One dashboard for a contracting company: what to actually track

August 18, 2026 · Joshua Hill

A contracting company needs one screen with about eight numbers on it: leads in by source, speed to first response, quotes sent, close rate measured in dollars, signed backlog, gross margin by job, cash on hand, and receivables past thirty days. Everything else is a report you pull when you have a specific question, not a number you watch every week.

Most contractor dashboards fail for the opposite reason people expect. They get stuffed with numbers nobody acts on, so the whole thing gets ignored by the third week and quietly stops being updated.

Numbers you watch versus reports you pull

Accounting software already tells you what happened. It exists to file taxes, settle disputes, and close the books, and it is very good at describing a quarter that is already over. A dashboard has a different job. It tells you whether this week is going the way you wanted while there is still time to change it.

The test for whether a number belongs on the dashboard is simple: would you do something different this week if it moved? Revenue year to date fails that test for most contractors, because by the time it moves the decisions that moved it are months old. Speed to first response passes, because a bad number there means somebody answers the phone differently starting tomorrow morning.

The eight numbers

• Leads in, by source. Counted the same way every week, and counted including phone calls. A lead count built only from web form submissions will tell you your best channel is your worst one.
• Speed to first response. The clock from when the lead arrives to when a human contacts them. This is the number that quietly decides how many of the other numbers matter, and we went through the timing in the post on estimate speed.
• Quotes sent and total dollars quoted. Volume out the door, which is what a slow month usually turns out to have been missing.
• Close rate in dollars, not in count. Ten small wins next to one lost large job reads as a strong month when you are counting jobs and a weak one when you are counting revenue. Dollars is the honest version.
• Backlog in signed dollars. Work sold and not yet completed. This is how far out you are booked, and it is the number that should drive whether you hire or whether you push marketing.
• Gross margin by job, actual against estimated. Run on every job that closed out this month. Estimating errors only show up here, which is the reason pricing for margin falls apart without it.
• Cash on hand against the next two weeks. Payroll and material commitments included. Profitable companies still go under on timing.
• Receivables past thirty days. One line, listed by customer, so an aging invoice has a name attached to it instead of sitting inside a total.

Where the numbers come from

You almost certainly already have all eight. They are scattered across the phone system, the email inbox, the estimating tool, and the accounting file, which is why nobody looks at them together. The work is collection on a schedule, so no one has to go hunting.

Start manual. A spreadsheet with one row per week and eight columns will beat a dashboard product you never finish configuring. Fill it in by hand for a month and something useful happens: you find out which numbers you cannot actually get. That is the real finding. A number you cannot pull is a process that is not being recorded anywhere, and it needs fixing before any software will help.

Then automate the ones that hurt to gather. Phone systems export call logs and answer times, estimating tools export quotes sent and accepted, and the accounting file carries margin, cash, and aging. Pulling those exports into one sheet on a schedule is most of the build, and it is the kind of plumbing the system we run is made of.

How often to look at it

Weekly is right for the operating numbers. Monthly is right for margin, because a job has to close out before its real cost is known. Daily applies to one thing only: response time on new leads, the only number here that goes bad in hours rather than weeks.

A dashboard nobody reads is decoration. Give it a standing time on the calendar, same day and same half hour every week, and read the numbers in the same order each time so a change stands out. Consistency matters more than the tool you build it in.

What we do in our own companies

We run painting and flooring companies in Bellingham, and our dashboard started as a spreadsheet. Parts of it still are one. The two numbers we look at hardest are speed to first response and actual margin against estimated margin on closed jobs, because between them they catch most of what goes wrong before it compounds.

Website visits are not on the operating dashboard. Traffic is a marketing question with its own review, and mixing it in with cash and backlog gives the weekly meeting somewhere pleasant to go instead of somewhere useful.

The honest part: no dashboard fixes anything by existing. It shortens the gap between a problem starting and somebody noticing, and that gap is where most of the money in a contracting company gets lost.