Five numbers every flooring or remodeling business should be able to state without looking them up. In our experience most owners can state one.
They are not five separate facts. They are one chain, and each one hands its answer to the next.
Cost per lead
What one inquiry costs. Widely quoted, least useful in isolation, but required to calculate the second.
How to work it out. Everything you spent to generate inquiries in a period, divided by the inquiries it generated. Everything means the ad spend plus whatever you paid somebody to run it, not just the platform invoice.
The trap. It is the easiest of the five to make look good. Widen a radius, take a question off the form, and it falls while quality falls with it. A cost per lead moving in the right direction on its own is not news.
Cost per booked estimate
What one booked appointment costs, before anybody shows up. This is the operative figure. Inquiries that never convert to an appointment still carry full acquisition cost, so a fifty percent booking rate silently doubles the true cost of every estimate booked.
How to work it out. The same total spend, divided by the number of estimates that got booked. Then run it a second time against the visits you actually attended, because a no-show cost you the drive as well, and judge the money on that one.
The trap. Owners quote this as though it were cost per lead with a better name. It is not. It is the number that tells you what a homeowner in front of you costs, which is the only version of the question your calendar cares about.
Booking rate
Inquiries converting to a scheduled visit. Below the mid twenties, the constraint is usually response speed or absent qualification rather than lead source.
How to work it out. Visits attended, divided by inquiries received, over the same period. No money involved, which is why it is the one to start with.
The trap. Blaming the source first. Before you change where the inquiries come from, check how long they waited for a call and whether anybody asked the four qualifying questions. Both of those are free to fix and both move this number faster than a new lead source will.
Close rate
Attended estimates converting to signed work. When this is weak the common correction is a price reduction. The more productive question is who is being quoted, because quoting unqualified buyers depresses close rate independently of price.
How to work it out. Jobs won, divided by estimates attended. Count the ones that went quiet as losses, because they are.
The trap. Reading it as a report card on your selling. Split your last twenty by source and the two piles will usually behave like two different companies, which tells you the number is describing your funnel more than your performance in a living room.
Average job value
Ten jobs at $15,000 and thirty at $5,000 can produce similar revenue while representing entirely different operations, cost structures and constraints.
How to work it out. Revenue from completed work, divided by the number of jobs. Then look at the spread as well as the average, because one large job can hide a month of small ones.
The trap. Treating it as a fact about your market rather than a consequence of your choices. It moves when the product mix moves, when the qualifying questions change, and when you stop attending the small jobs that eat a full day.
Agree the definitions before you count anything
The usual way this exercise fails is not laziness. It is that the definitions drift, so month three measures something month one did not, and the trend line becomes fiction.
Write three sentences down and keep them where you can see them.
- A lead is a named person, with a way to reach them, who wants a job doing. A wrong number is not a lead. A supplier calling you is not a lead.
- An attended estimate is a visit that actually happened. Booked and cancelled does not count, and this is the one people quietly fudge because it flatters everything downstream.
- A won job is signed and scheduled, not “they sounded keen on the phone.” Pick the moment and use the same moment every month.
Then keep referrals and repeat customers in their own column rather than folding them into the total. They behave nothing like paid inquiries, and mixing them in makes your advertising look better than it is.
How the five chain together
What follows is an illustration and not anybody’s real figures. The arithmetic is the point. Put your own numbers in.
| The step | Company A | Company B |
|---|---|---|
| Spent in the month | $3,000 | $3,000 |
| Inquiries | 60 | 40 |
| Cost per lead | $50 | $75 |
| Booking rate | 25% | 60% |
| Estimates attended | 15 | 24 |
| Cost per booked estimate | $200 | $125 |
| Close rate | 30% | 45% |
| Jobs won | 4 or 5 | 10 or 11 |
Company A has the better cost per lead and it is the one in trouble. That is the whole argument for tracking more than the first number, and it is why the second one exists.
Notice also what Company B did not do. It did not spend more and it did not find a magic source. It answered faster and qualified earlier, and the two numbers in the middle of the chain moved.
Read them in pairs
One number moving tells you very little. Two moving together tell you what actually happened.
- Cost per lead down, booking rate down. The inquiries got cheaper and worse. Nothing improved.
- Booking rate up, close rate down. You are attending visits that were not qualified. Something got loosened, usually the budget question.
- Close rate up, average job value down. You are winning the easy small ones. Check what that did to the calendar before you celebrate it.
- All four flat, revenue up. The job mix moved. Find out why, then find out whether you can do it on purpose.
The one page that carries all of it
None of this requires software to begin. A written record and honest inputs will get most businesses the majority of the value.
One row per inquiry, six columns, filled in as it happens rather than reconstructed at month end.
- Date the inquiry landed.
- Where it came from, named specifically rather than “online.”
- Whether a person actually spoke to them, and how long it took.
- Whether a visit was attended, and on what date.
- Won, lost or quiet.
- The value, when it is won.
That sheet answers four of the five directly and the fifth once you add your spend at the bottom of the month. It takes about thirty seconds per inquiry, and the discipline of filling it in matters more than the tool you fill it into.
What to do with them once a month
- Booking rate down? Look at response time first, then at whether anybody is qualifying. Do not touch the ad account yet.
- Close rate down while booking rate is up? You loosened the qualifying. Tighten one question and watch it for a month.
- Cost per booked estimate rising? Either the inquiries got more expensive or fewer of them turned into visits. The chain tells you which in about a minute.
- Average job value falling? Check the product mix and check what you are agreeing to attend. Small jobs cost the same half day as big ones.
The one you cannot answer
The number you cannot currently answer is usually the one costing you the most.
For most owners that is the second one, which is why it has a piece of its own: cost per booked estimate. If booking rate is the one that is off, the two arguments behind it are what a free estimate really costs you and who is calling your inquiries. And before you hand these numbers to anybody else to manage, run the seven questions on them.
