How centralised routing gets each enquiry to the right location immediately, without head office standing in the middle of every call.
Key takeaways
Most multi-location businesses don't lose enquiries outright — they misplace them. A customer near one location fills out a form built for the network as a whole, or calls a number that happens to ring at head office, and the enquiry lands with whoever is free rather than whoever is closest. By the time it's forwarded, re-explained or simply dropped, the customer has often already called somewhere else.
The pattern repeats across phone, web forms and messaging, and it gets worse, not better, as a network grows. Every additional location adds another place an enquiry can land in the wrong hands, and no manager wants to spend their morning re-routing calls that were never meant for them.
Effective routing starts with reading the enquiry itself — the postal code on a form, the area code on a call, the location named in a message — rather than assuming the channel someone used tells you where they belong. AI is well suited to this because it can parse unstructured input, like a free-text form or a voicemail transcript, fast enough that the routing decision happens before the customer notices any delay.
The system then delivers the enquiry straight into the queue of the correct location's manager, with the same urgency as if it had come in on their own line. Nothing about the customer's experience should reveal that routing happened at all.
Centralising routing is easy to mistake for centralising control, and that mistake is what makes managers resist it. The fix is a design choice: the routing layer only decides where an enquiry goes, not how the location handles it afterward. The manager still calls back, quotes and closes the deal exactly as they would if the enquiry had reached them directly.
That distinction matters for buy-in as much as for customers. A manager who sees the system as head office deciding who gets their leads will work around it. A manager who sees it as leads reaching them faster than before will actually use it.
No routing rule catches everything — a postal code is missing, a caller mentions the wrong city, a form gets filled out generically. The difference between a good system and a fragile one is what happens next. A good fallback puts the enquiry somewhere a human notices quickly, rather than letting it sit unclaimed in a queue nobody owns.
Building that fallback in from the start is cheaper than fixing it after a customer complains nobody called them back, which, without one, is usually how the gap gets discovered.
A 30-minute call is enough to tell you whether AI pays for itself here.