Most owners check one number every week: revenue, or how full the diary is. Both are useful. Both are late. By the time revenue drops, the cause happened weeks ago, somewhere between the ad and the signed contract.
The fix is a short weekly review of the numbers that come before revenue. Seven numbers, fifteen minutes, every Monday. Here they are, in the order a customer passes through them, with how to work each one out and what to look at first when it moves.
1. Leads and cost per lead
What it is: new enquiries this week, and total marketing spend divided by those enquiries. Track it by channel.
Why it matters: it is the top of the funnel and your earliest warning. If leads drop this week, surveys drop next week and sales drop the month after.
When it moves: check the obvious first. Did an ad stop, a budget run out, a form break, or a Google listing go down? Then look at seasonality and creative fatigue. A rising cost per lead is not always bad: tighter qualifying raises it while improving the leads.
2. Contact rate and speed to lead
What it is: the share of new leads you actually spoke to (or got a reply from), and how long it took to make first contact.
Why it matters: an uncontacted lead is money spent for nothing. And speed matters: enquiries that wait hours often go to whoever called first.
When it moves: look at when leads arrived. Evening and weekend enquiries are the usual gap. AI replies and booking out of hours close it. See why speed to lead decides who wins.
3. Lead to appointment rate
What it is: surveys, visits or design appointments booked, divided by leads.
Why it matters: it shows whether your leads are real and whether your setters or office team are turning conversations into appointments.
When it moves: split it by channel and by the person booking. If one channel drops, the lead quality changed. If one person drops, it is a skills or effort issue. If it falls across the board, look at the first conversation script and the qualifying questions.
4. Show rate
What it is: appointments that went ahead divided by appointments booked.
Why it matters: a no-show wastes a surveyor's or salesperson's half day, the most expensive time in the business.
When it moves: look at how far ahead appointments are booked (long waits cause no-shows), whether confirmations and reminders are going out, and whether both decision-makers were booked in. We cover this in detail in fixing no-shows before the visit.
5. Close rate (on sat appointments)
What it is: sales divided by appointments that actually happened. Use sat appointments, not leads, so this number measures selling, not marketing.
Why it matters: it is the clearest measure of your sales process and your offer.
When it moves: split by rep and by lead source. A drop for one rep is coaching. A drop for one source is lead quality. A drop across everyone often means price, finance or a competitor offer. For what "good" looks like and how to measure it fairly, see what's a good close rate?.
6. Average order value
What it is: total sold value divided by number of sales.
Why it matters: two firms with the same close rate can have very different businesses. Add-ons, upgrades and finance all move this number.
When it moves: a drop can mean reps are discounting to close, or the product mix has shifted towards smaller jobs. A rise with a falling close rate can mean you are only closing the biggest buyers. Watch the two together.
7. Cost per sale
What it is: total marketing spend divided by number of sales, overall and by channel.
Why it matters: this is the number that tells you whether marketing is making money. It pulls numbers 1 to 5 together into one figure. If you want the full method for tracking it by channel, read tracking every lead to a sale.
When it moves: go back up the list. Cost per sale only changes because something above it changed.
The weekly dashboard
Put the seven numbers on one screen, this week against the last four weeks:
| Number | This week | 4-week average | Direction |
|---|---|---|---|
| Leads | |||
| Cost per lead | |||
| Contact rate | |||
| Lead to appointment | |||
| Show rate | |||
| Close rate | |||
| Average order value | |||
| Cost per sale |
Compare against your own recent average, not against someone else's benchmark. Your trade, area, price point and lead sources make your numbers yours.
Two extra numbers worth a glance
- Pipeline value: the total value of open quotes. If it shrinks for several weeks, a quiet month is coming. See feast or famine.
- Forward bookings: how many weeks of installs are sold. It tells you when to push marketing and when you have room to be choosy.
How to run the review
- Same time every week. Monday morning works for most.
- Numbers first, stories second. Look at the dashboard before anyone explains anything.
- Pick the one number that moved most against the four-week average.
- Go up the chain to find the cause.
- Agree one action, one owner, one date. Not five. One.
- Check last week's action before starting.
The review should take fifteen to thirty minutes. If gathering the numbers takes longer than reading them, that is the first problem to fix.
Where the numbers come from
All seven come from three places: your ad accounts (spend), your CRM (leads, contacts, appointments, outcomes) and your sales records (values). If those sit in different systems, someone has to stitch them together every week, and it stops happening by week three. A CRM built for installers does it automatically. Our Installer Flow reporting puts pipeline, reps, cost per lead and cost per sale on dashboards, so the Monday review is just reading, not counting. For home improvement firms with long sales cycles, that view of the pipeline is what stops surprises.
If you want a weekly dashboard set up for your business, book a call and we will show you what it would look like with your own numbers.
Questions owners ask
Which KPI matters most for a home improvement business?
If you only track one, track cost per sale, because it ties marketing spend to signed work. But it arrives late. The earlier numbers in the chain, such as contact rate and show rate, tell you what is about to happen to cost per sale, so they are where you spot and fix problems.
How long should a weekly numbers review take?
Fifteen to thirty minutes, if the numbers are already on a dashboard. Look at the seven numbers, compare with the last four weeks, pick the one that moved most and agree one action. If it takes hours to gather the numbers, that is the first thing to fix.


