Stop Routing by Location: Sort Two-Clinic Calls by What the Caller Actually Needs

The complaint isn't really about routing

When a clinic owner says "my Location A desk is covering Location B," the usual fix attempt is to change the routing tree. Add a menu. Change the ring order. Set a fallback. Then a month later the same coordinator is fielding the same calls for the other site, and everyone concludes the phone system is bad.

The phone system usually isn't the problem. The problem is that the routing was built around buildings, and most of your calls have nothing to do with buildings.

A caller asking what a consult costs, whether you offer a service, what your hours are, or whether they can move Thursday to Friday does not care which desk answers. They care whether the person answering can finish the request. If your routing sorts those calls by geography, you've created two queues that each need full coverage at all times — and whichever desk is quieter that hour absorbs the overflow. That's not unfairness in the rotation. That's a design that guarantees spillover.

Sort your calls into two buckets

Before touching a single setting, spend a week writing down what people actually call about. You don't need software for this. A notepad at each desk and a tally is enough. Then split every reason into two columns.

Location-neutral. Anyone with access to your schedule and your price list can handle it. New consult inquiries. "Do you treat X?" Hours and parking. Rescheduling. Confirming an appointment. Financing questions. Intake forms not received. Callbacks on a missed call. Requests for a call back from a provider.

Location-bound. Genuinely requires someone physically at that site or with that site's context. "I'm in your parking lot and the door is locked." "I left my sunglasses in room two." "I'm running fifteen minutes late for my 2:00 here." A post-treatment concern that needs the provider who performed it. Anything involving a package or balance recorded only at one site.

Most clinics are surprised by the ratio. The location-bound column is short. It's also the column that genuinely justifies interrupting a desk mid-check-in.

Build the shared layer first

Once you've sorted, the design writes itself. Location-neutral calls should never be routed to a location at all. They should hit one shared layer that answers every time, regardless of which site is busy, which coordinator is in a treatment room, or which number the patient dialed.

That shared layer can be a person — a coordinator who isn't assigned to either front desk and whose only job is the phone. If you have the volume and the payroll room, that's a clean solution, and it removes the spillover argument entirely because there's no desk to spill onto.

If you don't, the shared layer is where AI answering earns its place. This is the bucket of calls that is repetitive, scriptable, and doesn't require anyone to be standing in a specific building: answer the question, check real availability, book or reschedule, capture the name and number, log it. SmartSyncLink handles this layer across both locations at once, which matters more than the per-call handling — it means the shared inbox and calendar are one view, not two you have to reconcile at the end of the day. When a caller says "whichever location can see me sooner," that's answerable in one step instead of a transfer.

Location-bound calls then go where they belong: directly to that site, with a short ring and a clear fallback. Volume on those lines is low enough that a desk can actually pick up.

The number a patient dials is not your routing decision

A common mistake is collapsing to a single public number to simplify routing. Don't. Each location should keep its own local number on its own Google Business Profile, its own directory listings, and its own signage. Local presence matters for how people find you, and a shared number attached to two profiles creates problems you don't need.

Keep both numbers public. Route both into the same shared layer. The dialed number becomes useful data — it tells you which site the caller associates with, which is a reasonable default when they don't state a preference — but it stops being the thing that decides who picks up.

Measure load by minutes, not by call count

If you want to know whether coverage is actually lopsided, count handling time, not calls. Twelve two-minute reschedules is a lighter morning than three fifteen-minute consult conversations with a treatment history. Front desks feel the minutes; owners look at the call log and see something that appears balanced.

Track, per site: calls answered, calls that required a transfer, and rough minutes on the phone. Review it monthly. If one site is consistently carrying more minutes on location-neutral work, your sorting is wrong somewhere — a call reason is landing in the wrong bucket.

Write down one tiebreak rule

Every two-location clinic eventually hits the ambiguous call: new patient, no location preference, wants the first available appointment. Without a written rule, coordinators improvise, and the improvisation always favors the site the person answering works at.

Pick a rule and put it in writing. First available slot. Closest to the caller's stated area. Preferred provider. Whichever site has the open injector hours you're trying to fill. Any of these is defensible; what isn't defensible is leaving it to whoever picks up. Once it's written, it can also be automated, which removes the question entirely.

Then handle the handoff

The last piece is what happens when a location-neutral call turns out to be location-bound halfway through. The rule that matters: the person who took the call owns it until someone else confirms they have it. No blind transfers between sites, no "I'll have them call you." A logged note with a name, number, reason, and who's responsible — visible to both locations — is what keeps a two-site clinic from losing patients in the gap between desks.

Route by request, not by building. The coverage complaints tend to go quiet on their own.