How to Shortlist Web Development Companies in One Week: A Day-by-Day Process

Most business owners who need a new website do not lack options. They have too many. A search returns pages of agencies, directories list hundreds of firms, a colleague recommends someone, and an AI assistant suggests a few more names. Every one of them shows attractive work and promises a site that brings in leads. Without a process, the choice ends up being made by whoever answered the email first or whoever had the nicest homepage.

A shortlist fixes that. It turns a vague search into a short series of decisions, each with a reason you can write down. The goal of this guide is to help you shortlist web development companies in five working days, moving from roughly thirty candidates to three finalists who each receive the same brief and the same questions. It stays neutral about providers and does not rank anyone.

If you have not started looking yet, our guide to finding the best web development companies in the USA covers where buyers look and which sources deserve trust. This article picks up from there and focuses on the routine: what to do each day, how long each step should take, and how to keep the comparison fair.

What does it mean to shortlist web development companies?

A shortlist is the small group of companies you will actually spend time with: calls, proposals, reference checks and a final decision. Everyone outside the shortlist gets no more of your time. That makes the shortlist the most important filter in the whole hiring process, because the best proposal you will ever see can only come from a company that made the cut.

To shortlist web development companies well, you need three things. First, a written description of the project, so every company is measured against the same need. Second, a fixed set of criteria, so you do not change the rules halfway through because one firm made a good impression. Third, written reasons for every inclusion and every rejection, so the final choice is something you can explain to a partner, a board or yourself six months from now.

What a shortlist is not

• It is not a ranking. The three finalists are not first, second and third. They are three companies that met the bar and deserve a full comparison.
• It is not a price list. Price matters, but you cannot compare prices until the scope is the same. That happens after the shortlist, when finalists respond to one brief.
• It is not a popularity contest. A directory position, a badge or a large social following tells you about marketing. The shortlist is about evidence that a company can deliver your specific project.

The most common trap is falling for one company early and using the rest of the week to confirm the choice instead of testing it.

How many companies should be on the shortlist?

There is no official number, and you should be suspicious of anyone who claims there is. The right size is the number of qualified companies you can evaluate consistently with the time you have. For most small and mid-sized business projects, that usually works out to a funnel shaped like this: around thirty names at the start, about ten after a quick desk review, and three finalists after verification and calls.

The numbers are a planning tool, not a rule. Thirty is large enough to include different types of firms (small studios, larger agencies, platform specialists, local and national teams) without becoming a research project. Ten is a number you can verify properly in a day. Three is the number most owners can compare in depth: three proposals, three sets of references, three teams to meet.

If your project is small and simple, two finalists can be enough. If it is large, regulated or highly technical, four finalists can be worth the extra time. What matters more than the number is that every company at each stage was judged by the same criteria.

A good shortlist is not the three companies you liked most. It is the three companies that survived the same tests, with the reasons written down.

Before day one: write a one-page brief

The week only works if you know what you are buying. Before you search for anyone, spend an hour writing a one-page brief. It does not need to be polished. It needs to be specific enough that two different companies would read it and understand the same project.

What to put on the page

• What the website has to produce. Calls, form submissions, bookings, sales, applications. Name the action that matters most.
• Who it is for. Your customers, where they are, and what they need to see before they contact you.
• The pages and features you know you need. Our guide on what pages a business website should have is a good starting checklist.
• Tools it has to connect to. CRM, booking system, payments, email marketing, phone tracking.
• Platform preference, if any. If you are unsure whether you need custom code or WordPress, say so. Our comparison of custom websites and WordPress explains the trade-offs.
• What already exists. A current site, brand assets, content, rankings you want to protect.
• Your constraints. A rough budget range, a target launch window and who signs off on decisions.

The brief has a second job. It tells you what kind of company belongs on the list at all. A booking-heavy service business needs different strengths than an online store or a membership portal. If your project is closer to software than to a marketing site, read our guide on website companies versus web app developers before you start, because it changes who you should be looking for.

Day 1: build a longlist of about thirty names

Day one is about breadth, not judgment. You are collecting names from several sources so that no single channel decides who you consider. Give yourself two to three hours and a simple spreadsheet with one row per company: name, website, where you found it, and one line on why it might fit.

Where the thirty names come from

• Search. Run searches that describe your project rather than generic phrases. “Custom WordPress site for a law firm” finds different companies than “web design company.”
• Directories. Review platforms list many firms with client reviews. Our comparison of Clutch, DesignRush and GoodFirms explains what each directory shows and how positions are earned.
• Referrals. Ask peers in your industry who built their site and whether they would hire them again. Our guide to finding web developers through Reddit, LinkedIn and referrals covers how to ask and how to weigh the answers.
• Sites you admire. Look at the footers of competitors and businesses you respect. Many sites credit the company that built them.
• AI assistants. They can produce names quickly, but treat every suggestion as unverified until day three.

Do not filter hard yet. Include a mix of sizes and models: small studios, larger agencies, freelancers if your project suits them, local teams and national ones. If you are unsure about the trade-offs, our guides on boutique versus large agencies and local versus national web development companies help you decide which mix is worth keeping on the list.

By the end of day one you should have around thirty rows. If you have far more, you searched too broadly; tighten the brief. If you have far fewer, add a source you skipped. If you would rather talk through the brief with someone before you search, a Blueprint Briefing is a short call about what your website has to produce, and the brief you leave with works for any company on your list.

Day 2: cut to ten with a desk check

Day two is where the real filtering starts. You will spend about ten minutes on each company, looking only at what is public: its website, its portfolio and the way it describes its work. The question for each one is simple: is there evidence this company has done a project like mine?

The ten-minute desk check

1. Read the services page. Does the company actually do the type of work in your brief, or is it one line in a long list of services?
2. Open three portfolio items closest to your project. Same industry, same kind of site, or the same technical challenge. If there are none, that is a strong signal.
3. Check that portfolio items are real. Click through to live sites. Screenshots without links are weaker evidence.
4. Look for who does the work. Named people, roles and locations are better than stock photos and a generic “our team.”
5. Read how they describe their process. A clear sequence from discovery to launch suggests experience. Vague promises do not.
6. Note what they say about ownership. Who owns the domain, hosting and code after launch is a question you will ask later, but some firms answer it on their site.

Give each company a quick mark in your spreadsheet: keep, maybe, or cut, with one sentence of reasoning. Keep the companies with clear evidence of similar work. Cut the ones with no relevant examples, no visible team, or a site that contradicts what they claim to sell. A web company whose own site is slow, broken on a phone or hard to navigate is telling you something.

If the labels confuse you at this stage (design studio, development company, digital agency), our guide on web development companies versus web design companies explains what each one usually delivers. Aim to end the day with about ten names marked keep.

Day 3: verify the ten in public records

Day three moves from what a company says about itself to what can be checked independently. None of these checks proves a company is good. Together, they show whether it is real, established and honest about who it is. Plan on fifteen to twenty minutes per company.

Business registration

Every state runs a business entity search, usually through the Secretary of State’s office. Look up the legal name of the company, confirm it is active, and note when it was formed and where it is registered. Compare that with what the company claims on its site. Our guide on how to verify a U.S.-based web development company walks through the full set of checks, including how to tell a U.S.-based team from a U.S. sales office.

Domain history

The ICANN Lookup tool shows registration data for a domain, including when it was created. A company claiming many years in business with a domain registered last year is not necessarily lying, since businesses rebrand, but it is a question worth asking on the call.

Reviews, read carefully

Read reviews on more than one platform, and read the negative ones as closely as the positive ones. Look at how the company responds. The Federal Trade Commission’s rule on consumer reviews and testimonials, which took effect on October 21, 2024, prohibits fake reviews, reviews bought for a required sentiment, undisclosed insider reviews and review suppression; the FTC’s questions and answers on the rule explain each one. You cannot audit a company’s reviews, but you can notice patterns: a sudden burst of five-star reviews with no detail, or reviewers who only ever reviewed this one business.

For local signals, our guide on using Google Business Profile reviews to vet a web development company covers what the profile shows and what it hides. For the broader question of what “established” should mean, see what makes a web development company established. By the end of day three, most lists shrink to six or eight companies.

Day 4: test the portfolios on live sites

Screenshots show how a site looked on launch day. Live sites show how it works now, which tells you far more about the company that built it and, often, about how it supports clients after launch. Pick two or three live portfolio sites per company, the ones closest to your project, and spend a few minutes on each.

What to test on each live site

• On your phone. Open the site on a phone, not only a laptop. Can you find the main action in a few seconds? Do menus, forms and buttons work?
• Speed and stability. Run the page through PageSpeed Insights, Google’s free tool. Do not treat the score as a verdict, since content, hosting and later changes by the client all affect it, but a pattern of very poor results across a company’s portfolio is worth asking about.
• The path to contact. Try to reach the business: a form, a phone link, a booking widget. If you cannot, neither can that client’s customers.
• Basic accessibility. Can you move through the page with the keyboard? Is text readable against the background? Do images have meaningful descriptions?
• Signs of care. Broken links, placeholder text, outdated copyright years and missing pages suggest the site was launched and left.

Remember that clients sometimes change their sites after launch, so one bad example is not proof of anything. Look for the pattern across several projects. A company whose recent work is consistently fast, clear and working is showing you what it builds by default.

Day 4, afternoon: send the same first message to everyone

Once the portfolio tests are done, contact the companies still on the list, usually four to six. Send every company the same short message and the same one-page brief. This is the first time you see how each company behaves with a real prospect, and it is one of the most useful tests of the week.

What the first message should ask

• Whether the project fits their work, and if not, why not.
• One or two live examples closest to your brief, with a note on what their team did on each.
• Who would lead the project and who would build it, with names and roles.
• Availability for a twenty-minute call on day five.

Pay attention to the replies. Did the company answer the questions you asked, or send a generic sales deck? Did someone read the brief? A company that says honestly that your project is not a fit is doing you a favor and deserves a note in your spreadsheet as a company to remember for other work.

Silence is information too. If a company does not reply within the time you gave, it is reasonable to drop it. How a company treats you before it has your money is rarely worse than how it treats you after. If you want GoSocial to receive the same message, you can book a Blueprint Briefing instead, and we will answer the same four questions on the call, whether or not we end up on your list.

Day 5: short calls and the final three

Day five is for conversations. Book twenty to thirty minutes with each company still on the list and ask to speak with the person who would lead your project, not only a salesperson. Use the same questions with everyone, take notes in the same format, and score each call right after it ends, while it is fresh.

Questions for the call

1. What would you build for this brief, and why? A good answer refers to your goals, not to their favorite platform.
2. Who exactly would work on my site? Ask whether those people are employees or contractors and where they are based.
3. What does your process look like from discovery to launch? Listen for clear steps, review points and who approves what.
4. What do you need from me, and when? Content, access, decisions. Projects slow down when this is unclear.
5. Who owns the domain, hosting, code and analytics after launch? The answer should be you.
6. What happens after launch? Updates, fixes, support, search visibility, and how those are priced.
7. What would make you turn this project down? Companies that can answer this tend to know their strengths.

After the calls, pick three finalists. Use your notes, not your memory of who was most charming. If you are torn between two for the last spot, keep the one with stronger evidence of similar work. A company that was honest about a weak area can still be the right choice if that area is not central to your brief.

The shortlist worksheet: criteria you can score

A worksheet keeps the week consistent. Each criterion below is something you can observe or check, scored on a simple scale such as 0 (no evidence), 1 (some evidence) and 2 (clear evidence). The weights are yours to set based on the brief; a project with heavy integrations should weigh technical evidence more than visual style, for example.

Do not let the total score make the decision for you. Use it to spot gaps. A company with a high score but a zero on ownership terms needs a direct conversation about that zero before it goes further, because some gaps are harder to fix than others.

An illustrative week, step by step

The example below is invented to show how the process plays out. It does not describe any real business or company.

The brief. A home services company with two locations wants a new site that produces booked estimates. It needs online booking connected to its scheduling tool, service pages for each service, a reviews section and call tracking. The owner is open to custom code or WordPress and wants to keep the site’s existing search traffic.

Monday. The owner collects thirty-one names: twelve from search, nine from a review directory, four from peers in a trade association, four from the footers of sites they admire and two from an AI assistant.

Tuesday. The desk check cuts the list to eleven. Eight companies show no home services or booking work at all. Five list dozens of services with thin portfolios. Four portfolios link only to screenshots. Three company sites are hard to use on a phone.

Wednesday. Verification cuts three more. One company’s registration does not match the name on its site and it has no explanation online. One of the AI suggestions turns out to have closed. One has a cluster of short, identical reviews posted in the same month. Eight remain.

Thursday. Live portfolio tests remove two companies whose recent sites have broken booking forms. The remaining six receive the same message and brief. Five reply; one sends a generic deck without reading the brief.

Friday. Four calls happen. One company is honest that booking integrations are not its strength. The owner keeps three finalists, each with written notes on why, and sends them the full brief for proposals.

The point is not the exact numbers. It is that every cut had a reason the owner could repeat later.

Document every inclusion and rejection

Writing down your reasons feels like extra work during the week. It saves far more time later. When a finalist’s proposal surprises you, you can check your notes to see what you expected. When a partner asks why a well-known firm was not considered, you have the answer. When the project ends and you look back, you can see which signals were accurate and use them next time.

A simple format for each company

• Status: kept, cut, or finalist, and on which day.
• Reason in one sentence: “No live examples of booking integrations” or “Strong relevant portfolio, named team, clear ownership terms.”
• Evidence: the link, record or note that supports the reason.
• Open questions: anything you would ask if the company came back into consideration.

Rejection reasons should be about evidence, not feelings. “Did not like their website” is a weak reason. “Their site is hard to use on a phone, and two of three portfolio sites have the same problem” is a strong one. The discipline of writing it down tends to make the reasons better.

This record also protects the fairness of the comparison. If you notice that you cut one company for a missing case study but kept another with the same gap, you can correct it before it shapes the final decision. If you want a team to look at your notes and your brief with you, you can schedule a Blueprint Briefing; bring the worksheet and we will tell you where we would score ourselves.

Shortlisting with less time, or with a team

Not everyone has five days. Some owners need a decision faster; others need to involve partners, managers or a board. The process adapts in both directions.

If you only have two or three days

• Start with fewer names, around fifteen, drawn mainly from referrals and the closest portfolio matches.
• Combine the desk check and verification into one pass per company.
• Test one live portfolio site per company instead of three.
• Hold calls with your top three or four only.

What you should not compress is the brief and the written reasons. Those two pieces make the comparison fair, and they take less time than they save.

If several people are involved

• Agree on the brief and the worksheet criteria before anyone starts searching.
• Split the desk check: each person reviews part of the list using the same criteria.
• Have at least two people on every call, scoring separately, then compare scores.
• Name one person who makes the final decision, so the process does not stall.

Using AI search to help, carefully

AI assistants and AI search features can speed up parts of the week. They are useful for summarizing what a company says about itself, drafting your first message or turning call notes into a clean table. They are less reliable as a source of names or facts about specific companies.

The main risks are that suggestions can be outdated, that companies can be mixed up, and that an assistant may present a company’s own marketing as an independent view. Our guide on how to evaluate AI recommendations for a web development company explains how these suggestions are produced and how to check them.

Good uses of AI during the week

• Turning your notes into a first draft of the brief.
• Generating a list of questions for the call, which you then edit.
• Summarizing a long services page so you can compare it with others.
• Formatting your worksheet and notes.

Things AI should not decide for you

• Whether a company exists, is active or is based where it says.
• Whether its reviews are genuine.
• Whether its portfolio sites work today.
• Who belongs in the final three.

Every name an AI tool suggests goes through the same day three verification as every other name. No exceptions.

Red flags that remove a company on any day

Some findings are serious enough that a company should leave the list whenever you discover them, regardless of how it scores elsewhere. Write them down in advance so you are not tempted to make an exception for a company you like.

• Promises of guaranteed results. No one controls search rankings or citations by AI systems. A company that promises a specific position or result is either misinformed or misleading you.
• Refusing to name the team. If you cannot learn who would work on your project, you cannot evaluate them.
• Ownership kept by the provider. If the domain, hosting or code would stay in the company’s name, you may find it hard to leave later.
• Records that do not match. A business name, location or history that contradicts the public record, with no reasonable explanation.
• Review patterns that look manufactured. Bursts of identical, detail-free reviews, or reviewers with no other history.
• Pressure tactics. Discounts that expire by the end of the call or a push to sign before you have seen a written scope.
• Portfolio items that are not theirs. Sites that credit another company, or examples the company cannot explain.

Not every concern is a red flag. A small team, a young company or a portfolio with few examples in your industry can still be a good fit. Red flags are about honesty and ownership; ordinary gaps are things to ask about.

What happens after the shortlist

The shortlist is the end of the search and the start of the decision. Send each finalist the same full brief, with the same deadline and the same request for a proposal format: scope, process, team, timeline, what is included after launch and how the price is built. Proposals that follow the same structure are far easier to compare.

The next steps, in order

1. Send one brief to all three finalists. Answer every question in writing and share the answers with all three, so no one has an information advantage.
2. Compare proposals against the brief. Look first at what is missing, then at what is extra, and only then at price.
3. Check references. Ask each finalist for two or three clients with similar projects, and ask those clients what went wrong and how it was handled.
4. Read the contract. Confirm ownership, payment milestones, change requests, support terms and what happens if either side ends the project.
5. Decide, and tell everyone. Let the other finalists know. A short, honest note keeps the door open for future work.

If the project includes a redesign of an existing site, ask each finalist how it will protect the pages and search visibility you already have. Our website redesign process guide shows the steps that matter. And if you are still weighing whether a freelancer could handle the work instead of a company, the comparison of a web development agency versus a freelancer is worth reading before the final decision. When you are ready to hear how GoSocial would approach your brief, book a Blueprint Briefing and we will walk through it on a short call.

Common mistakes when shortlisting web development companies

Most shortlisting problems come from skipping a step, not from choosing the wrong criteria. The table lists the patterns that come up most often, why they cause trouble and what to do instead.

A fixed process is what keeps urgency from making the decision.

Where GoSocial fits

Since you are reading this on our site, it is fair to say where we sit. GoSocial is a custom web design company in Boca Raton, Florida, building websites for businesses across the United States, and it has worked across web development and digital marketing since 2016. One accountable team, CEO-led: the people who plan the site build it. You can read more about how we work on our web development services page.

GoSocial builds both custom-coded websites and custom WordPress websites, and recommends the fit for each project. Every project follows the same process: discovery, strategy, architecture, design and content, development, testing, launch and measurement. Every GoSocial website launches with GA4 and Google Search Console connected and tracks form submissions, phone clicks, WhatsApp clicks and bookings. Our websites are strengthened with SEO, AEO and GEO on one foundation, as our SEO, AEO and GEO services page explains, and you can see live examples in our portfolio.

With GoSocial, the domain, hosting account and code belong to the client, so another team can maintain the site if the client ever chooses. We do not promise a performance score, a ranking or an AI citation; we build for the metrics and show the measurements. If your project needs a specialist we are not, we would rather say so on the first call. You can also reach us through our contact page.