Most owners look at two numbers: revenue and jobs completed. Both are lagging indicators — they describe a month that is already finished and cannot be changed. By the time revenue tells you something is wrong, the cause was six to ten weeks ago and the evidence has gone cold.
The seven below are chosen for a different property: each one moves before revenue does, each has a threshold where you should act, and all seven fit on one screen in about ten minutes on a Monday. This is not a dashboard project. It is a habit.
1. Median first response time
From lead arriving to your first outbound contact, measured as a median rather than an average — one lead answered nine days late will drag a mean into fiction. Split it by channel, because in almost every business one channel is dramatically worse than the rest, and it is usually the newest one nobody owns.
Act when: the median passes 15 minutes for two consecutive weeks, or any single channel exceeds an hour. The fix is nearly always an orphaned channel rather than a lazy team.
2. Quotes with zero follow-up
The count of open quotes that have received no contact since being sent. This is the most direct measure of money left on a table you already paid to set — the call, the drive, the measure, and the pricing are all sunk before this number even exists.
Act when: it is anything other than zero. There is no acceptable level, because a scheduled sequence makes zero the default and any other number means the sequence is not running or not covering something.
3. Open pipeline value
The total dollar value of quotes sent and not yet decided. Owners consistently underestimate this and consistently over-index on last month's revenue instead. Pipeline value is the closest thing a service business has to a forecast, and it is the number that tells you whether next month is already in trouble.
Act when: pipeline falls below roughly one month of target revenue. That is a marketing problem with a six-week lead time, so noticing it late is the expensive part.
TARGETS
4. Days to invoice
Average days between job completion and invoice sent. Not days to payment — days to invoice, because that portion is entirely within your control and it is usually where the delay actually lives. Most businesses discover the gap is a week or more, and that every day of it is a day of their own cash sitting in someone else's account.
Act when: it rises three weeks running, or exceeds two days. Both usually point at the same cause — invoicing happens at a desk in the evening instead of at completion on site.
5. Rescheduled-by-us rate
The share of booked jobs that you moved, as opposed to the customer. This is the best available proxy for whether your schedule has any slack in it, and it is a leading indicator of both bad reviews and staff burnout — two things that show up in revenue much later.
Act when: above 5% of booked work. The cause is almost always booking to full capacity, which turns a single emergency into four apologetic phone calls.
6. Repeat and referral share of revenue
What percentage of the month's revenue came from existing customers or people they sent. This is your retention engine and your true cost of acquisition in one number. A business at 60% is compounding; a business at 15% is renting its revenue from an ad platform every single month.
Act when: the share falls for two consecutive months, or sits below about 30%. The remedy is maintenance plans and reactivation, not more advertising.
7. Reviews in the last 30 days
A raw count, not an average rating. Rating is a slow-moving stock; count is a flow, and flow is what local search actually responds to. It is also the most sensitive early indicator of whether your post-job process is still running, because the review request is the first thing that quietly stops happening when a season gets busy.
Act when: the count drops below your monthly job count divided by five. That usually means the request has stopped firing, not that customers stopped being happy.
The one-screen version
| Number | Healthy | Act when |
|---|---|---|
| Median first response | Under 5 minutes | Over 15 minutes, two weeks running |
| Quotes with no follow-up | Zero | Anything above zero |
| Open pipeline value | 1.5–2× monthly target | Below 1× monthly target |
| Days to invoice | Same day | Over 2 days, or rising 3 weeks |
| Rescheduled by us | Under 3% | Above 5% |
| Repeat + referral share | Above 40% | Below 30%, or falling twice |
| Reviews last 30 days | Jobs ÷ 5 | Below that for a month |
Two rules that make this stick
First: if a number takes more than a minute to produce, it will not be produced. Any metric requiring a spreadsheet export and a pivot table gets checked twice and then never again, which is why the list stops at seven and why they all need to come off one screen.
Second: a number without a threshold is decoration. "Response time is 22 minutes" prompts nothing. "Response time is 22 minutes, which is over the line, so today we find out which channel is dragging it" is a decision. Write the thresholds down before you look at the numbers, because deciding them afterwards is how every uncomfortable number gets rationalised.
Revenue tells you about a month you can no longer change. These seven tell you about the one you are still in.
Common questions
What KPIs should a service business track?
Seven that lead revenue rather than lag it: median first response time, quotes with zero follow-up, open pipeline value, days to invoice, jobs rescheduled by you, repeat and referral share of revenue, and reviews in the last 30 days. Each has a threshold at which you should act.
How often should a small business owner review metrics?
Weekly, in about ten minutes, on the same day each week. Monthly is too slow to catch a leak while it is still cheap, and daily turns normal variation into noise you will start ignoring.
Why measure days to invoice instead of days to payment?
Because days to invoice is entirely within your control, and it is usually where the delay actually sits. Most businesses find a week or more between job completion and the invoice going out — every day of which is their own cash sitting in someone else's account.
What is a good repeat and referral share for a service business?
Above 40% suggests the retention engine is working and acquisition costs are compounding downward. Below 30% means most revenue is being rented from an ad platform each month, and the fix is maintenance plans and reactivation rather than more advertising.