The most expensive ERP rollout in a small company is one that ends up on the shelf. The second most expensive is the one where, a year in, you realise you paid for modules nobody ever turned on. The decision on which system to buy takes about three weeks. You live with it for seven years.
Most ERP presentations we've sat through with clients look identical. The sales rep opens a pre-baked demo with a fictional "Kowalski Ltd." company, clicks through neatly prepared scenarios, says "and here we can also...", leaves an hour later with a PDF quote. The client comes away convinced they saw a working system. What they actually saw was clickable slides.
This article is for the SME owner who just received three ERP quotes and has no idea how to compare them. Not on a marketing feature list — on what actually matters.
First, write down what you really need
The single most common mistake: the conversation with the vendor starts before the conversation inside your own company. You walk into a demo not knowing what you want to see — so you see what the vendor wants to show. The result: you buy a system for scenarios you don't have.
Before you book the first demo, put on one page:
- 5–10 critical processes the system has to handle. Specifically: "intake of an order from a retail-chain customer", "issuing a goods-issued document with a reservation from a specific batch", "monthly production settlement with material cost", not "sales support".
- Numbers: documents per month, items in the catalogue, concurrent users, number of warehouses, number of legal entities in the group.
- Integrations that must work on day one: bank, courier, webshop, e-invoicing (KSeF in Poland), retail-chain EDI, your existing CRM.
- Hard constraints: regulated industry, HACCP certification, batch traceability, multi-currency.
That one page is worth more than all the quotes put together. Because suddenly you have something to ask about.
At one of our clients — a 30-person furniture manufacturer — the first version of the RFP read "we want a modern ERP with a production module". Three quotes came back between PLN 180k and 340k and couldn't be compared. After two days of workshops we wrote out 12 processes and five integrations. The second RFP brought back three comparable quotes — and the company picked Comarch Optima with a dedicated production module for about PLN 130k instead of Comarch ERP XL at PLN 280k plus PLN 60k of annual licensing. The system has been running for three years and they use ~85% of what they paid for. Savings: ~PLN 150k on the rollout alone, plus ~PLN 45k a year on licences.
A demo is a show, not the product
A standard sales demo doesn't show you the system — it shows you the system's best moment. Your job is to break the show.
Rules that help:
- Don't watch the demo on their data. Send yours. 50 of your real SKUs, 20 of your customers (anonymised), three of your typical orders. Ask them to run the process on those.
- Make them show the ugly paths. What happens when a customer orders something out of stock? How do you reverse a wrong invoice already submitted to e-invoicing? What does the sales rep see when the warehouse has already reserved the goods?
- Ask about response time on realistic volumes. "What does the annual sales report look like with 800k documents in the database?" The demo usually has 200.
- Put an implementation consultant next to the sales rep. Those are two different people. The sales rep sells; the consultant is the one you'll live with. If the vendor doesn't want to show the consultant, that's a signal.
Read the contract carefully — especially four things
Most ERP contracts get read once, on a Friday, in a hurry, because "the client wants to start Monday". That's the moment you lose the project.
Four places worth pausing:
- Scope of rollout vs. "additional work". What the sales rep showed in the demo often appears in the contract as a "possible extension". Make them list the specific processes from your page — with numbers. Anything outside the list goes by the hour, and that hourly rate is worth negotiating up front, not mid-project.
- Acceptance criteria for each stage. What exactly has to work for you to sign off? How long do you have to raise objections? What happens if you don't sign? Without this the vendor invoices 70% of value after "go-live" and you're left with broken reports.
- Post-implementation support — what it covers and what it doesn't. How many hours a month? Does it cover configuration changes or only bug fixes? SLA on a critical ticket — 4h, 24h, "next business day"?
- Exit and ownership of data. In what format do you get your data when you part ways? Are the reports (custom printouts, forms) yours or the vendor's? Do you have the right to let another integrator in?
You're not a lawyer, but these four things you'll understand in one read. Hold the line on these four.
Ask about the 3-year total cost
The headline price in the quote is the tip of the iceberg. The full cost of an SME ERP is:
- licences (one-off or annual — and that makes a giant difference over three years),
- rollout (analysis, configuration, data migration, training),
- integrations (bank, courier, webshop, e-invoicing — each separately),
- modifications (customisation, custom printouts, custom reports),
- annual support (typically 18–22% of the licence value),
- updates (sometimes included in support, sometimes not),
- infrastructure (on-prem server, vendor cloud, database — sometimes a separate SQL Server licence).
Build a 3-year table and add it all up. The vendor who's cheapest in year one is often the most expensive in year three, and vice versa. Ask outright: "let's count three years together, in one file."
You're choosing the vendor, not just the system
An SME ERP stays in place for 7–10 years. You're sitting on the same sofa with the vendor that whole time. Worth knowing who you're inviting in.
Questions worth asking before you sign:
- How many people at your company know this system? If the answer is one consultant who just wrapped up a rollout somewhere else — that's risk.
- Which of your clients are in our industry? They'll name three — call two. The question isn't "are they happy" (the answer is never "no"), it's "what would you do differently today?".
- What if the person leading our project leaves your company? Is there a backup? Does the project freeze?
- Can we talk to the consultant who'll actually lead us, before we sign? If "no" — be very careful.
The best system with the weakest implementer loses to a mediocre system with a strong implementer. Every time.
What not to buy
A few things simply not worth it in an SME context:
- "The full package, because you'll need it eventually anyway." You won't. Buy the working minimum plus a clear path to extend.
- "Turnkey" configuration without your people in the room. An ERP rolled out without your team at the table is an ERP for the bin.
- Changing everything at once. ERP + WMS + CRM + webshop in one project is a doomed project. Stages.
- No outside pair of eyes on the quote. A second opinion from someone independent of the vendor pays for itself before the first meeting.
What next
Choosing an ERP isn't a decision about a system. It's a decision about a business partner for the next several years, and about how much you'll pay not for the software but for actually using it well.
If you have three quotes on your desk right now and don't know how to compare them — call. A half-hour conversation, where someone external reads your RFP and the quotes, is usually enough to cut more from the budget than the whole engagement costs.
The most expensive ERP isn't the one with the highest quote — it's the one sitting on the shelf a year later.