Can Shipsy's DSD engine handle both pre-sell and van-sell orders on the same truck?

Yes. A single route can carry pre-sell orders for confirmed outlets and van-sell capacity for opportunistic selling. The driver app separates the flows and reconciles them at trip end.

How does the system handle cash-on-delivery and collections at the outlet?

The driver app captures cash amount and method per outlet drop, with trip-end reconciliation against shift totals. Nexa reconciles the trip against the settlement file.

What happens if an outlet closes unexpectedly during the route?

Astra re-sequences the remaining outlets against the live picture and surfaces the skipped outlet to the beat supervisor for tomorrow's planning. The impact on OTIF is measured and visible.

Does this work for FMCG categories with temperature requirements?

Yes. Temperature-controlled DSD runs on the same routing layer with additional constraints for vehicle capability and dwell-time tolerance per stop.

How fast is a DSD rollout?

Most operators pilot in one city cluster inside the first operating quarter and expand geographically after the productivity metrics stabilize. Full geographic rollout typically runs over two to four quarters. --- **Download the full DSD optimization playbook (PDF)** or [subscribe to The Shipsy Brief](/subscribe) for monthly FMCG benchmarks.

Direct Store Delivery (DSD) optimization: routing, pre-sell vs van-sell, and driver productivity

DSD is where FMCG unit economics live or die. Shipsy’s DSD engine runs micro-cluster routing with retailer-tier awareness, supports both pre-sell and van-sell execution models on the same driver app, and measures productivity per truck per day at the SKU level. The routing unlock, the model unlock, and the measurement unlock only compound when they run on one platform.

The finding

Direct Store Delivery is the highest-touch, highest-complexity execution model in FMCG. Every truck carries sales authority, cash, inventory, and returns simultaneously. Every outlet decision affects productivity, inventory accuracy, and cash handling in the same stroke. Shipsy aggregate data across FMCG operators running DSD shows driver productivity moving materially — more outlets served per truck per day, at the same or better SLA — when routing is micro-clustered against retailer priority, when pre-sell and van-sell coexist on one execution layer, and when outcomes are measured per stop rather than aggregated per route.

Why it’s happening

Three mechanics compound.

1. Micro-cluster routing tuned to DSD drop density. Urban DSD routes can hit 30-60 outlets per day. Dense-cluster routing with time-window discipline is the only way to push productivity without breaking SLA. Astra re-sequences routes when outlets reschedule or cancel intra-day.

2. Unified pre-sell and van-sell on one driver app. Some FMCG categories run pre-sell (order today, deliver tomorrow). Others run van-sell (salesperson sells from the truck on the spot). Many run both on the same route. Shipsy’s driver app handles both modes natively with SKU-level allocation, ePOD, and cash collection.

3. Productivity measured at outlet level, not route level. Shipsy surfaces outlets-per-hour, revenue-per-stop, and returns-per-outlet so supervisors can identify underperforming territory assignments rather than underperforming drivers. The feedback loop into route redesign and beat planning is what moves the needle sustainably.

Net: DSD stops being a “high driver variance” business and becomes a data-driven productivity business, where route design, driver development, and outlet performance are all visible in the same system.

What it means for FMCG operators

Operators split by how they treat DSD variance.

Driver-luck operators blame bad days on driver variance — slow driver, difficult outlets, traffic. Their response is incentive restructuring or driver rotation. This handles symptoms rather than causes.

System-driven operators treat DSD variance as a system output and re-engineer route density, outlet assignment, and inventory loads against the data. Driver variance drops as a downstream effect.

DSD capability	Traditional approach	AI-native approach (Shipsy)
Route planning	Static beats	Micro-cluster routing, re-optimized intra-day
Pre-sell + van-sell	Two systems, reconciled manually	Unified execution on one driver app
Outlet time windows	Best effort	Time-window routing with tracking
Proof of delivery	Paper, next-day	ePOD with geofence, real-time
Cash handling	Manual reconciliation	Driver app reconciles at trip end
Productivity metrics	Route-level, monthly	Outlet-level, live
Returns management	Informal, trip-end	SKU-level returns flow inside the app

Three implications.

Driver productivity is a system property. It is designed in, not squeezed out.
Returns and cash are first-class flows. Treating them as afterthoughts hides real cost.
Outlet-level metrics drive beat redesign. Aggregate numbers never surface the outlets that are dragging the route.

What to do about it

Audit your DSD productivity at outlet level before redesigning any routes. Most operators discover that 10-15% of outlets consume 30% of the effort. Pilot micro-cluster routing plus unified pre-sell/van-sell on one urban cluster and measure outlets-per-truck-per-day alongside OTIF. And treat the driver app as the primary operational substrate — if it doesn’t handle pre-sell, van-sell, ePOD, cash, and returns natively, you are reconciling across systems for the rest of the year.

For how secondary distribution feeds DSD, read our secondary distribution guide. Explore Shipsy for FMCG operators and Route Optimization and Planning.