How to Measure Job Backlog (Weeks of Work Booked)
The short answer: Job backlog is signed work you have not yet completed. Measure it in weeks of crew capacity, not dollars: take the remaining labor hours on every accepted job and divide by the crew hours you can actually deliver in a week. Four to eight weeks is a comfortable range for most service businesses. Under two weeks means sell now; over twelve means you are quoting work you cannot staff. CRMb tracks accepted quotes, scheduled jobs, and crew capacity on one record, so backlog is a view you open rather than a spreadsheet you rebuild every Monday.
Most owners answer "how much work do we have booked?" with a dollar figure and a shrug. Dollars are the wrong unit. A $200,000 backlog means nothing until you know whether it is four weeks of work for your crew or six months of it.
What actually counts as backlog
Backlog is signed, not yet delivered. That definition rules out more than people expect.
Counts:
- Accepted quotes with a scope the customer has approved.
- Jobs scheduled but not started.
- Jobs in progress, but only the remaining portion.
- Contracted recurring maintenance, only for the term already committed.
Does not count:
- Open quotes still waiting on a decision. That is pipeline, not backlog, see sales pipeline stages for contractors.
- Verbal "we'll definitely use you in spring" promises.
- Work you completed but have not invoiced. That is receivables, see how to track accounts receivable.
Mixing pipeline into backlog is the most common mistake, and it is the dangerous one. It makes a slow quarter look full right up until the day the crew has nothing to do.
Measure it in weeks, not dollars
Dollars vary wildly per labor hour. A material-heavy install and a labor-heavy repair can carry the same price and consume completely different amounts of the only resource you cannot buy on short notice: crew time.
The calculation:
Backlog weeks = remaining estimated labor hours / weekly crew capacity
Remaining estimated labor hours. Every accepted job's estimated hours, minus hours already burned on jobs in progress.
Weekly crew capacity. Field-productive hours per week, not payroll hours. A four-person crew at 40 hours is 160 payroll hours, but drive time, shop time, and callbacks realistically leave 120 to 140 billable. Use your real number, see crew utilization rate.
A worked example:
| Job | Estimated hours | Hours burned | Remaining |
|---|---|---|---|
| Miller deck | 120 | 45 | 75 |
| Ridgeview retaining wall | 200 | 0 | 200 |
| Ortiz drainage | 60 | 0 | 60 |
| Hillcrest maintenance (Q3 term) | 90 | 30 | 60 |
| Total | 395 |
At 130 productive crew hours a week, that is 395 ÷ 130 = 3.0 weeks of backlog. Not four months of security. Three weeks.
What a healthy number looks like
There is no universal target, but the bands behave consistently:
- Under 2 weeks. Selling is now the only priority. You are one cancellation from an idle crew, and idle crews are the fastest way to convert a profitable month into a loss.
- 2 to 4 weeks. Workable but thin. Keep quoting aggressively and be careful about discounting to fill the calendar, discounts taken from panic are rarely recovered.
- 4 to 8 weeks. The comfortable zone for most residential trades. Enough runway to schedule efficiently, cluster jobs geographically, and say no to bad work.
- 8 to 12 weeks. Strong. This is where you start raising prices instead of adding crew, see how to run a profitable service business.
- Over 12 weeks. A risk, not a trophy. Customers who wait three months cancel, competitors quote two weeks out and win, and your reviews start mentioning delays.
Seasonal trades read these bands differently. Snow removal and landscaping run their backlog against a calendar window that closes whether or not the work is done, so a July backlog of ten weeks in a business with a fourteen-week season is close to full, not comfortable.
Track the trend, not the snapshot
A single backlog number tells you today. The direction tells you next quarter.
Record backlog weeks on the same day each week. Three consecutive declines mean sales is not replacing delivered work fast enough, and you have roughly your current backlog length to fix it before the crew notices. Three consecutive climbs mean you are selling faster than you can deliver, which is the moment to raise prices, extend lead times honestly, or hire, in that order.
Pair the trend with two supporting numbers:
- Backlog by service line. A healthy total can hide a dead segment.
- Backlog concentration. If one customer is 60% of it, that is not backlog, that is a single point of failure.
What to do with the number
Backlog is falling. Convert stalled quotes before you cut price, see how to follow up on a quote. Reactivate past customers on maintenance work. Fill gaps with recurring contracts, which are the ballast that keeps backlog from collapsing, see recurring maintenance contracts.
Backlog is rising past comfortable. Raise prices on new quotes before you hire. Price is the fastest lever and the only one that improves margin instead of adding fixed cost. Quote longer lead times honestly rather than promising dates you will miss.
Backlog is stable and healthy. Use the runway to be selective. Decline the job types your post-job profitability review shows underperform, and cluster the calendar to cut drive time.
Backlog and cash are not the same thing
A full backlog with no deposits and 60-day payment terms will still put you short on payroll. Backlog measures whether the crew has work. Cash measures whether you can pay them for it.
Protect both: collect deposits on larger jobs, see how to collect a deposit before work, bill in stages on long jobs with progress billing, and watch collection speed with days sales outstanding. A business can go under with a record backlog. It happens every year.
FAQ
What is a good backlog for a contractor? Four to eight weeks of crew capacity suits most residential service businesses. Commercial and new construction routinely run longer because contracts are larger and lead times are already long. The right target is the one that keeps your crew busy without pushing customers past what they will wait.
Is backlog measured in dollars or hours? Track both, but manage by hours. Dollars tell you revenue coverage; hours tell you whether the crew has work next month. Since crew time is the constraint that actually limits delivery, hours drive the decisions.
Do open quotes count as backlog? No. An unsigned quote is pipeline. Multiply it by your historical close rate if you want a forecast, but never add it to backlog, or you will hire for work that never arrives.
How does backlog differ from work in progress? Work in progress is the portion of started jobs not yet completed or billed. Backlog is broader: everything signed and undelivered, started or not. WIP is a subset of backlog.
How often should I check it? Weekly, on a fixed day. Backlog moves as jobs close and quotes get accepted, and a monthly check catches a decline four weeks too late to respond calmly.
What if my estimates of labor hours are unreliable? Then your backlog number is unreliable in the same direction. Fixing estimate accuracy fixes backlog accuracy for free, which is why comparing estimated to actual hours on every finished job pays off twice.
Getting started with CRMb
Backlog is hard to measure because the pieces normally live in different places: accepted quotes in email, the schedule on a whiteboard, actual hours in a timesheet app, and estimates in someone's head.
CRMb keeps them on one record. Quotes move through a pipeline you can view as a board, table, or timeline, so accepted work is unambiguous. Scheduling and Capacity show what the crew can absorb week by week. Crew hours punched in Punch land on the job automatically, so remaining hours on in-progress work are a real figure rather than an estimate of an estimate. Reports then show margin on those real punched hours, so the backlog you are protecting is backlog worth having.
Start a 14-day free trial at crmb.io and calculate your backlog this week. Most owners find it is shorter than they assumed.