Paid Search for Builders
You’re not buying a job. You’re buying a conversation that starts eight months early.
Your job values change what’s affordable
An extension is £19,450. A loft conversion, £17,200. A garden room, £12,850.
At those numbers you can afford clicks that would bankrupt a gas engineer. If a booked extension costs you £400 in advertising, that’s around 2% of the job. Very few marketing channels in any industry look that good.
The catch is the timeline. Your customer isn’t buying today. They’re eight months out, researching, talking to their partner, getting three quotes. So a click today produces an enquiry in October, and if you judge the campaign on this month’s bookings you’ll switch it off long before it’s had a chance to work.
That mismatch is why paid search has a bad reputation among builders. It’s not that it doesn’t work — it’s that it gets measured on the wrong timescale and killed early.
Handled properly, it does something quite specific: it puts you into the conversation early, then keeps you there through a long decision, which is exactly where most builders lose people they’d already half-won.
What actually works
Bid on the job, not the trade. “House extension cost”, “loft conversion [town]”, “garage conversion” — specific, high-intent, and cheaper than “builders near me”.
Landing pages per job type, with real cost information. Someone clicking “extension cost” wants a number. Give them a range or they’ll leave.
Remarketing is the important part. Someone who read your extension page in February needs to see you again in April, June and August. Over an eight-month decision that’s what wins the job, and almost no builder does it.
Lead magnets work unusually well here. A cost guide or a planning checklist in exchange for an email means you can stay in touch across the whole decision rather than hoping they remember you.
Negatives for jobs, courses, materials and DIY. Building searches attract huge amounts of trade traffic.
Skip “builders near me”. Contested, directory-heavy, and it brings small work.
What we’d actually do
Campaigns by job type — extensions, lofts, garages, garden rooms — each with its own landing page carrying real cost ranges.
A cost guide as a lead magnet, so enquiries that aren’t ready yet still enter your list rather than disappearing.
Remarketing across a long window, because that’s where the actual return is.
Then measurement over six to nine months rather than monthly, with early indicators — guide downloads, page depth, returning visitors — to tell us it’s working before the bookings arrive.
And follow-up that survives the timeline, because there’s no point generating an enquiry in February if nobody contacts them again before October.
The first 90 days
The honest framing for your trade is that 90 days is not long enough to judge this, so here’s what the first 90 days should produce instead of bookings.
Weeks one to three. Landing pages built per job type — extensions, lofts, garages, garden rooms — each carrying real cost ranges. Plus the cost guide that acts as your lead magnet. This is most of the work.
Week three onwards. Campaigns live on job-specific terms. Not “builders”. Modest budget, tight geography, heavy negatives against trade and career traffic.
Month two. Remarketing switched on, which is genuinely the half that produces the return over an eight-month decision. Guide downloads should be arriving by now — that’s your early signal that the right people are landing.
Month three. Expect enquiries, not many bookings. A handful of people who downloaded the guide in month one may now be talking to you. Most of your February traffic will book in the autumn.
The number we’d hold you to at 90 days is guide downloads and enquiry quality, not jobs won. If you judge this on three months of bookings you’ll switch it off right before it starts working — which is exactly why paid search has a poor reputation among builders, and it’s a measurement problem rather than a channel problem.
What good looks like
Cost per booked job under £600. At £12,850 that’s under 5% and comfortably worth it.
Guide downloads within the first month. Your early signal that the campaign is reaching the right people.
Enquiry-to-booking measured over six months, not four weeks.
Remarketing running continuously, because that’s the half that does the work.
The mistakes we see most
Judging it monthly. The commonest and most expensive error in your trade.
Bidding on “builders”. Expensive, contested, small jobs.
No cost information on landing pages. People searching cost want cost.
No remarketing. You pay to reach them once and then let them forget you.
No follow-up system. An eight-month decision needs more than one phone call.
No lead magnet. Most of your traffic isn’t ready to enquire yet, and without one they’re gone.
What’s it worth?
Average job £12,850 · cost per booked job £520
Every £520 returns £12,850. The only thing that breaks it is measuring too early.
Every enquiry lands in Gaffer
Enquiries arrive tagged with the job type and page. Follow-up sequences keep you in front of them across months, which is where most builders lose people. Quotes go out from templates, and long-running opportunities don’t quietly fall off the list.
Questions builders ask
How long before I know it’s working?
Six months for bookings. Six weeks for the early signals.
Isn’t it expensive per click?
Yes, and it doesn’t matter at your job values. Cost per booked job is the number.
What budget?
Start modest and let the cost per job decide. We’d rather prove it small and scale something that works than spend big on something unproven.
Can I pause it when I’m booked up?
Yes, but be careful — pausing a builder’s campaign stops the pipeline eight months out, not today. Better to keep it ticking at a lower budget than switch it off entirely.
Ad spend separate?
Always, in your own account.
Free review · Bigger outfit? Talk to us
See what we'd do for builders
Put your website in and we'll tell you what's losing you work — speed, mobile, your Google profile, reviews and the pages you're missing. Two minutes, no call unless you want one.