Choosing the wrong ERP is one of the most expensive mistakes a business can make. A company sits through a slick demo, falls for the system that looks great, signs the deal, and spends the next few years regretting it. And here's the part worth sitting with: it's usually not because the software was bad. It's because they never worked out what they actually needed first.
We've circled this topic plenty of times on the podcast — the pitfalls of picking the wrong one, upgrade versus migrate, what to ask a vendor before you buy. But we'd never done an episode specifically on how to choose, from both sides: the inward look at your own business, and the outward look at the market. This is that episode written down.
The headline, before any of the detail: almost all the work that decides whether your ERP project succeeds happens before you email a single vendor.
First, look inward
Start with the most basic question, and answer it honestly. Why do you need an ERP at all?
The most common answer is "ours is old". And old brings real problems — end-of-life support, security exposure from missed patches, a dwindling pool of people who can support it, knowledge walking out the door as the people who understood the system retire. All legitimate. But old, on its own, is not a reason to buy a new ERP. It's a symptom.
The sharper question is this: what can you not do today that's costing you money? A fifteen-year-old system that still fits the business and does the job is not a problem. The issue is never really the age — it's the fit. Has your business changed shape since you implemented? Different products, different customers, customers who now buy through an Amazon shopfront instead of a straight manufacturing-to-customer line? If the system no longer fits the business you actually run, that's your reason. "It's fifteen years old" is not.
And while you're being honest, consider the option nobody likes to put on the table: doing nothing. A lot of what people expect from a new ERP, they can do without one. "Our data is a mess" is not an ERP problem — go and clean it up. You'll have to clean it up for the new system anyway; you can't carry ten-year-old open transactions into a fresh implementation. Same with messy processes. Always ask what the value of doing nothing is, because sometimes it's higher than it looks.
Know who you are
Once you know why, work out who you are — and here it really is just the facts. What's your revenue? How many actual users of the system (not employees — the two are rarely the same)? How many sites? How many countries? What industry, and what regulations and compliance regime sit on top of it? Gather that set of facts and a surprising number of options take themselves off the table. A small manufacturer doesn't need Dynamics F&O. A husband-and-wife operation isn't shopping at the enterprise tier.
One thing to get right at this stage: choose on fit first, cost second. It's tempting to lead with the budget, and it's the wrong move — it quietly steers you toward systems that were never built for your business. The return on an ERP doesn't show up in year one. It shows up in years three, four, five and six, in how fast you can grow and which markets you can move into. Lead with cost and you optimise for the wrong horizon.
And knowing who you are doesn't mean reaching for a general-purpose ERP that does a bit of everything. Those tend to be good at nothing. There's a reason Acumatica ships different editions — the manufacturing edition comes with the BOM management and the rest of what a manufacturer actually needs. There's a reason QAD is everywhere in automotive: it was built for it. Know yourself well enough to find the system catered to you, and to confidently discount the ones that aren't.
Know your processes (this is the big one)
If you spend more time on one step than any other, make it this one.
Every ERP digitises your processes — exactly as good or as bad as they currently are. Most systems can be configured to do almost anything you ask, well or badly. So if you don't genuinely know your processes going in, you'll overspend, overrun, and still not solve the problems you set out to solve.
Here's the test. Can you describe your business processes without opening the ERP? If you can't, you don't know your processes — you know your software. Those are different things, and the gap between them is where projects go to die.
The more useful version of the question isn't "what do you do?" — it's "why do you do it?" Let your project manager be the internal consultant and go asking why, before any external consultant turns up and uncovers it for you during discovery, when it's far more expensive to change. By the time you've selected a system, most people are already locked into the process they want to replicate — "we need to be able to do the same thing in the system" — and that's exactly the moment everything stalls into a 50- or 100-hour customisation. Have the open conversation now, while you're still choosing.
Be warned: this gets uncomfortable. Start picking at the band-aids and you uncover a lot of complexity and workarounds. The CFO who's never seen the shop floor doesn't know that Julie in accounting is running a macro-ridden spreadsheet because nobody understood the system after the last accountant left and took the tribal knowledge with her. Get the people closest to the work in the same room as the C-suite and there are usually a few genuine shocks about how the business actually runs versus how leadership assumed it did. That's not a bad thing — it's the whole point. It's a chance to reset, clean house, and stop digitising habits you should have killed years ago.
One more distinction worth nailing down: complexities versus non-negotiables. Not every quirk matters. Jane's enormous macro spreadsheet doesn't influence which ERP you pick. A non-negotiable has a clean definition — it's something that, if it's missing, loses you money, damages a customer relationship, or lands you with a compliance penalty. Everything else is a preference in disguise, dressed up as a requirement. Be ruthless about telling the two apart.
Now look outward
If you've done the inward work, the outward part is far easier — you're no longer staring at a blank Google page, you're working against a shortlist you've already built. The extremes have come off: too big, too small, wrong industry, all gone.
For the market research itself, there are good impartial starting points. G2 and Compare Software for reviews and side-by-sides. Research firms like Info-Tech that publish proper industry-specific roundups. Your trade associations — ask what the businesses like yours actually run. And the communities where real users complain: Reddit, sysadmin forums, vendor community boards. Treat it like buying a car. You'd find the owners' group and see what current owners moan about; do exactly that for ERP. Check the feature-request forums too — are the vendors delivering on them, or is there a wishlist that's been ignored for three years?
AI is a perfectly good starting point now, precisely because you've done the inward work and can ask a sharp question: "I'm a B2C e-commerce business with £50m revenue choosing a new ERP — which systems should I even be looking at?" That's a question you can only ask well once you know who you are.
Demos: your job is to break them
Then come the demos, and the single most important thing to remember is that every demo works. That is the entire point of a demo. Flashy, two clicks, month end closed in ten seconds — it's never like that in real life, and it's astonishing how many companies still get duped by it.
So watch the demo for the things that matter — your end users will stare at this UI eight hours a day, so it needs to be clear and navigable — but spend your real energy trying to break it. If you can get a trial environment, try to break it. If a partner is demoing live, get them to run it on your data, and bring the hard questions you found while mapping your processes. Don't let them show you a sales order going through cleanly. Show me month end. Show me how you fulfil a back order as a priority. A good partner will welcome this — ours builds flowcharts of the major processes and demos on the prospect's own data precisely because demos go by fast and people need to picture themselves using the thing. Be wary of YouTube demos; they're generic at best and misleading at worst.
The partner matters as much as the system
This is the part people underrate. More than one ERP will work for you — they're not all the same, but they're all ERP, and several will do the job. Which means a bad implementation of the slightly-better system is worse than a good implementation of the slightly-worse one. An ERP can earn a place on your shortlist purely because you know a strong implementation partner for it. It can lose its place because nobody in your network can find decent support for it.
Look for a partner who'll say no to you. If you offer them thirty grand of customisation, the good ones tell you that you don't need it. Check they're certified, that they have real depth on the product rather than two consultants bolted onto a cybersecurity firm, and that ERP is their actual focus. Watch the handoff: the slick salesman who knows the product inside out is rarely the person implementing it. Find out who's doing the implementation, the training, and the ongoing support — and notice how much they listen. If they're doing 80% of the talking, walk. They should be asking about your business, not running a PowerPoint of logos you don't care about.
And mind the red flags. A 30-minute demo followed by a contract in your inbox the same afternoon is not a process — it's a sales tactic. Choosing an ERP properly takes four to eight months. You don't buy it like a shirt.
Where this leaves you
Choosing an ERP comes down to three moves, in order. Look inward until you genuinely know your business, your processes, and your non-negotiables. Research second, against a shortlist rather than a blank page. And never trust a demo you haven't tried to break.
Do that and you stand a real chance of choosing well. Skip it and you become the cautionary tale — the company that fell for the demo and spent the next three years wishing it had heard this episode first.
This is from episode S4E13 of The ABCs of ERP & Beyond
