A field service platform will give you hundreds of reports. Almost none of them answer the question an owner asks on Monday morning: did more jobs get booked last week than the week before, and why?
That question is the whole point of any effort to fix the phones, follow up faster or wake up old customers. If you cannot answer it in five minutes, you cannot tell a working system from an expensive one. This post lays out a one-page report with six numbers, where each comes from, and how to read it. It is the measurement layer under the where-to-start guide for contractors.
Two ground rules. First, pull the same six numbers the same way every week, even if the definitions are imperfect. A consistent rough number beats a precise one that changes meaning. Second, report by branch if you have more than one. A company-wide average hides the location that is dragging it down.
The six numbers
1. Inbound leads by source
What it is: every new request for work that reached you last week, split by where it came from: phone, web form, text, chat, Google Business Profile, lead marketplaces, referrals.
Why it matters: it is the denominator for everything else. If leads fell 20% because an ad campaign paused, booking rate did not cause that.
Where to pull it: your call tracking platform for phone leads, your website form or CRM for web leads, and your field service software's lead or request list for the rest. Housecall Pro's lead conversion reporting, for example, breaks leads down by lead source, though per its help documentation it is limited to Pipeline users. Jobber tracks inbound requests as their own record type. The practical problem is almost always source tagging. If your CSRs pick "other" or leave it blank, fix that before you build anything else.
2. Answer rate
What it is: of the calls that came in, the share a person answered.
Why it matters: this is the leak at the top of the funnel. Published figures vary widely because they measure different things. Invoca's July 2026 home services benchmark put the share of callers who spoke with a person at 52% across all calls, rising to 65% when calls under 15 seconds were excluded. CallRail's January 2025 figure for home services missed calls was 14%. Do not compare yourself to either. Compare yourself to last week.
Where to pull it: your phone system or call tracking report. Pull it by branch and by hour if you can. Overnight and 8 a.m. are usually where it breaks.
3. Response time to web leads
What it is: minutes between a web form or chat submission and the first real human or system response.
Why it matters: speed is the part of a web lead you control. A Valve+Meter secret-shopper study of more than 450 home services companies, reported by Professional Remodeler in 2018, found that a company that had not responded within six hours had a 75% chance of never responding at all. It is an old study from a performance marketing company, and a small sample, but it points at the right problem: a lot of web leads are never answered.
Where to pull it: compare the form timestamp in your CRM with the first outbound call, text or email. Most field service tools log the lead's creation time and the first activity. If yours do not, submit a test lead every Monday and time the response yourself. That crude method is better than nothing.
4. Booking rate
What it is: of the calls and leads that were real opportunities, the share that ended up as a scheduled appointment.
Why it matters: this is the number your CSRs control. ServiceTitan's analysis of more than 3,000 trade businesses (June 2022 data) put the typical shop at 42%, and ServiceTitan says its software can show a call booking rate per CSR. Our call booking rate benchmarks post goes through how to read that number against your own size and trade.
Where to pull it: your field service platform, if your CSRs classify calls consistently as opportunity or not. Without that classification the number is noise, so audit a sample of 20 calls a month. Jobber's quotes report, found under Insights then Reports then Work Reports, shows quote status and the share converted to jobs, which is a quote-based version of the same idea. The vendor definitions differ, so write down which one you use.
5. Booked jobs and booked revenue
What it is: the count of jobs scheduled last week and their booked dollar value, split by new customers and existing ones.
Why it matters: this is the number that pays the bills. Report booked, not completed, because completed lags by days or weeks and tells you about last month's decisions.
Where to pull it: the job or appointment list in your field service software, filtered by created date and by status. Include the customer type and the lead source, so you can later answer which marketing channel produced booked revenue and not just leads.
6. Recovered jobs
What it is: jobs that were booked because of a recovery step rather than a first-touch answer. Count three buckets separately:
- Missed-call recoveries: callers you missed who ended up booked after a callback or text.
- Estimate follow-ups: open estimates that closed after follow-up. See unsold estimate follow-up.
- Reactivations: past customers who booked after an outreach campaign. See how to reactivate old customers.
Why it matters: this is the line that shows whether a recovery system is paying for itself. It is also the one most often missing, because it requires tagging.
Where to pull it: add a "recovery" tag or source value to the job, set automatically when the booking follows a recovery message or callback. Then filter by it. If your system cannot tag automatically, a CSR checkbox works as long as someone owns the habit.
What the page looks like
Keep it to one screen. A table with this week, last week and four-week average per number is enough.
| Number | This week | Last week | 4-week avg |
|---|---|---|---|
| Inbound leads (by source) | |||
| Answer rate | |||
| Web lead response time (median) | |||
| Booking rate | |||
| Booked jobs / booked revenue | |||
| Recovered jobs (by bucket) |
Use median, not average, for response time, because one lead that sat over a weekend will wreck an average.
How to read it
The numbers form a chain, and the break point tells you what to fix:
- Leads steady, answer rate down: a coverage or routing problem. Look at which hours and which branch.
- Answer rate fine, booking rate down: a scripting, training or scheduling problem. Listen to ten calls.
- Web response time long: web leads are going to a shared inbox nobody checks. Fix the notification path.
- Booked jobs flat while recovered jobs rise: your recoveries are replacing first-touch bookings, not adding to them. Check whether the main funnel got worse.
- Booked revenue down with jobs up: you are booking smaller tickets. Check whether you are filling the schedule with low-value work.
Keep it honest
A few habits keep the page trustworthy:
- Freeze definitions. Write each number's definition at the bottom of the page and change it only with a note.
- Set a baseline. Run the report for four weeks before you change anything, so you know what normal looks like.
- One owner. Someone produces it every Monday, and someone else reads it. If it is never read, stop producing it.
- Be honest about attribution. If you launch three things in a month, you cannot credit one of them. Change one variable at a time when you can.
If you would rather not build this by hand, weekly booked-job reporting is one of the eight components of our Booked-Jobs Recovery System, and wiring it into your field service software is a large part of the first 30 days. If you want to talk through which of the six numbers you cannot pull today, book a strategy call. Either way, the stack underneath matters, so see how AI connects to ServiceTitan, Housecall Pro and Jobber for what each platform will and will not expose. Platform features and plan limits change, so check your own vendor's current documentation.