Switching answering services without dropping a call: the migration checklist
Most contractors stay with a service they dislike because changing feels risky. Done in the right order, the switch takes an evening and nothing goes dark.
We hear the same sentence from contractors who are unhappy with their current answering service: “I know it is not great, but I cannot risk the phone going down.”
That fear is reasonable and almost always misplaced. The risk is not in switching — it is in switching in the wrong order. Done properly, both services are live at once for a week and nothing is ever uncovered.
Here is the sequence.
Before you touch anything
1. Get your baseline. Pull last month’s call detail record from your carrier and count answered calls, missed calls and after-hours volume. Without this you cannot tell whether the new service is better; you will be comparing a feeling to a feeling. How to run that in twenty minutes.
2. Read your contract. Specifically: notice period, minimum term, and whether the number you publish belongs to you or to them. That third one matters more than the other two combined.
3. Find out who owns your number. If your published number is a tracking number issued by your current provider, cancelling means losing it — and it is on your trucks, your Google Business Profile and your yard signs. If it is yours, this whole exercise is a forwarding change and you are fine.
If they own it, you have two options: port it out (allow two to four weeks, and never cancel before the port completes), or keep the old service running on a stub while you migrate your published number elsewhere. Do not let anyone talk you into cancelling first.
Write down what your current service actually does
Most shops have never documented this, which is why migrations feel scary — the knowledge lives in a script somewhere at a vendor.
Ask your current provider for a copy of your call script. Then confirm on paper:
- Your greeting, word for word
- Your service area boundaries
- Your diagnostic fee and your waiver rule
- What counts as an emergency, and what happens when one comes in
- Who gets called or texted for an emergency, and in what order
- Your hours, including holidays
- Anything you have told them never to say
This document is the actual product you are buying from any provider. Once it exists, switching becomes configuration rather than reinvention. It is also worth having regardless — if your current service loses it, you will be rebuilding from memory.
The overlap week
This is the part that removes the risk.
Day 1 — configure the new service. Load the script above. Do not point any live traffic at it yet.
Day 2 — test it yourself. Fifteen calls minimum, deliberately awkward: interrupt it, give a partial address, ask the fee, describe an emergency, go quiet. Whatever you are buying — human or software — this is when you find out what it does off the happy path.
Day 3 — send it a slice. Forward one segment only. The best first slice is after-hours on weeknights, because those calls are currently going to your old service or to voicemail, and the comparison is clean. Your daytime coverage is untouched, so the downside is bounded.
Days 4 to 10 — run both. Old service keeps daytime and weekends, new service takes weeknight evenings. Compare like for like: how many calls, how many booked, how many complaints.
Day 11 — decide with numbers. Bookings per hundred calls is the metric. Not how it sounded to you; what it produced.
Day 12 — move the rest, then cancel. In that order. Give notice only once the new service is carrying full load and you are satisfied.
Nothing in that sequence has a moment where calls are uncovered, and every step is reversible with a forwarding change.
What to watch during the overlap
Four things tell you most of what you need to know.
| Signal | What it means |
|---|---|
| Bookings per 100 calls | The only number that matters. Everything else is diagnosis. |
| Calls under 20 seconds | Hang-ups. If these rise, the greeting or the wait is wrong. |
| Emergency handling | Did anything urgent get treated as routine? One failure here outweighs a lot of convenience. |
| Repeat callers | Are existing customers being recognised, or asked to re-explain everything? |
Two mistakes worth avoiding
Cancelling before the new service has run a full week. Every contract has a notice period, and none of them refund you for cancelling early. There is no saving worth an uncovered Saturday.
Migrating in peak season. June and July are the worst possible weeks to change anything about your phone. If you are drowning right now, get through it and migrate in September. Coverage bought during a heat advisory tends to be badly configured, because nobody has time to test it. Peak season is a capacity wall, not a slope — do not add a migration to it.
If you are switching to something automated
One extra step: run the integration round trip before you send it live traffic. Book a test job, then open your field service software and check every field arrived. Do it twice with the same phone number to see whether it creates a duplicate customer record.
And keep the trial narrow to begin with. The strongest structure is after-hours only, because it cannot touch a call you are already answering — the worst case is that your existing setup carries on unchanged.
That is how ours works, if you want to test it: forward your weeknight evenings for two weeks and count what appears on the schedule. The demo line is (260) 264-9632 if you would rather hear it first.
Related reading: How to forward your business line · HVAC answering service cost in 2026 · Twelve questions to ask before you buy
Stop losing the calls behind this problem.
Forward your line tonight and see what a 24/7 answer rate does to next week’s schedule. Fourteen days free, no contract.
