Most redesigns start with a feeling. The site looks dated, or it does not bring in enquiries. A feeling is not a diagnosis, and a redesign built on one often reproduces the same problems in a newer style.

This checklist turns the feeling into a list. It takes an afternoon, a laptop and a phone. Score each check yes, partly or no. The "no" answers are your brief.

How to run it

Look at three pages: the home page, the page that sells your main service, and the contact page. Read them as a stranger would, on a laptop first and then on a phone. Give yourself ninety minutes and write down what you see, not what you intended.

Message: five checks

  1. The headline says what you do and who it is for. Cover the logo. Would a stranger know what the company sells?
  2. The first screen has one clear next step. Not three buttons of equal weight.
  3. The page names the problem in your customer's words. The phrases people use when they ask for help, not your internal service names.
  4. Proof sits next to the promise. Real clients, real results, real certifications. If you have none yet, say what you do to earn trust instead.
  5. There is no jargon a buyer would have to look up.

Structure: five checks

  1. Any service can be reached in two clicks from the home page.
  2. Navigation labels match what people expect. "Work", "Services" and "Contact" beat clever names.
  3. Every page has one job and one next step.
  4. Content follows the order of a decision. What you do, proof, how it works, what it costs, how to start.
  5. No page is a dead end. Each one finishes with something to do.

Trust: four checks

  1. Contact details are easy to find. An email address, and a phone number or a booking link if you take calls. If you work remotely, say so.
  2. People are visible. A name, a face or at least a plain explanation of who is behind the business.
  3. Policies exist. A privacy policy at minimum, and terms if you sell something.
  4. Nothing looks abandoned. Dated articles are recent, the copyright year is current and links work.

Mobile: four checks

  1. The headline and the main action fit on the first screen of a phone.
  2. Body text is readable without zooming. 16 pixels is a sensible minimum.
  3. Buttons and links are easy to tap. Aim for targets about 44 pixels square with space between them.
  4. Nothing scrolls sideways, and forms work with the phone's keyboard. An email field should bring up the email keyboard.

Speed: three checks

Google publishes three measurements of how a page feels, called Core Web Vitals, with a "good" threshold for each. Run your pages through PageSpeed Insights or the Lighthouse panel in Chrome and compare.

  1. The main content appears within 2.5 seconds (Largest Contentful Paint).
  2. The page does not jump while it loads. Layout shift under 0.1.
  3. The page responds quickly to taps and typing. Interaction to Next Paint under 200 milliseconds.

Lab tests on your own connection are a starting point. If the report shows field data from real visitors, trust that over a single run.

Accessibility and search basics: four checks

  1. Text contrast is at least 4.5 to 1, or 3 to 1 for large text. Grey text on white is the usual failure.
  2. Images that carry meaning have alternative text. Decorative images have none.
  3. The whole page works with a keyboard, and you can always see which element has focus.
  4. Every page has its own title and description, and one main heading. This is also most of what search engines need from a page.

Reading the result

Count the "no" answers by section.

  • Mostly yes everywhere: you probably do not need a redesign. Fix the details.
  • Most "no" answers in message and structure: the problem is content and information architecture. A new visual design on top of the same structure will look better and perform the same. Fix the words and the page order first.
  • Most "no" answers in mobile, speed or accessibility: the problem is how the site is built. A rebuild, or a focused technical fix, will pay off.
  • A mix: start with a review, so the fixes are ranked by impact and effort instead of by whoever shouts loudest.

What to do next

Rank the "no" answers by two questions: how much would fixing it help a visitor, and how much work is it? Message fixes are usually the cheapest and the most valuable. Do those first, then run the five-second test again: show the home page to someone who does not know the business and ask what the company does, who it is for and what they would do next.