Most ERP evaluations are built around a feature list, and the feature list is the least useful document in the process.
Every serious product on a shortlist answers yes to almost everything. Multi-currency? Yes. Approval chains? Yes. Works on a phone? Yes. Each answer is true, and together they tell you nothing about which of these projects will still be running in a year.
What decides that is almost never a missing feature. It is migration, adoption, and who is allowed to settle an argument.
The nine questions
Ask these of every vendor, in a meeting, and write down the answers. The differences will be obvious by the third one.
-
What will we stop typing? A system does not give you information you do not already have. It changes how many times the same information is written down. If nobody can name a task that disappears, nothing will.
-
Can you show it on our data? A demo runs on data built to make the product look good. Send three of your real documents and ask to watch them go through. Nothing separates vendors faster.
-
What happens to our accounting system? Does it stay as the book of record or is it replaced? What is exchanged, in which direction, how often? A vague answer here is the most expensive unknown in the whole project.
-
Which of our processes does this not fit? A vendor who says everything fits has not read them. A good answer names two or three and describes the workaround. That is a sign of competence, not of weakness.
-
Who approves what, and where is it recorded? Most companies buying an ERP are buying their way out of approvals that happen in chat messages. If the approval chain is not in the product, the chat stays.
-
Can someone do this in twenty seconds? Standing up, on a phone, with one hand, in a warehouse. A system people find slow gets filled in at the end of the day from memory — and then the numbers are memory again.
-
What does the second month look like? Everyone plans the go-live. Week six is when the trainer has gone, the team is tired and the old spreadsheet is still on a desktop. Ask who is present then.
-
If we leave in three years, what do we take? Ask about the export format and access to your own database. The answer rarely changes the decision, but the way it is given tells you how the relationship will go.
-
Who answers on the last day of the month? Month end is when the system is under load and a mistake costs most. Ask for the name, the hours, and what happens outside them. Then put it in the contract.
Four answers that should worry you
None of these is a lie. Each one is a sentence that sounds reasonable and hides the part that matters.
| What you hear | What it can mean | What to ask next |
|---|---|---|
| “Everything is configurable.” | Configuration is work, and somebody pays for it | “Who configures it, and how many days?” |
| “We will handle migration at the end.” | The riskiest part of the project is unplanned | “Can we do a dry run on our data first?” |
| “Yes, it does that.” | It may be on a roadmap, or possible in theory | “Can you show it now, on our documents?” |
| “Your processes are not standard.” | Sometimes true — and sometimes it is the product | “Which part, and what do firms like us do?” |
A demo is theirs. A dry run is yours.
A demonstration is a performance. The data was chosen, the path was rehearsed, and the awkward cases are not in it. You learn what the product can do on a good day.
A dry run is the opposite. You hand over your own records — customers, items, opening balances — and watch the system take them. It is the only part of an evaluation that tells you about migration, and migration is where these projects actually fail.
The three questions to ask yourselves
The vendor is only half of it. Before any of the nine, answer these internally, out loud, with the people who will have to live with the result.
- Who decides? When sales and finance disagree about how a discount should work, one person has to settle it. If no such person exists, the project will stop at exactly that argument, and no software will unstick it.
- Whose job gets harder? There is always someone, and it is usually the person who currently holds the information in their head. They are not being difficult. They are losing something real, and it is cheaper to acknowledge that early.
- What are we prepared to change? A system that fits your current habits perfectly is a system that changes nothing. If the honest answer is that nothing may change, buy nothing — you will save a year.
Before the first vendor call
Write one page first. Not requirements — processes, in the words your team actually uses, with a name beside each one.
Then mark the three or four that are genuinely yours, the ones where doing it the standard way would make you a worse business. Those are the only processes worth spending the meeting on. Everything else is ordinary, and every product on your list handles it.
If you want a second opinion on that page before you start talking to vendors, that is a conversation we are happy to have. If it ends with us saying you do not need a new system, that is a real outcome and we say it.
Projects behind this article
What we build in this area
-
ERP & business management systems
ERP for trade, distribution and service businesses — sales, stock, finance, payroll and approvals, connected to 1C and e-qaimə.
-
Data, BI & forecasting
Power BI dashboards, data integration and warehouses, telematics data platforms, forecasting and machine learning.