Your phone can show operational and still lose the call.
That is the trap tonight.
ServiceTitan opened a major incident at 8:38 p.m. Pacific on August 10 after outbound calling partially failed in Phones Pro and Contact Center Pro. At 9:10 p.m., ServiceTitan said its unnamed third-party provider had deployed a fix, moved both calling components back to operational, and began monitoring the recovery. But at 12:19 a.m. Pacific on August 11, the incident was still open. If your shop depends on those products for customer callbacks, sales calls, or dispatch, the safe move is simple: test the line before the morning board fills up.
There is a second signal behind the warning. Bandwidth opened an incident for its Los Angeles data centre 27 seconds before ServiceTitan opened its incident, reporting degraded U.S. outbound SIP calling and warning that calls routed through LAX might not complete. Bandwidth was still investigating at 12:19 a.m. Pacific. ServiceTitan did not name Bandwidth as its provider, so the matching timing and symptom are evidence of a likely connection, not proof. Either way, the operator consequence is the same: recovery inside ServiceTitan does not guarantee that every outbound route is stable yet.
The failure sits between the software and the call
This was not a confirmed shutdown of the whole ServiceTitan system. The official incident names two components only: outbound calling in Phones Pro and outbound calling in Contact Center Pro. ServiceTitan now marks both operational while it monitors the provider fix. The separate Bandwidth incident is narrower still, tied to U.S. outbound SIP traffic routed through its LAX data centre. That means one shop may place clean calls while another still sees failures, depending on product, route, number, and carrier path.
That selective failure is what makes voice outages expensive. A dead app is obvious. A call that appears to leave but never connects can sit unnoticed until the customer books someone else, the crew waits for access, or the office discovers a hole in the callback list. The danger is not drama. It is quiet lost work.
Operational is not the same as proven.
Who gets hurt first when outbound calls fail
The first exposure sits with service shops using ServiceTitan's phone products as the main route for urgent callbacks, after-hours booking, sales follow-up, and dispatch changes. A missed estimate call can turn into an empty slot. A failed access call can leave a crew at a locked gate. A callback that never connects can look like the customer ignored you when the failure was upstream all along. Those are small failures one by one, but they stack quickly when the office assumes the system is healthy.
Do not overread the alert. ServiceTitan has not reported that inbound calls, scheduling, invoicing, payments, or the rest of its software are down. Bandwidth has not reported that every U.S. outbound call is failing. Canadian shops are not named in the upstream incident, but voice routing can cross borders invisibly, so a Canadian office calling U.S. numbers should test instead of assuming geography protects it.
Do this before the morning board fills up
This is a fifteen-minute control check, not a migration project. Put one person in charge and close the gap before crews start moving.
- Place test calls. Call at least one mobile number and one landline or alternate carrier from both Phones Pro and Contact Center Pro if you use both.
- Use another line for urgent work. Move same-night callbacks, access calls, and first dispatch calls to a mobile phone or separate business line until the official incidents close.
- Audit the failure window. Review outbound attempts from 8:38 p.m. Pacific onward. Flag calls with no answer, short duration, or no confirmed customer contact.
- Retry with context. Tell customers you are confirming a call after a phone-service interruption. Keep it plain. Do not blame them for a missed connection.
- Record proof. Note the test time, destination number type, result, and alternate route used. If the failure affects booked work, preserve the call log for the vendor ticket.
- Watch the source pages. Stand down only after your own test succeeds and both official incident pages show full resolution.
What OPS expects next
- High confidence: ServiceTitan will likely close its incident before or near the upstream carrier because its two components already show operational. Keep testing until both pages agree.
- Medium confidence: Any remaining failures will be selective rather than systemwide because the active upstream warning is tied to outbound SIP traffic through one data centre.
- Watch item: Other phone products using the same route may report similar symptoms if the Bandwidth incident persists. No broader product impact is confirmed now.
If both incident pages close and your tests pass, return to the normal line. Still finish the callback audit. Recovery stops new failures; it does not recover the customer who already heard silence.
Sources and the duplicate check
This alert is based on the live ServiceTitan incident and the simultaneous Bandwidth LAX incident. OPS previously covered software selection in the 2025 construction-tech landscape. This is different: it is a dated availability failure with an immediate call-routing move, not another software comparison.
Test the line. Route around the weak link. Then get back to work.
