ERP: the five mistakes that cost SMEs a year of project time
Changing ERP is one of the heaviest projects an SME will undertake, and one of the riskiest: between 55 and 75 % of these projects fail partially or completely. The most troubling part is that the cause is almost never the software. It is human, organisational, and above all avoidable. Here are the five mistakes that turn a strategic investment into a lost year, and the method to outflank them.
The scenario is almost always the same. An SME decides to replace its ageing tools with a modern ERP that is meant to unify everything: sales, purchasing, stock, accounting. The software is chosen after a few seductive demos, the contract is signed, the project starts. A year later, the budget has doubled, go-live has been pushed back three times, the teams are working around the new system, and the leader wonders how a project meant to simplify their life ended up complicating it to this extent. They are not alone: this is the outcome of most ERP projects.
Two projects in three go off the rails
The studies converge, whatever the source. According to work by Panorama Consulting, Gartner and Deloitte, between 55 and 75 % of ERP implementations fail to meet their objectives, with frequent and sometimes spectacular cost and schedule overruns. In manufacturing, some analyses report average budget overruns of more than 200 % on failing projects. But the most important figure is not the failure rate: it is the cause. As with all transformations, an ERP failure is almost never technological. The software, most of the time, works. What fails is the way it was chosen, prepared and deployed.
Mistake 1: believing it's an IT project
This is the mother of all mistakes, the one from which almost all the others flow. An ERP touches every process, every department, every workstation. It is a project to transform the company that happens to rest on a piece of software, not a software project with a few side effects. The direct consequence: almost the entire budget goes on the tool and a derisory share on preparing the teams and on change management. Yet that is exactly where success is decided, as every study on adoption shows.
Mistake 2: choosing by the demo, not by the process
A good demo is built to seduce, not to reveal whether the tool fits your reality. Choosing an ERP on the strength of a presentation is like buying a house from a photo of the façade. The right method reverses the order: first map your real processes and your real needs, and only then hold the tools up against those requirements. An SME that knows precisely what it needs its ERP to do chooses well; an SME that lets itself be led by the most dazzling features ends up with a tool that is oversized, under-used and badly fitted to its trade.
« You don't choose an ERP, you choose a way of working. The software is only the vehicle that carries it. »
Mistake 3: wanting everything, all at once
The temptation of the big bang is irresistible and lethal: switching every department over on the same day, across its full scope. That means multiplying the risks by the number of fronts opened at once. Projects that succeed proceed in stages: a narrow scope first, stabilised and adopted, then progressive extension. Every step cleared strengthens the team's confidence and know-how. Every attempt to do it all in one go invites a chain collapse, where one blockage on a single module paralyses the entire go-live.
Mistake 4: neglecting data quality
A brand-new ERP fed with old, inconsistent or duplicated data faithfully reproduces the mess that came before, only faster. Data migration is the least visible and most underestimated part of a project: cleaning, de-duplicating, harmonising before migrating. It is thankless, it eats time, and it is non-negotiable. An ERP does not fix dirty data, it spreads it.
Mistake 5: letting go at go-live
Go-live day is not the end of the project, it is the start of the most delicate period. That is when teams hit real cases, when resistance finds its voice, when the real problems surface. Projects that fail cut the support at the precise moment it becomes most useful. Those that succeed keep close support in place through the first weeks, fix things fast, and reassure. Adoption cannot be decreed on day one, it has to be supported over the months that follow.
A leader's project, not an IT specialist's
The thread running through these five mistakes is clear: an ERP succeeds or fails on management and organisational decisions, not on lines of code. It is a leader's project that delegates technical execution, never a technical project the leader loses interest in. The SMEs that approach it this way join the third that succeed, the ones for whom the ERP becomes a real performance lever from the first year. The others discover, a year and a budget later, that the software was never the problem.
Sources
Panorama Consulting Group, Gartner and Deloitte, studies on ERP implementation failure rates (55 to 75 % partial or total failure; post-deployment correction costing 3 to 10 times more) · Sector analyses of budget overruns in manufacturing · Standish Group (CHAOS Report) on IT project performance · Documented experience reports on the human and organisational causes of ERP failure · LUCID field observations, assignments 2025-2026, anonymised references.
The first 30-minute conversation is free. That's where it gets framed, with no commitment.