Google Ads for insulation companies in the UK

January in a Bolton depot is not a branding exercise. Your estimators are staring at a diary with three ECO4 surveys, two cash-pay loft jobs, and a cavity-wall enquiry that still has not confirmed whether the property even qualifies. Meanwhile Google Ads is spending pounds on people typing “is spray foam a scam” and “ECO4 grant checker.” That mix is why so many UK insulation firms feel busy on the platform and empty on the van run.

Loft and cavity installer terms in competitive cities have been clearing £4-£8 CPC this winter - several times the cross-industry average. Google Ads still looks healthy on blended ROAS until you realise 20-30% of SEM budgets routinely leak into irrelevant queries, neglected negatives, or a homepage that talks about every product at once. For insulation companies, the leak has a local accent: grant tourists, DIY researchers, and out-of-area postcodes that will never pass PAS 2035 or TrustMark checks.

This piece is not another loft-versus-spray-foam campaign map. It is about the money decision UK marketing leads actually make: which queries you will pay for when ECO paperwork, cash loft installs, and commercial fabric upgrades all sit in the same account.

Where UK insulation Google Ads spend actually goes when you stop looking at “leads”

Pull last quarter’s search terms and you will usually find three buckets fighting for the same daily budget. First, grant hunters: “ECO4 insulation grant,” “free cavity wall insulation,” “warm home discount loft.” Second, problem researchers: “does loft insulation cause damp,” “spray foam mortgage issues,” “cavity wall insulation problems.” Third, job-ready searches: “loft insulation installers Manchester,” “external wall insulation quote Leeds,” “cavity wall survey near me.” Only the third group reliably books a survey that your PAS 2035 coordinator can stand behind.

Plenty of accounts still treat all three as conversions because the form says “Get a quote.” An ECO4 click from a tenant who cannot authorise works is not a qualified lead. A cash-pay homeowner in a 1930s semi with no loft hatch photos is closer. A facilities manager asking for 2,400 m² of factory roof insulation is a different product, margin, and sales cycle. If those three land in one Smart Bidding pool, Google will feed you whatever converts cheapest in the form, not whatever fills paying survey slots.

Mobile now takes 63%+ of paid search clicks, yet conversion rates often lag desktop by 30-40%. Insulation buyers on a phone at 21:10 want a postcode check and a call button, not a 4MB gallery of spray-foam cross-sections. If your page fails Interaction to Next Paint, they bounce before the estimator ever hears the postcode. Google’s CrUX data (Q1 2025) shows only ~56% of origins pass all three Core Web Vitals. Most insulation landing pages we audit fail INP - slow form rendering on mobile is the single biggest post-click drop. Slow pages plus grant-heavy queries is how you buy volume that never hits the diary.

AI Overviews now sit on a large slice of “how does ECO4 work” type queries and can cut CTR sharply even when you still show. You do not win those auctions by bidding harder. You win by refusing to buy them, or by sending remaining clicks to a page that qualifies eligibility in 40 seconds instead of explaining government policy.

If your team is still sending every ad to the homepage, that is the first operational break. Dedicated pages that match the ad (loft cash job, cavity survey, EWI commercial) keep bidding honest. If you want a partner to own that paid-search loop rather than another monthly dashboard, HeyLead’s Google Ads management for UK service businesses is built around booked work, not form spam.

A UK insulation PPC account structure that separates grant noise from paying surveys

You still need keywords. “Signals over keywords” is real for bidding, not an excuse to dump every insulation phrase into one Performance Max asset group and hope. Performance Max can drive bulk in mature accounts, but on a new or dirty insulation account it will happily harvest “free insulation grant UK” because those users fill Instant Forms. Keep Search as the spine for high-intent installer queries. Use PMax later, with strong audience signals from actual booked surveys, not every form submit.

Match types should follow job type, not agency habit. Exact and phrase for “loft insulation installer [city],” “cavity wall insulation quote,” “external wall insulation contractor.” Broad only where you have a thick negative list and a landing page that disqualifies grant-only traffic. High CPC pressure is pushing some teams back to manual bidding on new accounts. That can be rational for two weeks while you collect 30-50 genuine survey bookings. After that, tCPA or tROAS on a clean conversion (booked survey, not “submit”) usually beats daily micromanagement. Constant structure changes reset learning. Leave the skeleton alone and iterate ads and negatives.

Geo is not “United Kingdom.” It is the postcodes your installers can reach this week, plus a 20-30 minute radius you will actually honour in winter traffic. Exclude Northern Ireland if you do not operate there. Exclude London boroughs if your EWI crew is West Midlands only. Location bid modifiers still matter when a Sheffield loft job is profitable at £85 CPC and a Cornwall click is waste.

The grant-query trap and how UK firms keep ECO from poisoning cash-pay bidding

ECO4 and related schemes are not “bad traffic.” They are a different product. If your commercial model is funded measures through a managing agent, you should bid on those terms, with copy that asks for EPC band, tenure, and benefits status before a human rings. If 80% of your margin is private loft and cavity cash jobs, those grant terms are a tax. Twenty to thirty percent of budget waste in SEM is not mysterious. On insulation accounts it is usually “free,” “grant,” “scheme,” and “government” sitting inside a Broad match campaign named Insulation - All.

A West Midlands installer we modelled (composite, not a case study brochure) had tCPA of £41 on “loft insulation quote Birmingham” and £19 on “free cavity wall insulation.” The cheap ones were tenants forwarding a Facebook post. After they moved grant language to its own campaign with a hard eligibility form, cash-pay survey rate on the remaining Search budget moved from 4.2% to 11% in 17 days. One number, one mechanism: stop letting cheap grant forms train the bid strategy.

Call-only ads still work for emergency loft hatch and blown-fibre jobs when a homeowner has a roofer on site. They fail for EWI, where a facilities or homeowner needs photos, U-value talk, and planning notes. Do not copy a locksmith call-only template onto a 12-week external wall job.

Attribution will disagree with the van diary. Last-click will credit the branded “XYZ Insulation reviews” click that happened after a week of research. Google’s in-platform CPA will look prettier than CRM. Cross-channel attribution tools will tell a flattering story. Trust the van diary: surveys booked, show rate, jobs won, revenue. That loop is enough to stop automated bidding from hunting cheap grant forms. Treat Search as a contribution channel: booked surveys per £1,000, show rate, and jobs won, not a promised blended ROAS before anyone has seen your close rate. Nobody honest guarantees CPL before they see unit economics, crew capacity, and the landing page.

What a facilities loft job and a terrace cavity job need on the page after the click

A cash loft in a two-bed terrace wants: postcode, hatch access, day they can be in, and a number in £ they will not feel ambushed by. A commercial roof or factory lining job wants: square metres, access equipment, shutdown windows, and a PDF the FM can forward. One template cannot do both. That is why blended conversion rates look “acceptable” at 6% while estimators complain the leads are tyre-kickers.

Proof should be local and boring. Street-level before/after, not alpine chalets. Named product systems if you install them. Insurance and TrustMark in the first screen. Avoid stock “family in a warm living room” if the search was “external wall insulation quote.” Facilities people send pages to procurement. Homeowners send screenshots to a partner. Both groups bounce if the H1 says something generic like “Insulation experts.”

Forms: five fields max on mobile for residential. Commercial can be longer if you offer a callback slot. Call tracking numbers on every paid landing page. If you lose the source, you will argue with Google’s blended ROAS story using last-click from branded search. Unified measurement is the direction CFOs want. For an insulation SME, start simpler: UTM, call recording, survey booked, job won, revenue. That loop is enough to stop automated bidding from hunting cheap grant forms.

Page speed is not a vanity CrUX badge. A survey form that stutters on INP loses the 21:10 mobile click. Compress images of loft rolls. Lazy-load the gallery. Keep the click-to-call sticky. When paid traffic “does not convert,” watch the session: people often tap ECO FAQs and never reach the form. Fix the page, do not just raise budget.

What marketing leaders are seeing

Across accounts we audited in 2024, the pattern was consistent: once grant terms were split out, cash loft cost per booked survey rose ~20% but diary match rate more than doubled. Automated bidding was rarely the villain on its own - firing conversions on every quote form trained Google to buy cheap ECO eligibility clicks instead of paying surveys.

FAQs

Should UK insulation companies use Performance Max from day one?

No. Use Search until you have clean booked-survey conversions and a negative list that blocks grant and DIY language. Then add PMax with real job creative and audience signals. PMax driving “the bulk” is a mature-account story, not a dirty-account shortcut.

Is automated bidding just going to inflate spend?

It will if the conversion is a form fill. It usually will not if the conversion is a diary-confirmed survey and geos match your crews. Manual CPC is a temporary diagnostic, not a 2026 strategy for a live installer network.

What CPC should we expect for loft and cavity terms?

It varies by city and season. In peak winter (Nov-Feb) we have seen Birmingham loft installer terms clear £7-£12 CPC. Judge that against your gross margin on a typical loft day, not against a blog benchmark.

Do AI Overviews mean we should cut Google Ads?

They mean you should stop paying for encyclopaedia queries. High-intent “installer + city + quote” still clicks. Grant explainers often do not. Reallocate; do not abandon the channel.

How long before this looks stable?

Plan on 3-6 months of clean tracking, negatives, and landing tests. Anyone promising a fixed CPL before seeing your account, pages, and close rate is selling comfort, not a programme.

Putting it to work

This week

• Export search terms and mark grant vs cash vs commercial. Add the obvious negatives today.

• Change the primary conversion to booked survey or 90-second call.

• Point loft ads at a loft page, not the homepage. Check Core Web Vitals on that URL.

• Build UTMs so ECO and private pay never share one source in the CRM.

• Cap locations to postcodes your vans actually cover in January traffic.

Next 30 days

• Split campaigns if grant and cash still share a bid strategy.

• Test two RSA angles (price/speed vs TrustMark) to 50 clicks each.

• Watch form abandon on mobile and cut a field if drop-off sits on EPC questions for cash jobs.

• Only then trial tCPA or a tightly signalled PMax campaign.

Start by pulling 90 days of Google Ads search terms next to booked surveys, not form volume, and you will see whether grant language is training your bids. HeyLead can take that insulation-specific Search hygiene, landing match, and conversion setup off your plate so estimators see paying jobs instead of eligibility clicks. Chat with us on WhatsApp