Most multi-location businesses route leads with a round-robin or a zip code lookup. Both are guesses. Both leak about 30% of pipeline. Here's what AI-powered routing actually does differently and why it matters.
Why round-robin fails
Round-robin treats every lead the same. Lead arrives, system assigns to the next location in rotation. The system has no idea whether that location is the best fit for that lead's intent, segment, or geography. It just rotates.
The result: a high-intent lead asking about a specific service ends up at a location that doesn't specialise in it. A geographically obvious lead ends up at the wrong location because rotation didn't care. A reactivation candidate from an old database goes to a fresh sales rep who has no context.
Round-robin is not routing. It's distribution. There's a difference.
Why zip-code lookup fails almost as often
Zip-code routing is closer to right. A lead in zip 90210 goes to the location that serves that zip. The problem is that zip codes don't capture intent, history, language, or context.
A lead in zip 90210 might be a returning customer who's been shopping a competing service. A different lead in the same zip might be a high-intent first-timer. The two should be treated differently. Zip-code routing treats them identically.
Zip routing is necessary. It's not sufficient.
What AI-powered routing actually does
An AI routing layer reads the full context of a lead before deciding where it goes. Specifically:
Intent classification. What is the lead actually asking about? Is this a price-shopping inquiry, a high-intent buying signal, a support question that wandered into sales, or a reactivation of an old contact?
Segment detection. Which buyer segment does this lead fit? Different segments convert at different locations because the staff, the offerings, or the language match better.
Geographic preference (with override). Zip-code lookup as a default, but the AI can override when the lead's actual stated preference, language, or context points elsewhere.
Historical context. Has this contact been in the system before? Was there a previous interaction at a different location? AI reads the contact's history and routes to continuity rather than starting fresh.
Confidence scoring. When the AI isn't confident, it escalates to a human. Low-confidence routing decisions are flagged for review rather than silently sent to the wrong place.
The pattern that beats round-robin: AI reads the lead, scores intent, segment, and context, then makes a routing decision with rationale. The rationale gets logged so the team can see why the lead landed where it did.
The 30% leak
Across multi-location operations I've audited, the typical pipeline leak from generic routing is 25-35%. The leaks come from three places:
1. Mismatched routing. Lead lands at a location that doesn't fit. Conversation goes nowhere. Lead drops out without anyone noticing.
2. Slow routing. Round-robin or human-routed leads take hours to land. By then the lead has already booked with a competitor or lost interest.
3. Lost reactivation candidates. Old contacts re-entering get treated as fresh leads. Their history is invisible. They get the generic sequence and disengage.
AI routing fixes all three. Routing happens within seconds. The decision is informed by full context. Reactivation candidates are flagged and routed to the right path.
What this looks like in practice
For a 5-location preschool franchise I worked with, the routing layer reads each inbound contact and classifies:
- Age segment of the child (different rooms, different staff specialise)
- Intent (info-seeking vs ready-to-tour vs ready-to-enrol)
- Source attribution (Meta ads vs organic vs referral)
- Geographic preference (with override for stated location)
- Historical context (returning enquirer vs net new)
Each routing decision logs the rationale. The team sees not just where the lead went but why. When the AI gets it wrong (which happens), the rationale lets them fix the rule rather than guessing.
The franchise's enrolment SMS systems, driven off this routing layer, ended up handling 75% of inbound enrolment calls within the first month post-launch.
What this means for your operation
If you have three or more locations and you're using round-robin or simple zip routing, you're almost certainly leaking 25-35% of pipeline at the routing layer. The leak is invisible because you're only counting the leads that arrive, not the ones that disengage after hitting the wrong place.
The fix isn't more aggressive follow-up. It's a smarter front door. Once the routing is right, the rest of the funnel works with the leads it's designed for.
The economics are obvious. Recovering even half of a 30% leak on a multi-location operation is usually larger than the entire engagement cost in the first quarter.
Frequently Asked Questions
Running on a stack that grew by accident?
Tools added one at a time, never architected together. That's the problem I solve. Book 45 minutes and I'll map what moves, what stays, and what makes sense for your operation.
Book a Discovery Call