In SoudCoh's research sample of Australian trade and home-service advertisers, a median 14.7% of ad clicks became a recorded call or form (middle half 12%–19.6%): 12.3% for plumbers, 14% for electricians and 20.1% for cleaning businesses. Compare yours only if you count the same way — real calls and forms, not taps on a number. The method is on our Google Ads benchmarks page.
That is one of at least three conversion rates, and anyone who quotes a single number without asking what you count is guessing. A trades business has at least three conversion rates — visitor to enquiry, enquiry to quote, quote to booked job — and the only one that pays wages is the last. Measure all three separately.
That answer is less satisfying than a percentage, and it is the reason benchmark figures circulate so freely: a number is easier to repeat than a definition.
What follows is how to split the three apart, why published benchmarks will not transfer to your business, and how to build a benchmark of your own in an afternoon.
One number hides the problem
When an owner says the conversion rate is three per cent, the useful question is: three per cent of what, becoming what? Those two blanks are where the whole answer lives, and they are almost never filled in.
A conversion rate without a definition is not a measurement. It is a mood.
A trades business has at least three distinct rates stacked on top of each other, and they fail for completely different reasons. Rolled into one figure, a collapse in the third can be masked by a rise in the first, and the business spends money fixing the part that was already working.
Split them apart and each one points at a specific person, a specific process and a specific fix.
The three rates, and what moves each one
| Rate | What it measures | What moves it | Who owns it |
|---|---|---|---|
| Visitor to enquiry | People who land and then contact you | Page speed, whether the phone number is tappable, how much the form asks, whether the page answers the question that brought them | Whoever owns the website |
| Enquiry to quote | Contacts who receive a real price | How fast you answer, whether calls are missed, whether quotes go out the same day | Whoever answers the phone |
| Quote to booked job | Quotes that become paid work | Price, availability, trust, whether anyone follows up | Whoever quotes |
Notice that only the first is a marketing problem. The second is an operations problem and the third is a sales problem, and in most trades businesses the largest losses sit in the second and third. That is inconvenient, because the first is the one an agency can be paid to fix.
Multiplied together, the three give you the number that matters: the share of visits that become money. That is also the number that lets you work out what a visit is worth, which is the input every budget decision needs and almost nobody has.
Why published benchmarks will not transfer to you
Benchmark tables are compiled from businesses that count differently, which makes the averages arithmetically valid and practically meaningless.
Four reasons two identical businesses report different rates
- Calls. A business counting phone enquiries and a business counting only form submissions are not measuring the same thing. In trades, the gap between them is large.
- Denominator. All visits, or only visits from people in your service area? Job seekers, suppliers and competitors are in the first number and not the second.
- Counting rule. Someone who fills a form and then rings can be one conversion or two, depending on a setting nobody remembers changing.
- Urgency mix. A blocked drain at eleven at night converts at a completely different rate from a bathroom renovation, and both sit in the same account.
None of that means measurement is hopeless. It means the only benchmark worth having is your own, measured consistently, compared against yourself. And it means the definitions have to be written down once and left alone, which is a measurement discipline rather than a reporting one. It is also why SoudCoh's click-to-enquiry figures are published as a band across the research sample rather than one average rate — a single figure would hide exactly this spread.
A conversion rate without a definition is not a measurement. It is a mood.
Build your own benchmark in an afternoon
You need one quarter of history and three counts. Most businesses can assemble this from a phone bill, an inbox and an invoicing system without buying anything.
- Count every enquiry that arrived in the quarter — calls, forms, messages, walk-ups. One list.
- Count how many of those received a real price.
- Count how many of those became paid work.
- Divide each stage by the one before it. Those are your three rates.
- Write down what you counted, in a sentence, and keep it. The definition is the asset.
Do it once and you will usually find one stage that is far worse than the other two. That stage is the whole project. It is also, reliably, cheaper to fix than buying more traffic to pour into the same leak — which is the argument in favour of fixing what happens after the enquiry lands before increasing what you spend to generate more.
The rate nobody measures is usually the cheapest to fix
The three rates above all start from someone who got through. There is a fourth number sitting underneath them that most trades businesses never look at: how many people tried to reach you and did not.
The rate that costs the least to improve is almost always the one nobody is measuring: how many people never reached a human at all.
Missed calls during a job. Calls that rang out after hours with no message. Enquiries sitting in an inbox nobody opened until the next day. A profile that logged calls and directions from people who never touched the website at all. Each one is an enquiry you already paid to generate, lost before any conversion rate had a chance to apply.
Pull the call log for one week and count the unanswered inbound calls, then count how many of those numbers ever rang back. That number is usually the most expensive thing in the business and it costs nothing to look at. It also explains why two plumbing businesses in the same area, spending the same money, end the quarter in different worlds.
The rate that costs the least to improve is almost always the one nobody is measuring: how many people never reached a human at all.
What to do when a rate drops
Identify which of the three moved before changing anything. That single discipline prevents most wasted effort, because the fixes are not interchangeable.
If visitor to enquiry fell, something changed on the page or in the traffic reaching it. Open the site on a phone on mobile data and try to enquire. Most causes are visible in under a minute, and if the page is the constraint that is a website problem rather than an advertising one.
If enquiry to quote fell, something changed in the response. Somebody left, somebody got busy, or the phone started going to voicemail during the part of the day when most people ring.
If quote to job fell, something changed in the market or the offer. Prices moved, a competitor started undercutting, or lead times became unacceptable. That is a commercial conversation, and no amount of advertising fixes it.
Where those repairs have been made and the numbers published alongside the window they were measured in, our written-up engagements state the conversion definition beside the result, which is the only way a rate is worth comparing to anything.
Common questions
What conversion rate should my website have?
In SoudCoh's research sample of trade businesses, a median 14.7% of Google Ads clicks became a recorded call or form (12.3% plumbing, 14% electrical, 20.1% cleaning). That is the visitor-to-enquiry rate only; measure enquiry-to-quote and quote-to-job separately, and compare only against rates counted the same way.
Why do published benchmarks vary so much?
Because they aggregate businesses counting different things. One includes phone calls, another does not. One counts an unread form fill, another counts a booked job. Once the definitions differ, the averages describe nothing you can act on.
Which of the three rates should I improve first?
The one with the biggest gap between what happens and what should happen, which is usually enquiry to quote. Most trades businesses lose more work to slow or missed responses than to anything on the website.
Do phone calls count as conversions?
They must, in any trades business. If calls are not counted, the website conversion rate understates reality and every decision built on it skews toward customers who prefer typing over ringing.
How often should I recalculate?
Quarterly is enough for most trades businesses, and more often than that turns normal variation into false alarms. Recalculate immediately, though, if you change what you count — and note the date you changed it, because every comparison across that point is invalid.
Questions and answers
Should mobile and desktop visitors share one benchmark?
Keep the combined view, but inspect devices separately when their journeys differ. A mobile form or contact problem can be hidden inside an otherwise stable total.
How should repeat enquiries be handled?
Agree whether the measure concerns requests or distinct customers and apply that rule consistently. Keep the original records available so repeated contact is not mistaken for separate accepted work.