KPIs by Department for a Home Improvement Company

It is Monday’s leadership meeting. Marketing says leads are up, sales says the leads are junk, production says the backlog has never been bigger, and the controller says cash is tight. Everyone brought a number, every number is true, and you walk out still not knowing what to fix first.

The problem is scorekeeping: each department counts its own way, and nobody agreed which numbers belong to whom. The owner’s view starts with profit and a few numbers that ladder up to it; we cover that tree, and how to cut vanity metrics, in which KPIs actually tell you the company is winning. This page goes one level down: what each department should measure, how to calculate it, and what its manager should do when it moves.

Which KPIs should each department in a home improvement company own?

Three to five each, defined in writing, with one owner. In a company doing $10M or more, marketing, the call center and sales carry the front of the house; production and finance carry the back. The table below is the full set. Definitions vary from shop to shop, so the right answer is to pick one, write it down, and use it everywhere.

Read it left to right and every row answers three questions: what is the number, who owns it, and what do they do when it moves. A row that cannot answer the last question does not belong in the table.

What should a home improvement marketing team measure?

Cost per booked job by source is the number marketing should steer by. Cost per lead is only its input, useful for deciding where to test; a channel with cheap leads that never sign is expensive, not cheap (what a booked job really costs by channel shows the gap). Marketing cost as a share of net sales is the guardrail: if it climbs while net sales stay flat, you are buying volume that is not becoming signed work.

One warning before anyone moves budget: if the marketing report and the sales report disagree on what a channel produced, neither number can be trusted yet. That is a reconciliation problem to fix first, and it is usually where the marketing money is quietly leaking.

What should the call center measure?

Speed to lead first, because it drives the two numbers after it. A homeowner who just asked for a quote is often asking other companies too, and their attention fades by the minute. Reach them fast and you reach more of them at all (contact rate), and only a reached lead can book an appointment (set rate). Measure speed to lead as the median minutes from a lead arriving to real contact, not the average, because one bad Saturday hides the problem; the five minute window covers the research behind it.

What should the sales team measure?

Demo rate, close rate by rep, net sales per lead issued and cancellation rate. Together they show where appointments fall out between the kitchen table and a signed, kept contract, and they tell the sales manager who to coach and who should get the next lead.

Demo rate is the share of issued appointments a rep actually runs; a low number points to confirmation calls and no shows, not selling. Close rate by rep is sales divided by appointments run. Net sales per lead issued, often shortened to NSLI, is net sales dollars divided by appointments issued, where net means after cancellations. It folds demo rate, close rate and ticket size into one number per appointment, which makes it the fairest way to compare reps. Cancellation rate is the check on all of it: a signed job that cancels in the rescission window was never revenue.

What should production measure?

Backlog ready to start, on time start rate, cycle times and rework. Production’s job is to turn signed work into completed work without crews sitting idle or jobs bleeding margin in the field, and these four numbers show where that breaks down.

Backlog ready to start is the production number most owners never see. A big backlog looks like safety, but if most of it is waiting on a measure, a material order or a permit, crews sit idle while the backlog grows. Splitting backlog into ready and not ready, gate by gate, turns a vague worry into a weekly list of what to chase. On time start rate is the earliest signal that scheduling is slipping. Days sold to start and days start to complete show where jobs stall, and rework and punch list rate shows how often crews go back, which quietly eats the margin you thought you signed.

What should finance measure beyond profit?

Gross margin at signing, realized job margin, and cash collected against what was invoiced. Profit tells you how the month went; these three tell you why. Put the two margins side by side: gross margin at signing is what you expected a job to make, realized job margin is what it made after materials, labor and return trips, and the gap between them is where estimating, change orders and rework show up in dollars.

Cash collected versus invoiced closes the loop. Your CRM records what was sold and invoiced; your books record what actually arrived. Refunds, canceled financing and credit memos live in the gap, and a report built only on the CRM will never show it.

If your departments are measuring against different systems and you cannot tell which number is right, a free Profit Leak Audit reads your CRM, ad accounts and books together and shows where signed revenue is leaking.

How do you get every department to use the same definitions?

Write each definition down once, on one page, and make every report use it.

It is the end of the month. Marketing’s report calls it a record month for leads, sales says set appointments were flat, and the P&L shows less revenue than either one implied. Each manager defends their number, the meeting turns into an argument about whose report is right, and you walk out trusting none of them. A few months of that and it is tempting to stop reading reports at all and go back to gut feel.

Usually every report is right on its own terms; they count different things under the same name. Set rate is the classic case: one report divides by every lead received, another by the leads the call center actually reached. A sale can mean signed, or signed and past the cancellation window; revenue can mean invoiced, or collected.

Putting that fire out takes one meeting, not new software. Get the managers in a room, agree on one definition for each of those words, put them on a single page, and make every report use that page. From then on, when two reports disagree, check the definitions before anyone defends a number; often that is where the argument ends.

How should a home improvement company set up department KPIs?

You can set this up in one afternoon with your managers in the room, using the reports you already have. You walk out with a short list of numbers for each department, each one defined the same way by everyone and owned by one person. Work through it in this order:

1. Write profit at the top of a whiteboard. Every number you keep has to explain something about profit. If you cannot draw a line from a number to profit, it does not make the list.
2. Gather every report people use today. Pull up the CRM reports, the ad platform dashboards, the P&L and each manager’s spreadsheet, and write down every number on them. Expect a long list.
3. Ask two questions about each number: what decision does it change, and who makes that decision? “Cost per booked job by source decides where next month’s ad budget goes, and the marketing manager makes that call” passes. A number nobody can answer for gets cut.
4. Give each surviving number to one department and write its formula next to it. Use the table above as a starting point, and spell out the words people argue about: what counts as a lead, an appointment and a sale.
5. Decide how often each number gets looked at. Daily if someone acts on it within the hour, like speed to lead; weekly for the sales funnel and the backlog; monthly for cost, margin and cash. Note whether it is an early warning (leading) or a result (lagging).
6. Test one number against a second system before you trust it. Compare sales in the CRM with deposits in QuickBooks for the same month. If they do not match, fix the definition first, because a dashboard built on a number nobody trusts gets ignored.

How often should each department review its numbers?

Review each number as often as someone can act on it. Check it more often than that and people learn to ignore it; check it less often and you find the problem a month late.

Each department keeps its rows on its own screen; the owner’s screen holds only the handful that ladder straight to profit. How to assemble those views into one screen in Tableau, with job level profit from QuickBooks, is its own article. If a dashboard died in your company before, the reason it got ignored is usually that it displayed data instead of answering a department’s decisions.

When does a department not need its own KPI set?

When one person runs it alongside another. Below about $10M, one manager often covers marketing and the call center, and another covers production and the office. The rows in the table stay the same; the owners merge, and so do the screens.

Keep each department’s list short, and give every number an owner who acts on it. If no decision is waiting on the data yet, you probably do not need the expensive model. If the deeper question is how to buy analytics help at all, start with the buying guide.

Frequently Asked Questions