In short: choose a WMS by how your warehouse looks on a bad day, not by its feature list. At the demo, test with your own data the scenarios that cost you today: a receipt with shortages, an order with a swapped product, a count on one rack. Then ask the supplier what happens after the contract is signed.
Start from your warehouse, not from the feature list
Almost every WMS has receiving, picking, shipping and stock-taking. Feature lists look alike, and a comparison based on them usually ends with choosing the one with the most lines. The real differences are in the details: what the program does when someone scans the wrong product, when two pieces are missing from an order, or when a supplier ships goods without a barcode.
Before talking to any supplier, write three things on one page:
- What it costs you now: orders shipped wrong, hours spent looking for goods, stock counts with differences, warranty claims where you cannot find the serial number.
- What the warehouse looks like: how many items, how many locations, how many people per shift, how many orders a day and how different they are.
- What has to stay: the invoicing software or ERP, the printers, the terminals, the way racks are coded today.
That page becomes the basis of the demo. If the supplier asks you nothing about it before showing the program, that already tells you something.
Four types of solutions
Broadly, you will find four options. None of them suits everyone.
- The stock module of the ERP or invoicing software. It costs nothing extra and is already linked to invoices. It tracks stock by warehouse, rarely by location, and is not designed for scanning in the aisle. It is enough for a small warehouse with one storekeeper and a few hundred items.
- A standard international WMS. Feature-rich and proven in large warehouses. Implementation is usually long and expensive, changes go through the local partner, and the interface and support are not always in your language.
- A configurable WMS, implemented locally. An existing product adapted to your warehouse by a team in the same country. It is the usual compromise for mid-sized distributors and manufacturers, as long as the adaptations do not turn into a separate project.
- Development from scratch. It fits the process exactly, but you also pay for what every WMS has: locations, labels, scanning, roles, history. It only makes sense when the process really does not resemble anything on the market.
We wrote separately about choosing between an existing product and an application built for you: custom software or off-the-shelf product.
The criteria that matter
What a good WMS should contain is covered in our guide on what a WMS is. When choosing, the questions are more concrete:
- Locations: can you describe the warehouse in levels (zone, rack, shelf, position) as it is, or do you have to bend it to the program?
- Scanning: does it work on ordinary phones or only on dedicated terminals? Which codes does it read: EAN, Code128, QR, DataMatrix?
- Batches, serials, expiry: are they tracked from receipt to delivery, and can it block the sale of an expired batch?
- The link to your current software: via API, directly from the database or through files? Who builds it and who maintains it after go-live?
- Labels: does it generate them in your printer’s format, for items, batches and racks?
- Working without signal: what happens in a corner of the warehouse that WiFi does not reach?
- Hosting: in the cloud, on an instance that is yours only, or on your server? What happens to the data when the contract ends?
- Language and support: are the interface, the guide and support in your language? Who answers when shipping gets stuck on a Friday evening?
How to test a demo
A demo prepared by the supplier shows what works well. To see the rest, ask them to run your scenarios, with a few dozen real items from your catalogue. Five scenarios tell you almost everything:
- Receipt with differences: the supplier sent 48 pieces instead of 50 and a product that was not on the order. What does the storekeeper see on the phone, and what reaches the invoicing software?
- Picking with a mistake: the storekeeper scans a similar but wrong product. Does the program stop it or let it into the parcel?
- Incomplete order: two pieces are missing from the shelf. How is the line closed, what stays open and who finds out?
- Move and count on one rack: you move goods between two positions, then count the rack. How long does it take, and what happens to the difference found?
- Return: a customer returns a serialised product. Which rack does it go to, and how does it appear in the serial’s history?
Watch how many taps the storekeeper needs for each step. On a day with hundreds of order lines, the difference between two and five taps per line shows at the end of the shift.
Questions for the supplier
- Who does the implementation: the team that wrote the program or a partner?
- What is included in the price and what is paid separately: licence, implementation, the link to the invoicing software, training, support, later changes?
- How are changes requested and estimated after go-live?
- Who owns the code of the adaptations made for you?
- How do you get your data back if you leave: in what format and how fast?
- What response time does support have, and what does a recently solved problem look like?
- Can you talk to a customer using the program in a similar warehouse?
Warning signs
- The demo only runs on the supplier’s data, and they cannot load a few of your products.
- The answer to every question is “it can be configured”, without showing it on screen.
- The estimate arrives before anyone has seen the warehouse or the document flow.
- The link to your invoicing software is “your job”.
- It is not clear who has access to the data and where it is hosted.
Total cost, not just the licence
The licence or subscription is rarely the biggest part of the cost. Count the implementation, the link to the invoicing software, labels and printers, terminals if you need them, your team’s time in the first weeks and support over the following years. Compare offers over the same period, for example three years, not just the first one.
Frequently asked questions
How long does choosing a WMS take?
For a mid-sized warehouse, a few weeks are enough: a day to write down what costs you today, one demo on your scenarios with two or three suppliers, and a conversation with a customer of the finalist.
Is a WMS worth it for a small warehouse?
If you have one storekeeper, a few hundred items and the stock counts come out right, the module in your invoicing software is probably enough. A WMS starts paying off when several people work on the same goods, you have multi-level locations, or products with batches and serial numbers.
Do I have to buy scanner terminals?
Not necessarily. Many WMSs run on Android phones with a camera. Dedicated terminals pay off at high volumes, where reading speed and resistance to knocks matter.
What should I ask a customer of the supplier?
How long the implementation really took compared to the estimate, what did not work the first time, how quickly support responds and whether they would choose the same program again.
If you want to see the scenarios above in a WMS, DataFlows Depo runs on the Android phones you already have and connects to your invoicing software. Write to us and we will run the demo on your products.
