In this article (11 sections)
A distributor has a problem a retailer doesn't: the same product isn't worth the same to every customer. And that detail, which looks administrative, is why most off-the-shelf systems end up with a spreadsheet running alongside them.
Here's what a distribution system actually has to solve, and how to decide between buying one and building one.
The five problems specific to the trade
1. Customer-specific pricing
Different price lists, volume breaks, product rebates, promotional pricing for a period, special terms negotiated with the large account.
This is where most systems come up short. Plenty of them handle "list A, B and C" and that's the end of it. If your reality is that every significant customer has their own scheme, you need something that genuinely covers it.
2. Units of measure that don't line up
You buy by the pallet, stock by the case, sell by the each. Or you buy by weight and sell catch-weight cases.
A system that handles one unit per item forces you to do the conversion by hand, and that's exactly where the inventory discrepancies nobody can explain come from.
3. Inventory isn't one number
What's on the shelf, what's committed to confirmed orders, what's on order and in transit from a supplier, what's out with a sales rep.
Selling against on-hand instead of available-to-promise is the number one cause of orders that can't be shipped.
4. Picking and delivery
The order comes in, it has to be picked, checked, loaded onto the truck in an order that matches the route, delivered, and the signed proof of delivery has to come back.
That chain, run on paper tickets, is where deliveries get lost and where you get claims you can't disprove.
5. Accounts receivable and credit limits
How much each customer owes, how much they're approved for, what happens when an order pushes them over, and who authorizes the exception.
If that check isn't in the system, it's in somebody's head, and it fails the day that person isn't in.
What it has to have, concretely
- Per-customer pricing and discounts, with effective dates.
- Multiple units of measure per item, with automatic conversion.
- Inventory by warehouse, with committed and available tracked separately.
- Order, packing slip and invoice as distinct linked documents.
- A credit check when the order is confirmed, not when it's invoiced.
- Picking with verification, ideally with barcode scanning.
- A driver app: route manifest, delivery, proof of delivery, exceptions.
- An app or portal for the sales rep: order entry at the customer, with real stock and real pricing.
- Invoicing wired into your accounting, with sales tax handled.
The last two pay back fastest, and they're the two that most often get cut from the initial scope.
Buy or build
Buy a product if your operation looks like any other distributor's: manageable price lists, one warehouse, simple in-house delivery. There are distribution ERPs that cover this trade well, cost far less and are available tomorrow.
Build when the commercial logic is your differentiator and no product covers it: a pricing formula of your own, a rebate scheme based on combinations of products, a consignment operation, a workflow a large customer demands in a particular shape.
The short version of the question: if you explain your pricing scheme to an ERP implementer and their face changes, that's your answer.
The ranges
- Implementing an existing distribution ERP: licenses plus implementation, typically USD 3,000 to 12,000 the first year, depending on size.
- A custom distribution management system: USD 12,000 to 30,000.
- Adding a sales rep app or a delivery app: between USD 4,000 and 10,000 on top, each.
Where to start if you're going to build
Not with the whole system. With the process that hurts most, which in a distributor is almost always one of these two:
Order entry by the sales rep. If today the rep writes on paper or sends a message and somebody keys it in later, you have rework, transcription errors and delay. An app that takes the order with real pricing and real stock kills all three at once.
Pick verification. If claims for shorts or wrong items are frequent, barcode-verified picking pays for itself in months.
Pick one, do it well, and the operation itself will start asking for the rest.
The classic mistake in this trade
Digitizing delivery before the inventory is trustworthy.
A driver app on top of inventory that isn't real produces a new problem: now the failed deliveries are recorded, neatly, with a timestamp — and they still fail. The app doesn't create product.
The order that works is the reverse: first make the system's inventory match the warehouse's inventory, then make it impossible to confirm an order against something you don't have, and only then the delivery side.