We've heard the same sentence dozens of times this year. "Our website doesn't bring in leads."

The honest answer is that very few websites are broken in the way the owner thinks they are. They don't need a rebuild. They need five specific things checked, in order. We charge for the audit; you can do most of it yourself in an afternoon.

Here's the diagnostic.

1. The page promises the wrong thing

A homepage that says "We make beautiful websites" speaks to nobody. The visitor isn't looking for a beautiful website. They're looking for more customers, or fewer support tickets, or a property that sells in three weeks instead of five months.

Write the headline as a sentence that names the visitor's outcome, not your service. If you sell SEO, the headline shouldn't say "SEO Services Cardiff". It should say "Rank where your customers are searching."

A quick test: read your hero copy out loud as if you were standing across a desk from a real prospect. If it sounds like a brochure, it'll convert like a brochure.

2. The page is too slow

We see this on roughly 70% of audits. The site looks fine in the studio's lighthouse report because they tested on a 1Gbps connection with a Macbook. On a real prospect's phone, on 4G, in a car park, the page takes seven seconds to become useful.

Open your homepage on PageSpeed Insights and read the Largest Contentful Paint number. If it's over 2.5 seconds on mobile, you are losing leads before the page finishes loading. The usual offenders:

• A hero video or 4K hero image that should have been a properly compressed JPEG
• Five fonts being loaded when two would do
• A chatbot widget that ships 400KB of JavaScript before the visitor blinks
• A WordPress page builder that renders 18 nested div wrappers around a single button

Fix LCP first. The other metrics tend to follow.

3. There's no clear next step

Count the calls-to-action above the fold of your homepage. If the answer is more than two, the page has no point of view. The visitor doesn't know whether they're meant to book a call, request a quote, download a brochure, follow you on Instagram, or watch the founder video.

The fix is unglamorous: pick one primary action. Make it the most visually prominent thing on the page. Make every other link quieter, same font weight as body text, no border, no fill. The page should feel slightly imbalanced in favour of the action you actually want.

If your business genuinely has two equally valid actions (e.g. "buy now" and "book a demo"), pick the one with the higher long-term value and demote the other to a text link. Half a primary action is worth two thirds of an action.

4. The page targets the wrong search intent

This one is for the SEO-curious. If you're ranking but not converting, your content is matching the wrong query.

A page titled "Affordable web design Cardiff" attracts visitors looking for the cheapest possible website. They will haggle. They will not buy your £15k retainer. The mismatch isn't a traffic problem, it's an intent problem.

Pull your top 20 ranking queries from Google Search Console. For each one, ask: would this searcher actually want what we sell? If most of the answers are "no, they want something cheaper / smaller / different", the content strategy needs a rewrite, not the website.

The fix tends to be writing fewer pages, targeting higher-intent queries, and accepting that your traffic number will fall. Conversion rate should rise enough to more than cover it.

Quick checklist for intent alignment

• The query implies the searcher is ready to buy (e.g. "hire a web design studio"), not just browsing (e.g. "web design ideas")
• The page above the fold mentions price, process, or proof, anchors that filter out tyre-kickers
• The CTA matches the visitor's likely next step (book a call, not "subscribe to our newsletter")

5. There's no proof anyone trusts you

The five-second test from earlier has a sibling: the trust scan. Land on your homepage. Within ten seconds, can the visitor see:

1. A real client name or logo (not a stock-photo testimonial)
2. A real project you delivered (linked, with the outcome)
3. A real face, yours, or the team's

If any of those three are missing, the page is asking strangers to take a leap of faith on a brand they've never met. Most won't.

Stock-photo testimonials are worse than no testimonials. A quote attributed to "Hannah, Marketing Director" next to a smiling woman who's clearly an iStock model breaks trust in a way that's hard to recover. Use real names, real companies, and real photos with permission, or use just names and quotes, with no photo at all. Both work. The middle ground doesn't.

The follow-on question

After the five fixes, the question we get is "so do we need a new website?"

Usually: no. The structure of the site is fine. The copy and the conversion path need surgery. A full rebuild is six weeks of work and a notable expense; a headline rewrite, a hero compression pass, and a CTA hierarchy fix is about a week and 10 to 20% of the cost.

The exception is when the underlying tech is fighting the team, a Wix site that can't be A/B tested, a WordPress build whose page builder makes every edit a 40-minute affair, a one-page site that needs to become a 30-page authority site. Those are real reasons to rebuild. "It looks dated" usually isn't.

If you want a second pair of eyes on yours, we run a free 20-minute version of this audit on calls. No deck, no slides, we screenshare your homepage and walk through the five points. Reply to this post or book a call and we'll find a slot.