Your answering service says it booked more calls. Your plumber says half the visits on their list need an office call first. Which report should you trust?
Start with the jobs behind the count. A useful booking is a real visit your team can serve. It should have the right work, the right address and a time your team can honor. A full calendar is less useful if your office has to fix each entry.
This guide gives you a way to review a human answering service or an AI receptionist. It does not set a target rate for every shop. Your work, service area and open slots all matter.
1. Agree on what counts as a booking
Ask the service to explain its report in plain words. Does “booked” mean a caller asked for a time? Did someone leave a message? Or is a visit on the schedule, with the details your team needs?
Write down your own rule before comparing reports. For example: “We count a booking when a visit is on our schedule, the customer has agreed to the time and our team can do the work.” Check that the record meets that rule.
Keep an estimate visit separate from a sold job. A homeowner who books a furnace quote has agreed to a visit. They have not yet bought a furnace. Keep completed work and paid bills separate, too.
This is also a reason to check what your software means by each label. Jobber’s lead funnel, for example, tracks new leads, leads with a sent quote and leads with a created job as separate steps. Its count of leads with quotes counts each lead, not every quote sent. Source: Jobber Insights Dashboard
2. Sort calls before you judge the rate
Not every call is a chance to book new work. A supplier may ask about a part. A past customer may want an update. One homeowner may call twice about the same leak.
Keep these groups visible:
- New requests for work you serve.
- New requests for work you do not serve.
- Calls about an open job or quote.
- Other calls, such as suppliers and spam.
- Calls you cannot yet sort.
Keep repeat calls with the request they are about. If a customer calls twice, that does not make two new chances to win a job. A second call may still show a problem worth fixing, such as a reply that never came.
Use the same rules for your office, the answering service and any AI tool. Do not remove hard calls from one report while leaving them in the other. If the address or work type is missing, mark the fit unknown until someone checks it.
3. Look at one made-up plumbing example
The numbers below are fictional. They show how to read a report, not what your shop should achieve.
Over one review period, a plumbing shop logs 30 calls. Ten are about open jobs, suppliers or repeat contact on the same request. That leaves 20 new requests.
Of those 20, four are outside the service area and two ask for work the shop does not do. The remaining 14 fit the shop’s stated rules. Eight of those have verified visits on the schedule. Six do not.
For this example, the booking rate for eligible requests is 8 divided by 14, or about 57%. That is a different question from “What share of all 30 calls became bookings?” Neither figure shows how many visits took place or how many bills were paid.
Now review the six requests without a booking. Perhaps the shop had no open slot. Perhaps someone still needs a price answer. Perhaps the caller declined. Keep an unknown reason unknown. Do not mark all six as lost because of the answering service.
4. Ask what happened after the call
Choose a review period in advance. Include every request in that group, or use a clearly stated sample if the full group is too large. Do not choose only the best calls.
For each request, record:
- What work and service area the caller asked about.
- Whether it fit your rules, and why.
- Whether a visit was truly booked.
- Whether that visit took place or was canceled.
- Whether the customer chose the work.
- What still needs an answer, and who will reply.
Use call notes and job records your team is allowed to review. Avoid passing customer names or phone numbers around in a vendor comparison sheet. A private reference number can link back to your own records.
If a report says “booked” but the office cannot find the visit, ask the service to show the matching record. A caller saying “Tuesday could work” is not enough on its own.
Jobber’s request guide describes reviewing a request before moving it to a quote or job. An assessment can be part of that path, but not every request needs one. Your review should follow the steps your shop actually uses. Source: Jobber Converting a Request to a Quote or Job
5. Compare like work over like periods
Do not compare a quiet week of drain calls with a busy week of large install quotes and credit the whole change to a new service.
Use equal periods with time for visits and quotes to move ahead. Keep call sources, work types, hours and open slots visible. If those change, note it beside the result. Use the same amount of time to check outcomes for each group.
Look at the service bill and your office time as well as bookings. A lower fee may leave more cleanup for your team. A higher fee needs a clear reason in your own records. Use actual costs and paid work when they are known. Do not turn an open quote into revenue.
Small groups can point to questions worth checking. They do not prove one service will keep winning. If you cannot match calls to visits, say so before making a claim about results.
6. Fix the gap the records show
The next step depends on what went wrong:
Wrong-area visits: make the service-area rule clear, then check it on test cases.
Good requests with no open slot: review your capacity and the next step you offer. A new script cannot create crew time.
Missing job details: agree on the few facts your team needs before it can act.
Bookings that need office repair: check how the visit is made and who confirms it.
Open questions with no reply: name the person who owns the next call.
Test records should stay out of your real booking counts. Before changing a live process, agree on what a passing test looks like and what your team will do when a step fails.
Bring one unclear booking to the table
SimpliScale starts with your business, your tools and the step that is getting stuck. That review can cover the first reply, the booking step and the follow-up your team needs. A simple office change may be enough. If AI or a new connection looks useful, its scope and limits need checking before a build.
Bring one example with names, phone numbers and addresses removed. We can review what the caller needed, what the service promised and what your team saw next.
Common questions
What is a good booking rate for an answering service?
There is no single target in this guide. Define the requests your shop can serve, then check verified bookings from that same group. Compare similar work and periods before judging a change.
Should repeat calls count as new leads?
Keep repeat contact about the same request together. A past customer asking for a different job can be a new request. Write down that rule and use it in each report.
Does a booked visit mean we won the job?
No. Track the visit, the customer’s decision, completed work and payment as separate steps.
Can AI replace the review?
An AI label should be checked against the call and job records. Your team still needs a way to review unclear cases and correct mistakes. Any proposed tool or connection needs its own checks.