What a KPI is
- It shows one part of the business working. Cost per lead shows if your marketing works. Close rate shows if your sales work
- It has one owner, by name
- It has two goals: a blood goal, the number you have to hit, and a stretch goal, the number you push for
- It gets checked every week
KPI stands for key performance indicator. A business dashboard can easily show 20 numbers. If you try to watch all of them every week, you end up watching none. A number missing any of the four things above is just a number on a screen.
When a number belongs to everybody, nobody fixes it. When it belongs to one person, that person knows it is theirs and comes back with a fix.
Three departments, three owners
| Department | Its job | The numbers it owns |
|---|---|---|
| Marketing | Get the right homeowners to ask for a price, for the least money | Cost per lead, lead quality, cost to get a job by method |
| Booking | Turn leads into estimates on the calendar, and get them to show up | Speed to lead, booking rate, cost per estimate. Also watches show rate |
| Sales | Turn the estimates that happen into paid jobs | Close rate, cash per showed estimate. Also watches average sale |
Each department hands its work to the next one. When results drop, the problem is usually in one of them, not all three.
Marketing: cost per lead, lead quality, cost to get a job
| Number | How to work it out |
|---|---|
| Cost per lead | Ad spend ÷ leads, for the same week |
| Lead quality | Real leads ÷ all leads. A real lead is a homeowner with a real wood or LVP floor, inside your area, who picks up |
| Cost to get a job, by method | What a method cost ÷ the jobs it sold |
For home improvement lead ads on Facebook and Instagram, the average cost per lead is about $43, according to LocaliQ's 2026 benchmarks. That figure comes from an ad company and isn't specific to floors, but it's a starting point. Set your own goals after 4 weeks of your own numbers.
A cheap lead that never books is the most expensive lead you will buy. That is why lead quality sits next to cost per lead. Give every lead that does not book or buy a lost reason, like Wrong floor, Outside area or Renter. When those pile up, fix the ad: show a real wood floor, say "hardwood", name your towns.
A made-up example, not a result: in one month, Meta ads cost $200 for each job they sold, Google $300, door hangers $150 and referrals nothing. The next dollar goes to the cheapest paid method, and you never cut referrals.
Booking: speed to lead, booking rate, cost per estimate
95%
Of 466 home-service companies didn't reply to a web inquiry within 5 minutes (Valve+Meter, 2019)
40%
Never replied within five days (Valve+Meter, 2019)
Nearly 7x
As likely to qualify the lead, for firms that replied within an hour (Harvard Business Review, 2011)
| Number | How to work it out |
|---|---|
| Speed to lead | How fast a new lead gets a real phone call. Aim for within 5 minutes |
| Booking rate | Estimates booked ÷ leads. Half or more is a good target to start with |
| Cost per estimate | Ad spend ÷ estimates booked. In a made-up example, $30 a lead at 50% booked is $60 an estimate |
| Show rate | Estimates that happen ÷ estimates booked |
Speed decides who books. The Harvard Business Review study looked at 2,241 firms. Booking is whoever calls your leads, so Booking owns speed.
When the booking rate is red, check in this order: speed to lead, then lead quality, then how many times each lead was called and texted, then the call itself. When cost per estimate is red, it is always one of two numbers: cost per lead, which is Marketing, or booking rate, which is Booking.
Sales: why close rate can lie
| Number | How to work it out |
|---|---|
| Close rate | Jobs sold ÷ estimates that happened. No-shows don't count. They're a Booking problem |
| Cash per showed estimate | Cash collected ÷ estimates that happened |
| Average sale | Sales ÷ jobs sold. You set your own prices |
A made-up example, not a result. Two people each run 10 estimates on 1,150 square foot floors. Rep A closes 4 at $1.50 a square foot. Rep B closes 3 at $2.50 a square foot. Rep A has the better close rate. Rep B brings in more cash for every estimate that showed up: about $863, against about $690 for Rep A. Both prices sit inside the $1.50 to $3 a square foot that homeowners commonly pay for a recoat.
That is why cash per showed estimate matters most in Sales. It counts how often someone closes and how much they close for. Your goal for it is your close rate goal times your average sale, on your own numbers. A close rate of 30% is a target to aim for, with 35% as a stretch.
A red number isn't always Marketing's fault
Cost to get a job is the whole chain: cost per lead ÷ booking rate ÷ show rate ÷ close rate. In a made-up example, $30 a lead ÷ 50% booked ÷ 80% showed ÷ 30% closed is $250 a job.
So a red cost per job can come from any department. Find the red number furthest behind in the chain and fix that one first.
One owner per department
At the start, you own all three, and that's normal. Write your name next to each one anyway, so you look at each part of the business on its own.
When you hire, hand over a whole department and its numbers at the same time. If you hire someone to call your leads, they own speed to lead and booking rate from day one, so they know what a good week looks like.
The owner fixes the number. They don't just report it. They come to the weekly check with the reason and one fix.
Red, green, gold: the weekly check
- 1
Read your numbers every Monday
Last week, next to the week before
- 2
Color each one
Red is below the blood goal, green is at it, gold is at the stretch goal
- 3
Pick one red number
If more than one is red, fix the one furthest behind first. One fix you actually do beats three you forget
- 4
Check the fix next week
The first question next week is whether the fix moved the number
A made-up example: a blood goal of $30 a lead and a stretch goal of $25. $34 is red, $28 is green and $24 is gold. For costs, lower is better.
Your first step
Write an owner's name next to Marketing, Booking and Sales, even if every name is yours. Give each number a blood goal and a stretch goal. Then every Monday, each owner reads their own few numbers, picks the red one furthest behind, and fixes just that.