What makes a trades account different from every other account
If you run agency accounts across several sectors, home services will already feel like the odd one out. The campaigns are not harder. The measurement is, and the difference comes down to four structural facts about how these businesses get work.
The first is that the phone dominates. A homeowner with water coming through a ceiling is not filling in a contact form and waiting for a reply. They tap the first number that looks local and competent. Across the trades, the share of enquiries arriving by telephone rather than by form is routinely two-thirds and can be higher in emergency work. Any measurement setup that counts forms and estimates calls is measuring a minority of the business.
The second is urgency, and how compressed the decision is. In emergency call-outs the entire journey — problem, search, click, call — can last under three minutes, and the homeowner will ring three businesses and hire whoever answers. That collapses two things at once: attribution windows that assume a considered purchase are wrong, and the client's answer rate becomes a bigger lever than anything you can do to the campaign.
The third is seasonality, and it is more violent here than almost anywhere. Heating engineers have a November. Roofers have storms. Landscapers have a spring. Air conditioning has a heatwave and eleven quiet months. Reporting month-on-month in a seasonal business produces alarming charts that mean nothing, and produces them exactly when the client is most anxious about spend.
The fourth is job value spread. A boiler service and a full system replacement arrive through the same campaign, look identical as leads, and differ by a factor of thirty in revenue. Cost per lead is therefore a much weaker signal here than in sectors with uniform transaction sizes, and an agency optimising towards cheap leads can systematically fill a client's diary with low-value work while every number on the report improves.
The measurement gaps that cost agencies these accounts
Each of these is common, each is invisible in standard reporting, and each has ended a retainer that did not deserve to end.
The uncounted two-thirds
Reporting form fills to a business whose enquiries are mostly calls. Every derived number — cost per lead, channel ranking, return — is wrong by the same large factor, consistently, in the direction that makes you look worse.
The unanswered evening
Calls arriving after five going to an answerphone nobody checks. The client blames lead quality; the recordings show nobody picked up.
The seasonal panic
A heating client's February compared to their January. The chart collapses, the client panics, and the agency spends a meeting explaining weather.
The service calls in the lead count
Existing customers ringing the tracked number about a warranty, counted as new leads. Volume looks strong, conversion looks dreadful.
The high-value channel that looked expensive
A channel with a poor cost per lead producing the replacement jobs rather than the call-outs. Cut on unit cost, missed on revenue.
The van, the neighbour and the sign
A genuinely large share of trades work arrives through word of mouth and local presence. Reporting tracked leads as if they were all the leads invites a comparison you will lose.
How to set up a trades account so it survives its first winter
This is the sequence that works. It is ordered by how much each step changes what you can defend, not by how easy it is.
Track the phone before anything else
Dynamic numbers on the site so every call carries its source. Until this exists, every other number in the account is an estimate over the minority of leads you could see. It is a fifteen-minute job and it changes the entire picture.
Measure the answer rate immediately
In the first fortnight, find out how many calls go unanswered and when. On most trades accounts this is the largest single recoverable number, and it is not the agency's fault — which makes it the most valuable thing you can bring to the client early.
Route the gaps
Out-of-hours to a duty mobile or an answering service, lunchtimes to a second desk, paid traffic to whoever converts best. One rule at a time, each justified by the missed-call data.
Separate service from sales
Classify existing-customer calls out of the lead count. It lowers the volume you report in month two, which is uncomfortable, and it makes every ratio afterwards honest.
Get job values in, however roughly
Even a crude split between call-out, repair and replacement transforms channel comparison in a business with this much value spread. Push leads into whatever the client already uses so somebody actually marks them.
Set the report to year-on-year
Before the first seasonal drop, not after. A client who has been shown the seasonal shape in advance does not panic in February.
What each trade needs first
The pattern differs enough between trades to be worth planning around rather than applying one template to all of them.
| Trade | Journey shape | Where the money leaks | First thing to fix |
|---|---|---|---|
| HVAC / heating | Split: emergency breakdowns and considered replacements | Replacement enquiries judged on the same cost per lead as call-outs | Job value on won leads, so the two streams stop being averaged together. |
| Plumbing / drainage | Mostly emergency, minutes from search to call | Unanswered and after-hours calls | Missed-call reporting, then an out-of-hours routing rule. |
| Roofing | Considered, weeks long, storm-driven spikes | Attribution windows too short to hold the journey | Multi-touch with a 90-day window; first-touch credit for the content that introduced them. |
| Electrical | Mixed emergency and planned works, strong repeat base | Existing customers counted as new leads | Call classification, so repeat business stops inflating the lead count. |
| Landscaping / driveways | Highly seasonal, long consideration, high value | Month-on-month reporting during the off-season | Year-on-year comparison agreed with the client before winter. |
| Pest / locksmith / glazing | Emergency, extremely compressed, price-sensitive | Slow response and price questions handled badly on the phone | Call transcripts — the handling problem is usually visible in ten calls. |
The seasonality conversation, had properly
Every agency with trades clients has lost an account in a quiet season, and most of those losses were preventable. The mechanism is always the same: the client sees a falling chart, concludes the marketing has stopped working, and cancels at exactly the point when cancelling does the most damage.
It is worth understanding why cancelling in the off-season is so costly, because this is the argument that keeps the retainer. Organic visibility decays slowly and recovers slowly; a client who stops SEO in November and restarts in March does not resume where they left off, they resume several months behind. Paid campaigns lose their learning and their historical data. And competitors who kept spending through the quiet months have accumulated reviews, rankings and remarketing audiences that arrive in spring as a head start.
The way to win that argument is to have made it before the season turns. Show the client their own seasonal curve from last year in September. Agree that the winter target is not volume but position — reviews, rankings, content, and the groundwork that pays in April. Set the reporting to compare against the same month last year so the chart reflects progress rather than weather. All of that is a fifteen-minute conversation in autumn that replaces a difficult one in February.
There is a stronger version available where the data supports it. If last winter's spend produced this spring's ranking positions, and this spring's positions produced a measurable number of calls, that chain is visible in the reporting and is the single most persuasive thing you can put in front of a trades client. It turns the off-season from a cost into an investment with an observed return, which is a completely different budget conversation.
What a trades client actually wants to see
Trades owners are, as a rule, direct, numerate about their own business, and impatient with marketing vocabulary. Report accordingly.
- How many people rang. The first number, in the largest type. Not sessions, not impressions, not average position. Calls and forms, this month against the same month last year.
- How many we missed. Uncomfortable and enormously valuable. It is the number that changes the client's behaviour rather than yours.
- What each enquiry cost. In pounds, against what they know a job is worth. Trades owners do this arithmetic instantly and respect being given the inputs.
- Where they came from, in plain words. "Google searches", "the map listing", "Facebook" — not "organic / cpc / referral".
- The actual calls. Being able to listen to three enquiries from last week is what makes the totals real rather than something the agency asserts.
- What the reviews are doing. Rating, volume and whether anyone replied. In the trades this moves the map pack more reliably than anything else available.
The lead quality argument, and how to win it with evidence
Around month four of a working trades campaign, the client will say the leads are not converting. It happens on good accounts as often as bad ones, and the reason it is so hard to answer is that both sides are arguing from things the other cannot see. The client sees their engineers frustrated. You see rising volume at a falling cost.
With recordings and transcripts the argument becomes an audit, and on a trades account the audit almost always finds one of four things. A large share of calls never answered, usually clustered at lunchtime and after five. Callers asking what something costs, being told the business does not quote over the phone, and hanging up. Voicemails left and never returned. Or a genuine targeting problem, where a keyword is bringing in work the client does not do — commercial enquiries to a domestic business, or a service they stopped offering two years ago.
Three of those four are the client's to fix and one is yours. That split is why the evidence matters so much: without it, all four look identical from the client's side and are attributed to the agency by default. With it, you arrive at the meeting with eleven specific calls, a rough value attached, and a recommendation — which turns an accusation into a joint problem.
The one worth being scrupulous about is the fourth. If the transcripts show you are genuinely sending unqualified traffic, say so first and fix it. Agencies that find their own targeting problem in week six keep the account; the ones who discover it when the client produces evidence do not. Running the audit honestly means being willing to lose the argument occasionally, and that willingness is most of why the evidence is persuasive when you do win it.
Which plan a trades-focused agency actually needs
The honest answer depends on client count and call volume rather than on which features sound impressive, and trades accounts are unusually call-heavy — which is the number to plan around.
A consultant or small agency with three or fewer trades clients fits Starter. It includes call recording, tracked numbers, form capture and scheduled reports, which covers the measurement problem entirely. What it does not include is white labelling or a custom domain, so clients see the platform rather than your brand — acceptable when you are small, and the thing you will outgrow first.
Most trades agencies belong on Growth. It carries ten clients, twenty-five tracked numbers and three thousand calls a month, plus the two things that matter commercially: white labelling on your own domain, and CRM delivery so leads reach the client's job management software. Transcription is included, which on trades accounts is where the missed-call and call-handling findings come from.
Agency is the tier for an established shop with a real trades book — fifty clients, a hundred and fifty numbers, twenty thousand calls a month, and AI call intelligence for automatic classification across that volume. The classification matters more here than in most sectors, because separating service calls from new enquiries by hand across thousands of calls is not something anybody sustains.
The call allowances are the constraint worth checking before you choose. A single busy heating client in November can produce several hundred tracked calls on their own, and a seasonal peak is precisely when you least want to be thinking about limits. Size for the busy month rather than the average one.
Bundles that suit a trades book
Three shapes that come up repeatedly. Commit for longer and the discount rises: 15% at three months, 25% at six, 30% annually — which matters on a seasonal book, because annual billing smooths a cost that would otherwise land hardest in your quietest quarter.
| If you are | Start on | Because | Add when |
|---|---|---|---|
| A consultant with two or three trades clients | Starter | Tracked numbers, recording and reports cover the whole measurement gap. Three clients, five numbers, 500 calls a month. | You win a fourth client, or a client asks why the dashboard is not branded. |
| A specialist agency with a trades book | Growth | Ten clients, twenty-five numbers, 3,000 calls, white label on your domain, CRM delivery into their job software, and transcription for the call-handling findings. | Call volume approaches the ceiling in peak season, or you pass ten clients. |
| An established agency running trades at scale | Agency | Fifty clients, 150 numbers, 20,000 calls, and AI classification so service calls separate from new enquiries without anyone listening to them. | You start reselling to other agencies — sub-agencies live at this tier. |
Where to start on Monday
Pick the trades client you are least confident about — the one where you suspect the reporting understates what you are doing, or the one who has gone quiet. Put dynamic numbers on their site and leave it a fortnight.
Then look at three things. How many calls arrived, against the lead number you have been reporting. How many went unanswered, and at what hours. And which channel produced the calls, which is frequently not the channel producing the forms.
On a typical trades account those three figures reframe the relationship. The lead count roughly triples, which changes cost per lead by the same factor. The missed calls give you something genuinely valuable to bring the client that is not about your own performance. And the channel split usually credits organic search and the map listing far more than the previous reporting did — which is normally the work you have been doing and struggling to evidence.
That is a fortnight of elapsed time and about half an hour of work, and it is the cheapest way to find out whether this is worth doing across the rest of the book.
Common questions
Why does call tracking matter more for trades than other sectors?
Because most trades enquiries are phone calls rather than forms — routinely two-thirds, and higher in emergency work. Reporting form conversions to a plumbing or heating client measures a minority of their business, and every derived figure inherits the same large error in the direction that makes your work look worse than it is.
How do I handle a client's off-season without losing the account?
Have the conversation before the season turns. Show them last year's curve in advance, agree that the winter goal is position rather than volume, and set reporting to compare year-on-year so the chart shows progress instead of weather. A fifteen-minute conversation in autumn replaces a difficult one in February.
What about jobs that come from the van or word of mouth?
They are outside the tracked dataset and always will be, and for many trades they are a large share of the work. Say so in the reporting rather than letting the tracked number imply it is the whole picture — a client who compares your figure to their own diary will otherwise conclude your tracking is broken.
Can I stop existing customers being counted as new leads?
Yes — call classification separates repeat callers and service enquiries from new business. It lowers your reported volume the month you switch it on, which is uncomfortable, and it makes every ratio afterwards honest. On electrical and heating accounts with a strong repeat base this is a substantial correction.
Which attribution model suits emergency trades?
Time decay, because the whole journey often lasts minutes and the touches closest to the call genuinely did the work. For considered trades like roofing or landscaping, use position-based with a 90-day window — those journeys run for weeks and a short window erases the introduction entirely.
How do I show that answering the phone matters without blaming the client?
Frame it as recoverable revenue rather than as a failing. "You received 41 calls after six last month and answered four" plus their own average job value produces a number that usually exceeds the retainer. That reframes you as the only party who can see where the money is going.
Do I need transcription for trades accounts?
It is where the most actionable findings come from — price questions handled badly, voicemails never returned, a keyword whose calls are all wrong-number. You do not need it on every account permanently; turning it on for a month to answer a specific question is a perfectly reasonable pattern and usage is metered per client.
What plan should a trades-focused agency start on?
Growth suits most: ten clients, twenty-five numbers, 3,000 calls a month, white labelling on your own domain, CRM delivery and transcription. Starter works for two or three clients but has no white labelling. Size for your busiest month rather than your average one — a single heating client in November can produce hundreds of calls alone.
Can leads go straight into the client's job management software?
Yes, through native CRM integrations or a signed webhook for the trade-specific systems most of these businesses use. Speed matters more here than almost anywhere, since an emergency caller rings three businesses and hires whoever responds first, so delivery happens on arrival rather than on a batch.
How quickly will I see whether this changes anything?
A fortnight. Put dynamic numbers on one client's site and then compare the call count to the lead figure you have been reporting, look at unanswered calls by hour, and check which channel produced the calls. On a typical trades account the lead total roughly triples and the channel ranking changes.