| Plumber, number on two utes, a trailer and a decade of Google reviews | Assumes adding AI means a new number, so puts the whole idea off for a year | Number stays put, diverts after six rings, often live inside a week | Nothing gets re-signwritten and the Google Business Profile is left untouched, which matters because that number is part of the business details customers and search engines match across the web. |
| Dental clinic, reception drowning at 8am, existing on-site phone system | Calls ring out during the morning rush and patients ring the practice up the road | Overflow only, so reception answers first and the AI catches the queue | The clinic keeps its existing system and its number. We only change what happens on the calls reception cannot get to, which is also the easiest change to reverse if staff hate it. |
| Electrical contractor whose number is on an account in the old IT provider's name | Discovers the ownership problem halfway through a port and loses a fortnight | Ownership sorted first, then a divert while the account name is corrected | The ownership problem surfaces on day one from the bill rather than on port day from a rejection notice, so the AI can go live on a divert while the account name is corrected in the background. |
| Real estate office paying for eight lines it has not used since the office shrank | Keeps paying because nobody wants to risk touching the main number | Full port of the main number, the dead services closed at the same time | This is the situation where porting genuinely earns its keep. Where the monthly saving is real and the mess has to be sorted out eventually, doing it once alongside the AI build beats doing it twice. |
| Physio practice with a 1300 number for bookings and a local number on the door | Treats both as one job and is surprised when the timelines do not match | Two separate plans, local number diverted now, 1300 handled on its own track | Inbound 13 and 1300 numbers follow a different process to a standard fixed line, so we sequence them separately rather than letting one hold the other up. |
| Building company that wants to measure whether the AI is worth it before committing | No way to tell which calls the AI caught, so the decision stays a gut feel | New number on the AI for new marketing, old number diverts alongside it | Clean attribution from day one. After ninety days you can see exactly what the AI picked up and decide on the evidence whether to consolidate onto one number. |