At one location, call routing is simple: the phone rings in the office, and a CSR answers it. At three locations it starts to break. At eight it is a standing source of lost jobs, because every branch has grown its own way of handling the phone, and nobody can see which way works.
This post covers the four routing decisions a multi-location home services company has to make, the trade-offs on each, and how to tell from your own data whether the setup is working. It sits inside the broader framework in AI for contractors: where to start, and it applies whether your branches are your own openings or acquired brands.
The central versus local question
Companies usually frame this as a choice between a central call center and local offices. In practice you are choosing along several dimensions.
| Central team | Local office | |
|---|---|---|
| Coverage | Easier to staff evenings, weekends and peak hours across many branches | Limited to whoever is in the building |
| Market knowledge | Needs scripts, maps and service-area rules to compensate | Knows neighborhoods, lead techs and local quirks |
| Consistency | One process, one set of numbers | Each branch drifts |
| Dispatch fit | Must see live technician availability for every branch | CSR often sits next to the dispatcher |
| Cost | Shared headcount, but you add a management layer | Duplicated headcount per branch |
We did not find a source we trust that says one beats the other for home services, and we are wary of anyone who claims one does. Most companies land on a hybrid: local CSRs for daytime, a central team for overflow and after hours, and shared rules underneath. What matters is that the choice is deliberate and that you measure it.
The measurement matters because the shape of your call volume decides what you need. ServiceTitan's June 2022 analysis of more than 3,000 trade businesses showed call booking rates falling hard after hours: among large shops from 61% at the peak of the day to 21% after 6 p.m., and among small shops from 26% to 9%. That is ServiceTitan customers only, and the data is from 2022, but the direction is a useful reminder that evenings are a different problem from mornings.
Decision 1: Route by service area, not by which number was dialed
A branch's phone number is a poor routing key. Customers call the number on the truck, the Google listing, the old yard sign or last year's ad, and it often is not the branch that serves their address.
A better rule is routing by the service area of the caller:
- Ask for the address or ZIP code early in the call (or capture it from the web form)
- Match it to a service-area table that maps ZIP codes to branches
- Book the job to the branch that serves it, regardless of which line rang
A central team can do this with a lookup. A local CSR who gets a call from outside the area should be able to transfer or book to the right branch without phone tag. Keep the service-area table in one place and change it in one place. Branches that maintain their own copies end up disagreeing about who covers the town on the border.
Edge cases need a written rule: a customer on a boundary, a commercial account that spans branches, a trade one branch does not offer. Decide them once.
Decision 2: Set an overflow rule with a clock on it
A branch rings out when the CSR is on another call, the dispatcher is at lunch, or a storm tripled the volume. Without an overflow rule, those calls go to voicemail.
A workable overflow rule has four parts:
- A trigger. For example, a call that rings for a set number of seconds, or finds all lines busy.
- A destination. The next branch, a central queue, or an AI or human answering layer.
- Information passed along. The destination should know which brand and branch the caller dialed, so it can greet correctly and book to the right place.
- A booking path. Overflow that only takes a message is a slower voicemail. The overflow destination has to be able to book, or hand off with enough detail that a callback books.
Pick the thresholds from your data. Pull missed calls by hour and by branch. If most misses sit in two windows, the answer is targeted overflow for those windows, not a blanket change.
Decision 3: Treat after-hours as its own program
After-hours is where the booking rates in the ServiceTitan data fall, and it is also where the emergencies live: a no-heat call in January, a burst pipe at midnight. These need rules that daytime routing does not:
- What counts as an emergency, written down, with an on-call technician or escalation path
- What happens to non-emergencies, which are usually best booked for the next day rather than dispatched
- Who owns the follow-up, so the 11 p.m. lead is the first thing a CSR sees at 7 a.m.
- A single number or queue for the whole company, so you do not staff five separate after-hours setups
A central after-hours team covering all branches is often more practical than each branch covering its own. The detail is in after-hours call triage for HVAC and plumbing. Whatever you choose, the system needs to be able to book or create a job record, not just take a message. Our after-hours lead capture guide covers the capture side.
Decision 4: Report per branch, from day one
A company-wide average hides the location that is struggling. Report the same numbers for each branch, every week:
- Calls in, and answer rate
- Overflow volume and what happened to it
- Booking rate
- Booked jobs and revenue
- After-hours leads and how many were booked by noon the next day
The report in the weekly booked-jobs report is built for this. Run it by branch and look at the spread. If one branch answers 90% and another 60%, you have a coaching or staffing question for the second, and possibly a routing lesson from the first. If a branch's booking rate drops when overflow rises, the overflow destination is probably not booking effectively.
Field service platforms offer some multi-location reporting. Jobber's Jobber Central, for example, has a quote tracking view that compares business units by sent quote count and value, converted quote value and conversion rate. That covers quotes, not calls, so you will still need phone data from your call tracking or phone system. Capabilities vary by plan and change, so check your vendor's current documentation and confirm that branch tagging is consistent before you trust any roll-up.
A short example (illustrative)
Take a made-up company with four branches and one central after-hours line. Branch C's answer rate is 62% against 85% elsewhere, and the misses cluster between 7:30 and 9 a.m., when the dispatcher takes check-in calls from the techs. The fix is overflow from Branch C to the central team during that window, with a booking script and access to Branch C's schedule. It is not a new CSR, and it is not a new phone system. The numbers are invented for illustration, but the pattern, where misses pile up in a window tied to dispatch workload, is what the by-hour report is meant to surface.
Where AI fits, and where it does not
AI can take overflow and after-hours calls, ask the service-area questions, book into the right branch's schedule and send the result to the CSR who owns it at opening. It is a good fit for the overflow tier and a poor substitute for a clear routing rule. An AI that books into the wrong branch because the service-area table is wrong does so with more confidence and at higher volume. Fix the rules first. Our guide to AI receptionists for contractors covers what to test before you sign, and standardizing call handling across acquired brands covers the same problem from the platform side.
If you run three or more locations and want help working out which routing setup fits, the Booked-Jobs Recovery System is a fixed-scope engagement, live in 30 days and measured in booked jobs, with missed-call recovery, after-hours qualification and weekly booked-job reports among its components. To talk through your branches, book a strategy call.