How to choose an ERP system

Most ERP selections start in the wrong place: a list of products, a set of vendor websites and a series of demos. By the time the third demo finishes, everything looks capable, the differences have blurred, and the decision quietly reduces to price and whoever presented best.

The businesses that end up happy with their ERP almost always did the same thing first. They wrote down how the business actually runs, where it breaks, and what the new system has to change. That document is what turns a demo from a sales presentation into a test, and it is what makes competing quotes comparable.

It also protects you from the most expensive outcome in ERP, which is not choosing the wrong product. It is choosing a perfectly good product and configuring it around processes nobody examined, so the business pays to automate its existing problems.

The rest of this guide walks through the stages in order: requirements, shortlisting, demos, cost comparison, references and the partner decision. Expect the whole process to take eight to sixteen weeks for a mid-sized organisation.

Aim for three vendors. Two gives you no meaningful comparison and five spreads your team so thin that none of the evaluations is done properly. The work is in qualifying the long list down, and most of that can be done without a single demo.

Filter first on sector and size. A platform whose reference customers are 500-user manufacturers will be priced and staffed badly for a 25-user professional services firm, and the reverse is equally true. Ask each vendor directly for customers at your size in your industry before you book anything.

Filter next on the ecosystem. An ERP is a ten-year decision, so the availability of implementation partners, ongoing support, integrations and people who know the product in the job market all matter. A capable niche product with three UK partners is a genuine commercial risk.

Be sceptical of the requirements spreadsheet as a scoring tool. Vendors answer yes to almost everything, because almost everything is technically possible with enough configuration. What differentiates platforms is what they do well out of the box, which only shows up in a demo built on your own scenarios.

Build a three-year total for each shortlisted option rather than comparing monthly licence prices. Licensing is the visible number and often the smaller one. Implementation services, data migration, integrations and training usually exceed year-one licensing, sometimes by a wide margin.

Break licensing down by user type. Cloud ERP platforms differentiate between full users who work in finance and operations and light users who approve, submit timesheets or read reports. Getting that mix right changes the annual figure significantly, and vendors will quote whatever mix you give them.

Include the things quotes routinely omit: your own team's time, the data cleansing effort, testing, training beyond the initial sessions, and the change requests almost every business raises in the first year once people see what else the system can do. A contingency of ten to twenty per cent on services is realistic rather than pessimistic.

Then compare the support model. Response times, named contacts, whether training questions consume support hours, and how vendor release updates are tested against your configuration. Support is where most of the relationship actually happens and it is the part most often glossed over during selection.

Take references seriously and choose who you speak to. Ask for a customer at your size in your sector, and ask to speak to the finance lead or operations manager rather than the sponsor who signed the project off. The useful questions are what went wrong, what it cost beyond the quote, and what they would do differently.

Recognise that for most platforms the product decision and the partner decision are the same decision. The software is identical whoever supplies it. What differs is how well it is scoped, configured, migrated, integrated and supported, and those are entirely down to the people delivering it.

Weight the partner accordingly. Industry familiarity, the seniority of the named delivery team, and a willingness to push back on requirements they think are a bad idea are all stronger predictors of success than accreditation badges. A partner who agrees to everything tends to build expensive, fragile systems.

Finally, plan the internal side before signing. Name a project owner with authority to make decisions, protect their time, and agree who signs off process changes. ERP projects rarely fail on technology. They fail when nobody inside the business owns them.

If you are still working out whether an ERP is the right move, start with our guide to what an ERP system is, which covers the modules, how it differs from accounting software and the signals that a business has outgrown its current tools.

If you already know the direction and want to see how it applies to your sector, we have written practical guides on what an ERP changes across manufacturing, construction, professional services, wholesale, retail, healthcare and more.

And if Microsoft Dynamics 365 Business Central is on your shortlist, our pricing breakdown and partner selection guide cover the two questions that come up most often in the final stages of evaluation.