Most advertising guidance is written for businesses whose customers browse. Yours do not. Nobody scrolls Instagram and decides their boiler should break. They discover the problem, reach for their phone, and type — and whoever is at the top of that search gets the job, frequently without a second quote. That changes almost everything about how you should advertise and which tool is worth paying for.
Search first, and it is not close
For emergency and repair work the intent gap between search and social is enormous. A social ad interrupts someone who was doing something else. A search ad answers someone who has already decided to spend money and is choosing who to give it to.
Emergency and repair
Almost entirely search. Boiler failure, lockouts, leaks, breakdowns. Social is nearly worthless here because the need is unpredictable and immediate.
Planned work
Kitchens, extensions, landscaping, rewires. Search still leads, but social earns a place — people research these for weeks and can be reached before they search.
Contracts and commercial
Search for the enquiry, plus a credible website. Social rarely reaches a facilities manager in work mode.
So a tool that only does Meta is the wrong purchase for most trades, however good its creative is. If you buy one thing, buy the one that does Google properly.
The keywords that will waste your money
This is where trade campaigns leak, and where an AI tool either earns its fee or costs you. Build these negatives before spending anything:
- 1Job seekers. "plumber jobs", "electrician apprenticeship", "vacancies", "careers", "training", "course". A large volume of search around every trade is people wanting to become one.
- 2DIY intent. "how to", "DIY", "yourself", "fix my own". They are not hiring you.
- 3Salary and cost research. "plumber salary", "how much do electricians earn". Pure waste.
- 4Wholesale and parts. Someone searching for a boiler part wants a merchant, not an installer.
- 5Out-of-area. Radius targeting helps but keyword-level leakage is real — a Leeds firm can pay for "London" terms.
💡 Ask any vendor whether the tool builds a negative list at launch and keeps extending it from the search terms report. If it only generates keywords, it has done the easy half and left you the expensive half. For a trade, the negatives are worth more than the keywords.
The keyword-volume test
Ask whether keywords come from live Google Keyword Planner data through the official API, or whether a language model was asked which terms are popular. The second is common — Planner access requires an approved developer token, guessing does not — and it produces lists that read well and contain terms nobody searches.
Check it yourself: take five suggested keywords into Keyword Planner with the UK selected. If several show no volume, the campaign built on that list will get almost no impressions and nothing will explain why.
Radius, not region
You will not drive ninety minutes for a callout, so do not pay to be seen ninety minutes away. Radius targeting around your actual working area, with the radius set to where you genuinely go rather than where you would like to.
One detail that matters: set targeting to people *in* your area rather than people *interested in* it, which is the default on some platforms and will show your ad to someone in another country researching your town.
The phone is your conversion
Most trade enquiries are calls, not form fills. If your conversion tracking only counts form submissions, you are measuring the minority and optimising toward it.
- 1Set up call conversion tracking so a call from the ad counts. Without it the platform has no idea which keywords produce work.
- 2Call extensions so the number is tappable directly from the search result — many people will never reach your site.
- 3If you cannot answer during the day, a form or a message route matters more, not less. An unanswered call is a lost job and you paid for it.
💡 A campaign optimising toward a conversion that never fires is optimising toward nothing. For a trade this usually means calls: if the call is not counted, Google has no idea which keywords produce work and will optimise toward whichever ones produce form fills instead.
Handling the lead when you are up a ladder
The structural problem of every trade business: the person best placed to sell is the person currently unable to answer. Response time decides conversion — a lead contacted within minutes converts far better than one contacted an hour later, and by the evening most are gone.
What genuinely helps is an alert that reaches you rather than sitting in an inbox you open at nine at night. A WhatsApp notification the moment an enquiry arrives, with the name, number and what they asked about, lets you decide in ten seconds whether to stop what you are doing. Lead scoring — rating each enquiry on stated budget, timeline and how they arrived — matters more for trades than for most sectors, because the difference between a boiler replacement and a dripping tap is the difference between a good week and a wasted call.
Compliance, briefly
AI copy will write "the best plumber in Leeds" and "guaranteed same-day" without hesitating. Both are claims the ASA will hold you to. Replace superlatives with something checkable — "Gas Safe registered, same-day callout across LS postcodes" is both compliant and better advertising, because it is specific.
A sensible setup
Google Search first
Exact and phrase match, tight radius, negatives built before launch, call tracking on.
One properly funded campaign
Above the platform floor. Not three underfunded ones across three services.
Alerts to your phone
Not an inbox. The enquiry is worth what you can do in the first ten minutes.
Social later
Once search is producing, add Meta for planned work where people research before searching.
The enquiry reaches your phone, not an inbox
Adyft builds Google Search campaigns with negatives and call tracking handled, and alerts you on WhatsApp the moment an enquiry arrives. Fourteen-day free trial.
Start free — no card needed