Six years ago, Checkatrade was a decent deal for most builders. You paid a monthly fee, the odd £15 per-lead charge on top, and got three or four enquiries a week for loft conversions. A fair chunk of them turned into quotes. Some of those turned into jobs.

We don't see that anymore. Across 30-odd builders we've either audited or run pipelines for, the Checkatrade story has quietly gone the same direction: more money going in, fewer quotable enquiries coming out, and a growing share of tyre-kickers who've already sent the same job to four other firms on the same platform.

This post is what we actually see in the data, why it's happening, and where the better enquiries are coming from instead. No rant, no "Checkatrade is dead" clickbait. Just the numbers.

Here's an anonymised composite of a Surrey extension builder we audited in January. Turnover around £1.4m. Checkatrade spend had crept from £2,800 a year in 2020 to £9,600 a year by 2025.

• 2020: 184 enquiries · 41 quoted · 11 won · avg job £22k
• 2022: 201 enquiries · 34 quoted · 9 won · avg job £23k
• 2024: 232 enquiries · 22 quoted · 5 won · avg job £19k

Lead volume went up 26%. Quoted-to-won rate on the platform dropped from 27% to 23%. Quote rate from enquiry collapsed from 22% to under 10%.

Translated: he paid roughly £1,920 per won job in 2024, on work with a shrinking average ticket. His own referrals and direct website enquiries in the same period converted to quotes at 61% and to won jobs at 31%. Same firm, same quoting process, same pricing. The only variable was the source.

That pattern is consistent across the builders we see. It's not a one-off. So what's changed?

One. The platforms have made the signup funnel shallower on purpose. Homeowners can now post a job in under 30 seconds from a phone, with almost no qualifying questions. Easier to post = more posts = more revenue for the platform. Also: more half-serious enquiries, more people "just getting a rough idea," and more multi-platform shoppers who've already posted the same job on MyBuilder and Rated People.

Two. The per-lead fee has climbed while the cap on how many firms see each lead hasn't moved. You're now typically one of 3–5 firms racing to respond, on a lead that costs you £30–£60 before you've spoken to anyone. The race rewards whoever answers fastest, not whoever does the best work. That's fine for a drive cleaning firm. Not fine for a £40k kitchen extension.

Three. Homeowners who are serious about £20k+ projects have learned to Google first. When we interview recent customers of our clients, roughly 70% of the won extension and loft jobs involved the homeowner looking at the builder's actual website — often reading a case study or looking at photos — before any marketplace. The marketplaces have become a fallback search, not the primary one, for the money jobs.

The math changes. At £1,920 cost per won £19k job you're burning just over 10% of the project value on lead gen — and the project margin was already thin because competing bidders on the same platform pushed the price down in the quoting phase.

Compare that to a well-set-up Google Business Profile, a proper local SEO page for "loft conversion + [town]", and even modest ad spend on brand-defensive search terms. We see cost-per-won-job on that stack landing at £250–£600 across the builders we run pipelines for, on jobs with an average ticket 30–40% higher than what the marketplace produces. Because those enquiries land directly with the builder, not with five competitors at once.

None of this is clever. It's just unsexy and unfinished on most builders' sites. Doing these four, in this order, has moved the needle for every client we've taken on in the last 18 months.

1. Get your Google Business Profile properly filled out

Weekly, not once-a-year. Post photos from the last job. Answer Q&As. Reply to every review within 48 hours. Add service categories. Add service-area towns. Half the builders we audit have their own name misspelled, hours wrong, and two different phone numbers across their profile and their website. The platform rewards accuracy and recency because Google is trying to avoid embarrassing itself.

2. Publish one real case study a month on your own site

Proper page. Named client (with permission). Before/after photos. Actual numbers — start date, end date, final cost band, any problems and how they were solved. This is the single thing homeowners print out and show their partner before they phone you. It's also what Google indexes and ranks for "[your service] near me" queries.

3. Run defensive search ads on your own name

If someone types "[your firm] + reviews" into Google, an ad from Checkatrade, Rated People, or Houzz will show up above your own listing and skim off the call. £50–£100 a month on a brand campaign stops that. This is cheap defence and obvious once you've looked at the search results for your own firm.

4. Build a small local SEO surface — 5–10 pages, not 500

One page per service per town you actually work in. Not 200 programmatic junk pages — 5–10 pages that genuinely describe the work, link to a relevant case study, and list the postcodes you cover. These compound quietly and bring in the right kind of enquiry, because the homeowner has already read about the work before they phone.

Don't cancel Checkatrade on Monday morning. Some firms still get decent volume from it, especially outside the home counties. Instead, cap it. Pick a monthly budget — often £400–£600 is plenty — and treat it as one enquiry source among several. Measure it honestly. If cost-per-won-job is creeping up year over year (it will be), shift that money towards the four things above.

Our take, from running this playbook with about 30 builders: within 90 days you'll see more enquiries that named you specifically before they phoned. Within six months you'll stop caring about your Checkatrade ranking. Within twelve you'll be wondering why you ever paid per-lead in the first place.