A locksmith I work with phoned me last month. He'd been paying an SEO agency £800 a month for nearly two years and his rankings hadn't moved. He sent me his latest report. Forty-three pages. Generic checklist. Backlinks tracked, "domain authority" tracked, two dozen keywords tracked. Not one of those keywords was an emergency-intent query. Not one mention of his Master Locksmiths Association approval. Map pack performance? Not in the report.
This is the standard pattern. Most local SEO advice and most local SEO retainers were built for plumbers, dentists and restaurants. They get applied to trades wholesale, and the result is a lot of activity, a lot of reporting, and very few extra calls.
Three things the standard advice misses.
1. Keyword volume is the wrong metric for trades
Generic SEO tools are built around search volume. A keyword with 3,000 monthly searches looks more valuable than one with 30 because the tool is sorted by volume.
For trades, that's almost backwards.
The high-volume keywords in trade verticals are usually informational: "how to change a Yale lock", "how much does a boiler service cost", "what is Part P certification". The people typing those queries are researching, not buying. They might convert eventually, but they'll convert through someone they already trust by then — not through whoever ranked third on a how-to page.
The keywords that pay are the low-volume, urgent, geographic ones. "Emergency locksmith Sheffield S2", "boiler repair Manchester tonight", "electrician near me NICEIC". These have search volumes in the dozens, not the thousands. Generic SEO dashboards either don't show them at all or show them as "low priority" because the volume is small.
A keyword with 30 monthly searches at 80% buying intent will outearn a keyword with 3,000 searches at 10% buying intent every single time. We've seen this play out so consistently across trade engagements that we no longer even track high-volume informational queries on monthly reports. They're not the work.
2. Mass-producing location pages is killing trade sites
The other standard play is to build out hundreds of "location pages". One per postcode, one per nearby town, one per neighbourhood. The thinking: cover every variation a customer might search for, and you'll catch all the long-tail traffic.
What happens in practice: an agency builds 100–200 near-identical location pages, each with the trade name and the place name swapped in. Six months later the site loses 30–50% of its organic traffic. Sometimes the site gets manually penalised. More often it just quietly tanks under Google's helpful-content updates, which specifically target this pattern.
I've watched this happen twice this year on sites we audited but didn't take on. Both were locksmiths whose previous agency had built out 80+ pages for postcodes the business technically served but didn't operate from. The pages all read the same. Google figured it out.
The right approach is narrower. Top 5–10 postcodes by actual call volume get their own page with real local content — the landmarks in that area, the common lock or boiler types in that housing stock, the typical response time. The other 30 or 40 postcodes the business will travel to in an emergency get listed as a service area in the Google Business Profile. They don't get standalone pages.
This is a quality-over-quantity argument that everyone agrees with in principle and nobody applies in practice. But the maths is simple: ten genuinely useful pages outrank a hundred copy-paste ones, and they don't trigger the penalty.
3. The map pack matters more than the website
For most trade searches, Google now shows the map pack first — the three-result block with photos, ratings and call buttons. Then organic results. For "near me" queries and emergency queries, the map pack is often the only result anyone clicks. The blue links underneath might as well not exist.
Standard local SEO advice still treats the website as the centre of the work. The website is where the time goes: page speed, on-page optimisation, blog posts, internal linking. The Google Business Profile gets a tidy-up at the start of the engagement and then quietly stops being managed.
For trades, the priority order should be reversed. The Google Business Profile is the front door. The website is the back-office.
That doesn't mean the website is unimportant — service pages still need to exist, schema still needs to be right, page speed still matters. But the weekly attention should sit on the GBP: posts every week, photos every month, Q&A managed actively, fraudulent edits monitored and challenged within days of appearing, service categories tuned to match emergency-intent queries, hours set genuinely (with real out-of-hours coverage backing the claim).
A locksmith with a pristine website and a half-finished GBP loses to a locksmith with a basic website and a fully-optimised GBP. Every time.
What we measure instead
Five things, on a single page, every month:
- Map pack appearances by query. Not just "rank position" — rank position tied to specific searches, specifically the emergency-intent and "near me" ones. A plain rank tracker doesn't capture this; you have to track at the query-plus-location level.
- Out-of-hours call volume separately from daytime. For trades with significant emergency demand, the high-margin work is at unusual hours. If your monthly report shows "27 calls" but doesn't tell you how many came at 2am on a Sunday, you don't know whether the engagement is working for the part of the business that pays best.
- NAP consistency across the directories that matter for the trade. Not 200 generic directories — the 25–30 trade-specific ones (Master Locksmiths Association, Gas Safe register, NICEIC, Checkatrade, Trustatrader, FreeIndex, the trade-body site) plus the half-dozen that always matter (Google, Bing, Yell, Apple Maps). NAP drift in any of those drops your local pack ranking, and it's the boring monthly work most agencies skip.
- Trust-signal audit. MLA, Gas Safe, Part P, NICEIC, insurance approval, Trustpilot — present, displayed prominently, schema-marked, kept current. Audited monthly. Most trade sites have at least one of these displayed wrong.
- Conversion rate on click-to-call from the GBP and the website. Calls are the unit of work for most trades, not form submissions. If you're tracking form submissions but not call conversion rate by source, you're tracking the wrong thing.
What's not on that list: backlinks, "domain authority", page speed scores, blog post counts, keyword counts. These aren't useless — they matter at the margins. But they don't decide whether the phone rings.
Why most agencies measure the wrong things
Because the metrics on the standard reports are easier to track. Generic SEO platforms produce them automatically. Map-pack-by-query tracking and trust-signal audits and per-source call conversion are more work to set up. They require knowing the trade. They require manual checks.
So most agencies measure what's easy to measure and report what's easy to report. The deck looks impressive. The numbers go up most months. The trade owner reads it once, doesn't really understand half of it, and pays the invoice because the agency seems to be doing things.
The actual work isn't on the deck.
What this means if you're paying for SEO right now
If you're a UK tradesman paying an SEO retainer, look at your last three monthly reports. Three questions:
- Does the report break out emergency-intent and "near me" rankings separately, or are all keywords lumped together?
- Does it show out-of-hours call volume?
- Does it audit your trade-specific trust signals (MLA / Gas Safe / NICEIC / Part P / insurance approval) every month?
If the answer to all three is no, your agency is reporting on activities, not on outcomes. That doesn't necessarily mean they're bad operators — it usually means they're running a generic local SEO playbook on a trade business it wasn't designed for.
You might be paying for movement. You might be paying for theatre. The reports above don't let you tell the difference.