Most small distribution and manufacturing companies start their warehouse on paper, then move to Excel, and when on a Friday evening the stock in the system doesn't match the stock on the shelves by 15%, someone floats the idea: "maybe we should finally buy a WMS". Sometimes that's a good idea three years late. Sometimes it's two years too early.
This piece is about how to tell which of those moments you're in — and what to do in each. No pushing a "cloud warehouse platform" for half a million, but also no romanticising the paper notes where today someone is still writing quantities with a marker.
How to tell your warehouse is still "on paper", even if you have an ERP
We have clients who called us because they "want a WMS". On the first walk through the warehouse it turns out their ERP already has a WMS module — nobody just uses it. The warehouse keeper knows his shelves, taps everything in from memory, and the ERP gets a bulk correction once a week. That's not "no WMS". That's a paper WMS living in one person's head.
Signals we see most often at our clients:
- System stock regularly differs from physical stock by 5% or more.
- After the warehouse keeper's holiday the company "lightly panics".
- The annual inventory count takes 3–5 days and ends with a mass correction.
- "Wrong items shipped" complaints happen at least several times a month.
- Orders are picked from memory, not from a pick list.
- The price of a product you just sold came from an invoice three months old, because nobody updated the catalogue.
Each of these on its own is survivable. Together they are very expensive. Most of the time it turns out the company loses more on shipping mistakes, expired stock and stale prices than a proper WMS rollout would cost.
What a WMS actually delivers for an SME
A WMS in an SME is not "SAP-class warehouse software". It's usually a module inside your existing ERP, or a dedicated app (Comarch WMS, the warehouse module in enova365, an add-on for Subiekt GT), plus a few barcode scanners, a label printer, and the discipline that every stock movement passes through a scan.
What actually changes after rollout:
- Every stock movement is recorded in real time. Shipments, receipts, inter-warehouse transfers — visible immediately.
- Picking from a list, not from memory. The picker gets the list on the scanner in optimal order (from rack A to C), scans the product, and the system confirms it's the right one.
- Continuous inventory instead of one big paralysis. Each week you take one segment of racks, check it, correct it. After a year you've audited the whole warehouse and nobody lost a weekend.
- Expiry date and batch tracking. Critical for food, cosmetics, pharma. Without it, you write off stock every quarter because "it was at the back".
What a WMS in an SME won't do:
- It won't plan your production — that's MES or APS.
- It won't sort out your supplier negotiations.
- It won't replace your warehouse keeper. It will let one person do the work of two — but somebody still has to stand by the rack.
A real example: a 25-person cosmetics distributor
At one of our clients — a 25-person wholesale distributor of professional cosmetics running on Comarch Optima — the warehouse worked for years on a "the keeper knows where everything is" basis. System stock was updated through invoicing, which means it was never really live. The annual inventory was a two-day ordeal. Complaints from salon customers along the lines of "I didn't get what I ordered" averaged 8 per month.
We deployed Comarch WMS with four wired scanners, a thermal transfer label printer and a simple policy: every product entering the warehouse gets a label, every shipment is scanned line by line, the system rejects the pick if the stock in the location doesn't match. The first three months were painful — the warehouse keeper had to unlearn "walking from memory". Then it clicked.
Results after 9 months:
- Picking errors fell from 8 complaints per month to fewer than 1.
- Time to pick a typical order: from 22 to 13 minutes (-40%).
- Annual inventory count: from 2 days to 3 hours, because stock stays accurate and you only spot-check.
- Stock loss from expired products dropped by roughly 60% — the system watches dates.
The project itself — licences, hardware, integration, training — cost just under PLN 70,000. It paid for itself in under 11 months, mostly through avoided complaints and recovered labour time.
This isn't a spectacular case. It's the typical result when you take a warehouse running on "the keeper knows" and overlay the discipline of "every movement is scanned".
The hardware you don't get to skip
This is the part companies most often try to "trim" out of the quote. Don't trim.
- Barcode scanners — Zebra, Honeywell, Datalogic. Cheap consumer models off the shelf don't survive daily use in a warehouse. The number of service tickets after a year will tell you everything.
- Thermal transfer label printer — so labels don't fade after three months in the sun or in a cold store. Inkjet won't do.
- Wi-Fi covering the whole warehouse — a single dead spot equals one lost pick per day. Usually you need to add access points.
- Tablets or terminals for pickers — in a small warehouse a scanner is enough; in a larger one a tablet with the list helps.
Total hardware cost for a typical warehouse at a 20–30-person company runs to PLN 15,000–25,000. This is not the place to economise. Hardware that fails once a week kills enthusiasm permanently and the whole project loses its point.
Integrating with your ERP — the pitfalls
Most Polish SMEs run Comarch Optima, Subiekt GT, enova365 or Microsoft Dynamics. Each of those handles warehousing differently — and each has its own traps.
- Comarch Optima + Comarch WMS — the simplest integration, but watch out for per-seat licensing. Easy to buy, easy to overpay.
- Subiekt GT — works, but on larger volumes (e.g. 500+ lines per day) it starts to slow. This is often the moment an SME considers migrating to Subiekt nexo.
- enova365 — flexible, but needs a partner who knows the warehouse module, not just the accounting one.
- Microsoft Dynamics AX / 365 — usually in larger companies, but if you already have it, the built-in warehousing is strong. It does need an experienced consultant.
The worst scenario is two unintegrated systems: the ERP knows its own, the WMS knows its own, and once a day somebody "syncs" them by hand. That's worse than not having a WMS.
What not to automate
- Non-standard inbound (production returns, mixed pallets, complaints) — let the warehouse keeper enter those manually. A rule covering 5% of cases will take three months to write.
- Random storage locations instead of fixed ones — in an SME, fixed locations are usually simpler and cheaper to operate. "Chaotic warehousing" is a solution for 10,000 SKUs, not for 800.
- Voice picking — expensive, doesn't scale well with 5 people. Only makes sense from 20+ pickers up.
Where to start
- Measure today. How many "wrong items shipped" complaints per month? How long does the annual count take? How long does a typical shipment take to pick?
- Check whether a WMS already exists in your ERP and what's missing to switch it on. The answer is often "we have it, nobody uses it".
- Start with one category — usually the fastest-moving SKUs (typically 20% of items make 80% of shipments). That's enough to see the effect.
- Buy top-tier hardware for the warehouse. Cheap scanners are a false economy.
- Budget for training and a month of tolerance — the first weeks will be slower than before WMS. That's normal.
If you're not sure whether a WMS pays off for you today, or whether you already have everything you need in your ERP, we'll help you work it out from real numbers in your own warehouse. The first call is free and takes half an hour.
A WMS won't fix a warehouse where nobody wants to scan. But in a warehouse where people scan, a WMS turns two keepers into three — and better.