The Algorithm, the Trade-offs, and the Pricing
If your reps spend an hour every morning sequencing the day's visits in Google Maps, you don't have a planning problem — you have a routing problem. Multi-stop route optimisation is the difference between "this should work, probably" and "this is the shortest path through all 12 stops, factoring traffic, calendar windows, and territory rules." This piece explains how it works, why most field-sales tools get it wrong, and how Velocity X handles it differently as a buy-once template.
What Multi-Stop Route Optimisation Actually Does
The problem: a rep has N stops to make today. Find the order that minimises total drive time, subject to constraints — appointment windows, lunch break, territory boundaries, vehicle type. Solving this exactly for N > 15 is computationally hard (it's a variant of the Travelling Salesman Problem). Solving it approximately for N up to 50 is fast and good-enough using a few well-known heuristics: nearest-neighbour, 2-opt, simulated annealing, or Google's OR-Tools.
What "optimised" looks like in practice: a 12-stop day that would take 6 hours of drive time with naive sequencing drops to 4.5 hours with good routing. The rep gets back 90 minutes of selling time. Across a team of 10 reps over a year, that's roughly 3,600 hours of recovered time. Real money.
The Naive Approach Most Tools Use
Many field-sales tools advertise "route optimisation" but implement it as: sort stops by distance from current location. This is the nearest-neighbour heuristic and it produces routes that are 15-30% longer than optimal in typical territories. The math is on the user's side here — anything beats no routing — but "we have route optimisation" is technically true while leaving substantial time on the table.
The signal a tool is doing it properly: does the route change order when you swap two stop priorities, or just re-rank by proximity? Does it factor traffic at the time-of-day window? Does it respect appointment slots (a 2pm booking can't be moved to 11am)? If the answer to any of those is no, the "optimisation" is really just sorting.
How Velocity X Approaches It
Velocity X uses Google Maps' Routes API for the underlying distance/duration matrix, then runs an OR-Tools-backed solver on the constrained TSP. Stops are ordered by total drive time, subject to: appointment time windows, lunch breaks, territory boundaries, and rep-specified priority pins (lock stop A first, lock stop B last). The solve typically runs in under 400ms for a 25-stop day, fast enough to re-run live when the rep marks a stop done early and wants to recompute the rest of the day.
Live ETAs update every 60 seconds based on current traffic. When traffic hits a planned segment, the dashboard surfaces a "route change suggested" prompt that re-solves the remaining stops in two seconds. Reps can accept or override.
The Subscription vs Buy-Once Question for Routing
Most field-sales tools that include route optimisation charge per-seat per-month — Badger Mapping at $58/user/month, Optiway at $30+/user/month, LeadPlotter at $35/user/month. For a 10-person team, that's $300-$580/month or $3,600-$6,960/year — every year, forever, with prices that creep up.
Velocity X bundles the same routing engine into the buy-once template ($4,995 small / $9,995 medium / $24,995 large). The Google Maps API costs that drive the actual route solves run on your own Google Cloud account at usage-based pricing — typically under $5/month for a small team, scaling sub-linearly with usage. After year one, you're paying single-digit dollars per month for the routing infrastructure that the SaaS alternatives charge hundreds for.
Frequently Asked Questions
How does live traffic re-routing work?
The dashboard polls Google Maps' real-time traffic layer every 60 seconds for the planned segments. When a segment's expected drive time exceeds the planned time by more than 8 minutes, the system surfaces a re-route suggestion. The re-solve runs OR-Tools against the remaining stops + the rep's current GPS position. Total round-trip latency: under 3 seconds.
Does it work without internet?
Routes get cached locally on the rep's device once the day is planned. If signal drops mid-route, the rep keeps the cached sequence — they just don't get live traffic adjustments until signal returns. Stop check-ins and notes are queued locally and sync when bars return. This is the "offline field mode" pattern.
How many stops can it handle in one route?
Practical ceiling is around 35-40 stops per rep per day. The solver handles more in theory but solve times grow non-linearly past 50, and real reps physically can't make 50 stops in a workday. For multi-day routes (e.g. quarterly territory tours), the solver chunks into per-day legs.
What about delivery routing vs sales routing?
Same underlying algorithm. Velocity X is tuned for sales (variable stop duration, customer-facing windows, AI follow-up integration) but the routing primitive is industry-agnostic. Delivery companies can use it with minor configuration changes.
Can I override the optimal route?
Yes. Reps can pin stops to specific positions in the day, drag-reorder manually, or accept/reject re-solve suggestions. The optimisation is a recommendation engine, not a mandate.
What's the comparison to Google Maps' built-in multi-stop?
Google Maps maxes out at 10 waypoints, doesn't respect appointment windows, doesn't account for territory rules, and doesn't sync to your CRM. Velocity X uses Google's routing primitives underneath but adds the constraints that make it actually usable for field sales.
The Bottom Line
Multi-stop route optimisation in field-sales tools ranges from "barely useful sort" to "real OR-Tools solver". The difference shows up in your reps' actual day. Velocity X bundles the real version into a buy-once template, eliminating the per-seat SaaS bill that competitors charge for the same engineering work.