your eight-hundred-product catalog is a different problem. And if a case interests you, ask exactly which part they did: many portfolios include projects where only the interface was built on someone else’s strategy.
Price is the last question, not the first
Comparing proposals by their total is the fastest way to choose badly, because the total does not say what is included. A more expensive proposal with content, training, and maintenance included can end up costing less than a cheap one that keeps adding line items.
Sort proposals by scope first. When two proposals cover exactly the same ground, then yes, take the cheaper one: at that point the comparison means something.