# Route Optimization and Delivery Route Planning > The decision layer that builds routes against distance, time and capacity. For FMCG field teams and logistics: plan it, push it to the field, measure it. *Source: https://nextgeography.com/en/products/route-optimization/ · Language: en · Updated: 2026-08-25T11:04:56+00:00* *Site index for agents: https://nextgeography.com/llms.txt* A pale grey map of Istanbul with a pink route line running through numbered stops. PRODUCT · ROUTE OPTIMIZATION ## Which point, in which order, in which vehicle? The decision layer that builds a route against distance, time and capacity. For FMCG field teams and logistics distribution: it produces the plan, pushes it to the field and brings back what actually happened. USE CASES MEASURES TRAVEL MODES 01 · THE PROBLEM ### Routes get built, but nobody measures them. The same loss in FMCG field work and in distribution: excess kilometres, missed visits, late deliveries. #### The route is built from habit The order follows whatever the representative is used to; nobody computes the alternative. #### Capacity is not in the plan Vehicle volume, delivery window and working hours sit outside it. #### Territories are unbalanced One representative carries forty points, another eighteen; the workload is never measured. #### Drift is invisible The gap between the planned route and the visit that happened is not reported. Kilometres are the most invisible line in the budget: nobody asks at the planning stage, and everyone sees it at the end of the month on the fuel bill. 02 · THE APPROACH ### A route is an output; its input is data. The optimisation engine reads four layers together and produces a route you can actually run. A pale grey map of Kartal and Maltepe with a cloud of pale blue sales points; a pink route line follows the road network through them and stops at ring-shaped markers. The sales points in frame, and the visit order chosen through them. #### Point data Sales point, dealer or delivery address, tied to a coordinate and verified. #### Constraint set Vehicle capacity, delivery window, working hours, visit frequency. #### Network and time Distance and duration over the real road network, traffic profile included. #### Optimisation engine Order and assignment produced against a total distance, time and cost target. Output: a day-by-day visit order for every vehicle and every representative, estimated arrival times, total distance and time — on the map and in a table. 03 · USE CASES ### One engine, two different jobs. A representative's week and a vehicle's day out of the depot come from the same model. A pale grey map of Küçükçekmece and Bahçelievler with two routes: a pink round spread across a wide area on the left, and a compact navy round leaving a square navy depot marker on the right. Two rounds on one ground: visits spread across a territory, and a delivery round out of a depot. FMCG #### The representative's day, put in order. #### Planned by visit frequency Each sales point's frequency — weekly, fortnightly — becomes an input to the plan. #### A route per day The representative's week becomes a visit list split across days. #### Visit duration counts Average time spent at a point is included in the route. #### Runnable in the field The route lands in the mobile app; the representative checks in and runs the survey on site. LOGISTICS #### Every vehicle leaves the depot with its own route. #### Vehicle assignment Orders are distributed to vehicles by capacity; each vehicle takes its own route. #### Delivery window The customer's accepted hours enter the model as a constraint. #### Multi-depot network With more than one depot, an order is tied to the most suitable one. #### Return and second run The return to depot and a second run within the day are both calculated. 04 · TERRITORY AND DEPOT ### Territory before route: balance the workload. The field is divided into balanced, contiguous territories against the number of representatives and vehicles. A pale grey map of Kadıköy with three nested drive-time areas around a square depot marker; addresses inside them are deep blue, those outside are pale. Drive-time contours out of one depot; the addresses beyond them stay pale. #### Balanced workload Territories are equalised on point count and visit load. #### Contiguous territory Single-piece territories, not fragments — the transit kilometres between fragments disappear. #### Headcount from capacity How many representatives or vehicles each territory needs is calculated. Territories change once a year; routes change every week. Both come out of the same model. #### Is the depot in the right place? Drive-time analysis measures the reach of your existing depots and transfer points. #### Coverage by time The area reachable in fifteen, thirty, forty-five and sixty minutes, computed on the real road network. #### Points left outside Addresses that fall outside the target delivery time are listed. #### New depot scenarios Candidate depot locations are compared on the coverage they add. #### The same method fits a branch A dark store, a transfer hub or a dealer uses the same model. 05 · THE PRODUCT SCREEN ### Route, distance matrix and reach analysis on one screen. A pale grey map of Şişli with a web of thin blue lines joining the stops to one another; a pink route runs over it through numbered stops. Every link between the stops, and the order chosen, on one map. #### Stop list Reorder by hand, or optimise. #### Travel mode Car, on foot and by bicycle; the traffic profile can be switched on. #### Distance matrix A kilometre and minute table between points, exportable. #### Reach rings Ten, twenty and thirty minute reach areas, on the map. The values in the screen images are illustrative. 06 · IN THE FIELD ### Do not leave the plan on a screen; put it on the team's phone. The Sahanın Gücü mobile app carries the route into the field and brings back what happened. A pale grey map of Avcılar and Beylikdüzü with a wide pale pink ribbon following the road network; the addresses inside it are pink, those outside pale blue. The addresses that fall inside the buffer along the route are marked. #### Filter along a route Draw a route between two points, give it a buffer, and every point along it is listed. #### Visit plan The optimised order reaches the team with its date and the person assigned. #### Check-in and survey The visit is verified by the location it happened at; field data is written onto the point. #### Planned against actual Completed visits flow back to the centre live; drift is reported. 07 · WHAT WE MEASURE ### A saving is not a claim; it is a measurement. These are measured before and after during the pilot; the saving figure comes out of your own data. #### Total kilometres Distance covered per plan and per vehicle. #### Route duration Time on the road separated from time spent at the point. #### Visit completion How many of the planned visits actually happened. #### Cost per point The share of fuel, time and staff cost that falls on each point. #### On-time delivery The rate of compliance with the delivery window. #### Workload balance The spread of load between representatives and vehicles. The pilot runs four to six weeks: the first two measure the current state, the rest compare it against the optimised plan. 08 · SETUP AND INTEGRATION ### It installs beside your systems, not in place of them. #### Order / visit data A daily transfer from ERP, CRM or Excel. #### Address verification A free-text address is tied to a coordinate at building level. #### Optimisation Route and assignment produced together with the constraints. #### Write-back The result returns to your system and to the mobile app over the API. Cloud Next Geo hosts it; no separate server is needed. On-premise It runs on the organisation's own infrastructure. REST API Scoring, geocoding, routing and distance endpoints are ready. Setup averages two weeks including data preparation, and starts with a pilot territory. NEXT STEP ### Let's measure your routes together. We start with one pilot territory and your existing visit or order data. The first two weeks measure the current state; the saving figure comes out of your data.