Direct Store Delivery: A Practical Operations Guide

You can tell a lot about a distributor by asking one question: how long does it take to close yesterday's routes? If the answer involves a shoebox of paper tickets, a spreadsheet and a Thursday argument with a store manager about three cases of returns nobody can prove, you already know where the money is going.

Direct store delivery is the model where you own the last mile and the shelf. Your truck, your driver, your invoice, your merchandising — no retailer distribution center in the middle. It is how most beverages, snacks, bread, dairy, tortillas, beer and a long tail of specialty goods reach US shelves. It is also the operating model with the most places for cash and product to leak, because every transaction happens away from your building, in someone else's store, under time pressure.

This guide is about running it well: the economics, the workflow, and what to do when paper or a generic app stops being enough.

Where direct store delivery actually loses money

Nobody budgets for leakage, so it never shows up as a line item. It shows up as margin that is two points lower than the model said it would be.

The three leaks that hide in plain sight

Unverified credits and returns. A driver takes back six cases of expired product and writes it on a ticket. Maybe the store signs. Maybe the ticket gets wet. By the time accounting sees it, there is no way to tell a legitimate credit from a favor, a mistake, or theft. Credits as a percentage of gross sales is the single most diagnostic DSD metric, and most companies cannot calculate it by route.

Phantom out-of-stocks. The driver delivers to the back room, the product never reaches the shelf, and the store reorders. You paid to move inventory that generated no sale. Without a shelf-condition check captured at the stop, you will never see this pattern, only the symptom: erratic reorders from the same five accounts.

Settlement drift. Cash and check collections reconciled manually at the end of the week give small discrepancies a place to live. Same-day settlement is not bureaucracy — it is the control that makes the rest of your numbers trustworthy.

A cost-per-stop formula you can run this afternoon

Before buying software, find out what a stop costs. Pull one month of data for one route and compute:

• Driver cost per hour = (wages + payroll burden + benefits) ÷ paid hours
• Vehicle cost per mile = (lease or depreciation + fuel + maintenance + insurance) ÷ miles driven
• Cost per stop = (driver hours per route × driver cost per hour + route miles × vehicle cost per mile) ÷ stops completed

Then compare that number to average gross margin per stop. Any account where margin per stop is below cost per stop is being subsidized by the rest of your book. On a typical DSD route you will find three to eight of them, and the fix is rarely software — it is a minimum order, a frequency change, or a transition to a wholesaler.

Run the same math on time. If service time averages 18 minutes and drive time averages 9, your drivers spend two-thirds of the day in stores. That tells you to optimize the in-store workflow first and routing second — the opposite of what most vendors will sell you.

The workflow that holds up at the shelf

A good direct store delivery process is boring and identical every time. Here is the sequence worth standardizing.

Before the truck leaves

Load verification against a committed load list. The driver scans or confirms what is physically on the truck. Truck inventory becomes a real inventory location with a balance, not an assumption. Every case is either sold, returned, or still on the truck at end of day — and the three have to add up.

Route sequence with service windows. Most retail accounts have receiving windows, and a DSD route that ignores them produces drivers waiting at back doors. Capture the window per account and sequence against it.

Compliance basics. If you move food, the FDA's Sanitary Transportation rule under FSMA sets expectations for vehicle sanitation, temperature control and record keeping. Driver hours are governed by the FMCSA hours-of-service regulations, including short-haul exceptions that many DSD fleets rely on. Confirm how these apply to your specific operation with a transportation compliance specialist — the exceptions are narrower than they look.

At the stop: six steps, in order

1. Check in with a timestamp and GPS location, so service time becomes measurable.
2. Count the shelf and back stock before writing the order. This is what separates DSD from dropping pallets.
3. Build the order with suggested quantities based on recent movement at that account.
4. Merchandise — rotate stock, pull expired product, face the shelf, photograph the result.
5. Invoice and collect on the spot, printing or emailing the document the store needs.
6. Capture proof of delivery: signature, photo, exceptions and credits, all attached to the stop.

The reason this belongs in an app rather than on paper is not elegance. It is that each step produces a number you can manage: service time per stop, lines per order, credits per route, shelf-photo compliance, collection rate. Paper produces none of them.

When teams reach this point, most want the whole sequence in one tool the driver can run one-handed — which is exactly the job a sales rep app with a digital catalog is built for: order entry, suggested quantities, shelf photos, invoicing and proof of delivery in a single pass, online or off.

Off-the-shelf DSD software or custom: how to decide

Packaged DSD platforms exist and some are good. Use this checklist honestly before assuming you need something built.

• Does it work fully offline? Store back rooms are concrete boxes. If the app needs a signal to write an order, it will fail daily.
• Can it model your pricing? DSD pricing is messy: per-account price lists, promotional allowances, mixed cases, deposits, slotting. Most rejections of packaged software trace back to pricing rules.
• Does it integrate with your ERP or accounting system without a nightly CSV ritual?
• Does it handle your settlement model — driver as employee, independent operator, or consignment?
• Can one person at your company change a price list or add an account without a support ticket?

Four or five yes answers means buy. Two or fewer, especially on pricing and settlement, means you will spend more on workarounds than on a system that fits. That is the honest threshold for building custom — and it is also where a nearshore team helps: a Mexico-based development team working in US time zones can run daily standups with your operations lead and ride routes with your drivers, which matters more than it sounds when the requirements live in a driver's head.

A 90-day rollout drivers won't fight

DSD technology projects fail for human reasons. Drivers are paid on throughput, and anything that slows a stop is a pay cut in their eyes. Sequence the rollout accordingly.

Days 1–30: ride along, then measure

Pick two routes — your best driver and your most difficult account mix. Ride the full day. Time each step at each stop. Write down every workaround you see, because each one is an unwritten requirement. Establish baselines for service time, credits percentage, collection rate and lines per order.

Days 31–60: pilot on two routes

Run new and old systems in parallel for two weeks, then cut the paper. Give those drivers a direct line to someone who can fix things same-day. Target a service time no worse than baseline by week two and better by week four — if the app cannot beat paper on a good driver's route, it is not ready for the hard ones.

Days 61–90: scale and close the loop

Roll by route cluster, not all at once. Add a weekly review of three numbers only: credits as a percentage of sales by route, service time per stop, and same-day settlement rate. Three metrics reviewed every week beat a twenty-tile dashboard nobody opens.

What good looks like

A mature direct store delivery operation closes every route the same day. Truck inventory reconciles to the case. Credits are photographed and attributable to an account, a driver and a reason code. Cost per stop is known by account, and unprofitable accounts get a decision instead of a shrug. Shelf photos let a sales manager see 40 stores from a desk.

None of that requires a platform overhaul. It requires one disciplined workflow, captured where the work happens, and three numbers reviewed weekly.

If your direct store delivery routes still close on paper — or on an app your drivers route around — start with the cost-per-stop math on a single route this week. It usually makes the decision obvious. When you are ready to put a real workflow in your drivers' hands, get a quote for your project and bring your worst route with you; that is the one worth designing for.