Short answer: track four numbers every week, defined the same way every time: new leads, new customers, cost to sell and gross margin. Divide new customers by new leads for your close rate, and cost to sell by new customers for what each customer costs you to win. Compare each week with the same week last year so seasonality doesn't fool you.
The four numbers
| Number | What counts | What doesn't | Where the data lives |
|---|---|---|---|
| 01New leads | Anyone not already a customer who asked for a quote, visit or price, by any channel | Past customers asking for more work | Phone log, CRM, form inbox |
| 02New customers | Leads who became a paying job in the same period | Repeat customers, upsells on existing jobs | Closed jobs, first-time invoices |
| 03Cost to sell | Sales pay, commissions, sales managers, agency fees, every ad dollar (last 12 months) | Office overhead, your own pay | Payroll plus marketing lines on the P&L |
| 04Gross margin | Revenue minus crew, materials and equipment on the job | Office, trucks, insurance, you | Gross profit ÷ revenue on the P&L |
The three numbers you get from them
- Close rate = new customers ÷ new leads. How many people who ask end up buying.
- Cost per new customer = cost to sell ÷ new customers. What you pay to win one.
- What a customer is worth = average job revenue × gross margin. Whether that cost was worth paying.
An example
A company gets 80 new leads in a month and 24 become new customers: a 30% close rate. Over the last twelve months it spent $96,000 on sales pay, its agency and ads, and won 300 new customers: $320 per new customer. If an average first job is $900 at 50% gross margin, each new customer is worth about $450 on the first job alone, so the $320 was worth paying, and repeat work is upside. (Illustrative figures, not a client's.)
Where it goes wrong
- Counting repeat customers as leads. Marketing looks better than it is, and the cost of a truly new customer stays hidden.
- Leaving sales pay out of cost to sell. Ad spend is only part of what it costs to win a job.
- Reading only your CRM. Calls that never became a CRM record, and invoices that live only in accounting, go missing.
- Paying ads to win back your own customers. Without matching leads to the customer record, paid search takes credit for people who were coming back anyway. One client found 41% of a year of paid search going to existing customers.
- Comparing week to week. Home services is seasonal. Compare each week with the same week last year.
How often to look
Weekly for leads, new customers and close rate; monthly for gross margin, once the books close. The point is one page, the same definitions every time, on the same day each week.
Common questions
What is a good close rate for a home services company?
It depends on the trade and the lead source, so the useful comparison is your own close rate against the same months last year and by lead source. Track it weekly with the same definition: new customers divided by new leads.
How do I calculate cost per new customer?
Divide your cost to sell (sales pay, commissions, agency fees and all ad spend over the last twelve months) by the number of new customers in the same period.
Should repeat customers count as leads?
No. Count only people who were not already customers. Repeat business matters, but mixing it in makes marketing look better than it is and hides what new customers really cost.
Want these four numbers on one page every Monday? That's our Monday Report, built from your phone, CRM, invoices and ads and checked against your books. You can also send us the numbers you have, and we'll reply with what we see.
Book a 25-minute call →