Start with the job the website has to do
Most poor choices start with a vague request. Ask five companies for a new website and you will get five different projects back, with no fair way to compare them. Before you contact anyone, write down what the site is for and how you will know it worked.
Write down, in plain language
- The main action you want a visitor to take, such as requesting a quote, booking a call or buying
- Who the visitors are and what they need to see before they act
- What the current site fails at, with examples you can point to
- Systems the site must connect to, such as a CRM, a booking tool, a payment provider or inventory
- Who will update the site after launch, and how comfortable they are with web tools
- Languages, regions and any legal or accessibility requirements you must meet
One page is enough. For a fuller structure, use our guide to writing a website project brief. A written brief also tells you something about each company. The good ones respond to it, and question it.
Know the kinds of company you are comparing
| Type | Usually a good fit when | Watch for |
|---|---|---|
| Freelancer | The site is small, the scope is clear and you can manage the project yourself | Availability, and what happens when one person is away |
| Marketing or creative agency | Brand, campaigns and content are the main work and the site itself is fairly standard | Development handed to subcontractors you never meet |
| Design and engineering studio | The site needs custom design, integrations, interactive features or room to grow into a product | Whether the senior people you meet are the ones doing the work |
| Template or site-builder specialist | Speed matters most and an existing theme covers what you need | Limits you will meet later in custom features, performance and ownership |
None of these is better in general. The right type depends on the job you wrote down. A five page brochure site does not need an engineering team, and a quoting tool with a product catalog should not be forced into a theme. If you are weighing a site builder against a custom build, see Webflow vs custom development.
Seven things to compare, and what good looks like
| Criterion | What good looks like | Warning sign |
|---|---|---|
| Relevant work | Projects similar to yours in purpose and complexity, with an account of what the company did | A gallery of screenshots with no context, or work that all looks the same |
| Process | Named stages, what you approve at each one and how changes are handled | Design starts before anyone asks about your goals or content |
| People | You meet the designer and the engineer who will do the work | A sales contact now and an unnamed team later |
| Content and search | A plan for who writes the content, page structure, redirects and metadata | Content treated as your problem, to be dropped in at the end |
| Performance and accessibility | Stated targets, tested on real phones before launch | Neither is mentioned until you ask |
| Ownership | The proposal says who owns the design files, the code, the domain and every account | Ownership is not mentioned, or depends on staying a customer |
| After launch | A written support option and a handover another team could work from | Vague promises of help if anything breaks |
Ask every company to show evidence for each row. A claim is not evidence. A page speed report for a site they launched is. So is a handover document from a past project, with private details removed.
Questions to ask on the first call
Use the same questions with each company so the answers can be compared. Listen for specifics. Strong companies answer with how they did something. Weak ones answer with adjectives.
- 01
Which of your projects is closest to ours, and what was hard about it?
A strong answer coversA named project, the problem, the decision they made and what they would do differently now
- 02
Who will work on our site, and how much of their time will we have?
A strong answer coversNames and roles, and an honest account of the other work those people carry
- 03
What do you need from us, and when?
A strong answer coversA dated list of content, access and decisions, because missing content and slow approvals are a common cause of delay
- 04
How do you handle a change of mind halfway through?
A strong answer coversA change process that shows the effect on cost and schedule before the work is done
- 05
How will the site be tested before launch?
A strong answer coversDevices and browsers, accessibility checks, forms, redirects and a review with your real content in place
- 06
What will we own at the end, and where will it live?
A strong answer coversCode in a repository you control, design files, and accounts registered in your company name
- 07
What happens after launch?
A strong answer coversA defined support period, how to report an issue and how quickly they respond
How to compare the proposals
Proposals arrive in different shapes, which makes the cheapest one look best. Put them side by side on the same criteria before you look at the totals.
- 01
Check that each proposal covers the same scope
List the pages, features, integrations and content work in each one. Ask for anything missing to be added or stated as excluded.
- 02
Give each criterion a weight
Decide what matters most for this project before you score. A site that must bring in enquiries weights content and search higher than animation.
- 03
Score each company on each criterion
Use a simple scale and write one line of evidence beside every score.
- 04
Compare cost last
Look at cost against the weighted score, and at what each price includes after launch.
A scorecard for three proposals
An equipment supplier needs a new site that brings in quote requests and connects to its CRM. The team weights seven criteria and scores each proposal from 1 to 5. The companies and figures are invented to show the method.
| Criterion | Weight | A | B | C |
|---|---|---|---|---|
| Relevant work | 3 | 4 | 3 | 5 |
| Process | 2 | 3 | 4 | 4 |
| People | 2 | 3 | 5 | 4 |
| Content and search | 3 | 2 | 4 | 4 |
| Performance and accessibility | 2 | 3 | 3 | 4 |
| Ownership | 2 | 5 | 2 | 5 |
| After launch | 1 | 3 | 4 | 3 |
| Weighted total, out of 75 | 49 | 53 | 64 |
Proposal A is the cheapest, and it scores lowest where this project needs the most, content and search. Proposal B has the strongest team but would keep the site on hosting the supplier cannot move. Proposal C costs more than A and wins on the criteria that carry the most weight. The scorecard does not make the decision. It shows what each price buys.
Signs you have found the right company, and signs you have not
Good signs
- They ask about your customers and goals before they talk about design
- They tell you what they would not build, or what you do not need
- Their proposal restates your goals in their own words
- They explain trade-offs and leave the decision with you
- Past work comes with context, including what did not go to plan
Warning signs
- A price before any question about scope
- A promised position in search results
- Pressure to sign quickly to keep a discount
- No clear answer on who owns the code, the domain and the accounts
- Portfolio links that no longer work, or live sites that look nothing like the screenshots
A checklist to run before you sign
Pre-signing checklist
0 of 10 checkedIf a company resists putting any of these in writing, treat that as an answer.
What to do next
Write your one page brief, shortlist three or four companies whose work resembles your project, and send each of them the same brief and the same questions. If Alpha Expansion is on your list, our website design service page explains how we work and what you receive, and our work page shows the kind of sites we design.
You can also send us your brief before it goes to anyone else. On a free 30-minute scoping call we will tell you what we would ask, change or cut.
Common questions
How many website development companies should I shortlist?
Three or four is usually enough. Fewer gives you little to compare. More turns the selection into a project of its own, and the answers start to blur. Spend the time you save on a better brief and on speaking to people who have worked with each company.
Is the lowest price a warning sign?
Not by itself. A lower price can mean a smaller scope, a simpler approach or a team with lower costs. It becomes a warning sign when the proposal leaves out content, testing, redirects or handover, because someone still has to do that work. Compare what each price includes before you compare the prices.
Does the company need to be in my city or country?
For most website projects, no. What matters is an overlap in working hours, a named contact and regular reviews you can attend. Check that the company can meet any legal, language or accessibility requirements that apply where your customers are.
Should I choose a company that specializes in my industry?
Industry experience helps when your site has domain rules, such as regulated content or complex product data. It matters less than experience with the same kind of problem. A team that has built quoting tools for manufacturers may suit a logistics quoting tool better than a logistics specialist that has only built brochure sites.
What should the contract say about ownership?
It should say who owns each part and how you may use it. The parts are the source code or the builder project, the design files, the domain, the hosting and analytics accounts, and the documentation. Ask for this list in the proposal, not at handover.
