6 signs you need a new ERP (or to invest in your first one)

There's a moment when you realize the system itself is the problem. A planner out, a spreadsheet nobody else can read, a workaround that's been holding production together for 18 months.

If you're reading lists of signs you need a new system, that moment has probably arrived. But just to be sure, here are the six surefire ways to know you’re ready for something new.

Sign 1: One person knows where everything actually is

The real operations layer of your business lives in spreadsheets, paper, and a handful of people's heads. The official system, whether that's an ERP, a planning tool, or a stack of point apps held together with manual exports, is a record of what was supposed to happen, not a system for what's happening now. You can tell because when one person is out, the team struggles to function because they've lost access to the only working version of the data.

This is the shadow operating system most manufacturers run on. It shows up as the planning spreadsheet your team has been refining for two years, the WhatsApp thread between procurement and the warehouse, the printed sheet that never makes it back into the official system.

Your team isn't sloppy, the problem is that your software couldn't represent how the work actually happens, so your team built a workaround. The workaround is doing the job; the system is the problem.

Sign 2: Changing the system is harder than changing the business

A new product line, a new shipping partner, a new pricing rule. Each one is supposed to be a small adjustment. In a legacy ERP it becomes a quarter-long project, requiring scoping, an integrator, a quote, an internal approval, and a rollout plan. In a spreadsheet stack, it becomes a data migration nobody wants to own. The change has to be copied across five files, the new rule has to be documented somewhere people might find it, and you have to hope nothing else breaks.

These are opposite failure modes that produce the same outcome. Once data is locked into a rigid schema, changing it is hard. Legacy ERP vendors in particular built business models on top of that hardness. Case in point: SAP made 12-to-24-month ERP projects the norm.

If your operating tempo has accelerated and your system's cycle time hasn't, you're going to lose ground to competitors who can adapt faster.

Sign 3: Your team feeds the system instead of running the business

In a typical 100-person manufacturer, between 10 and 15 full-time equivalents are absorbed into system-adjacent work: data entry, reconciliation, fixes, spreadsheet workarounds, and chasing missing fields. That's 15,000 to 22,000 hours a year of skilled labor compensating for the system instead of creating value with it.

You'll see it everywhere once you look. "We need more headcount" becomes the only way to scale. Your operators have a side job filling in fields after the fact. Finance reconciles the same data three times across three modules.

The cost shows up in two places:

1. The salary line, where roughly €600k to €900k per year of pure salary cost generates no value, and
2. The second-order cost of brilliant people spending their time on work the software should be doing.

Sign 4: Reporting is always too late to act on

This week’s planning decisions are made on last week’s data. By the time the dashboard or the master spreadsheet is reconciled, the situation has changed. So the people closest to the work stop trusting the system entirely and run the business off side reports, which means the official reports are wrong by the next refresh, and the cycle continues.

The signal here isn't bad reporting tools, it's that the system is structurally a system of record. It stores what already happened. By the time the data is clean enough to analyze, the moment to act on it is gone.

A modern operations system does the analysis as the data lands and has the autonomy and authority to take action. For example, at L'Atelier du Ferment, Bonx generates manufacturing orders and procurement suggestions against sales, shelf life, and cold storage capacity. The system doesn't wait for a planner to compile next week's plan. It already knows the plan and has the agency to move things forward.

Sign 5: Every plan to fix it looks like a year-long project

For ERP shops, “fixing the problem” has always meant 12 to 24 months on the calendar, an integrator on retainer, and the threat of a phase three still running 18 months after kickoff. For spreadsheet shops, getting onto a more sophisticated ERP has historically carried the same price tag, which is why so many of them haven't done it. Both groups look at the cost of fixing the problem and decide to live with it for another year.

That math is finally changing. ERP implementations were normalized at 12 to 24 months because the underlying technology required it, and that's no longer true. Bonx deployed an ERP at a fast-growing food manufacturer in 42 days, in time for a national TV broadcast that 10x'd order volume overnight, with traceability intact on every package.

If you've been delaying the fix because the fix sounded worse than the problem, the assumption underneath that decision is out of date.

Sign 6: Good people are leaving

This one is the hardest to admit because it lands on culture before it lands on operations. You'll hear it in exit interviews and in hallway frustration. The work isn't what they signed up for. They came to plan production, manage suppliers, or run a line. They're spending their day on data entry.

For the generation entering manufacturing now, this is a non-starter. A capable 26-year-old isn't going to spend four decades typing into Excel or an ERP screen designed in the 1990s. The reality is, you just can't compete for that talent/

If your turnover in operations roles is outpacing your industry peers, and the reason your best people give in private is "the system," the system is now your recruiting problem.

The bottom line

Whatever you're using right now, whether that's an ERP, spreadsheets, or a mix of point tools, was built to be a system of record, meaning it stores states. The work of acting on those states, including deciding, executing, following up, and reconciling, has always belonged to humans.

That model was the only option for 40 years. It is no longer the only option, and it is the reason all six signs above sound like the same problem in different costumes. The system can't do work.

The next ERP isn’t a faster legacy ERP, and it isn’t a more polished spreadsheet stack. It's a system that does the work the old one made you do, and surfaces decisions to your team only when the decision actually needs a human. Operators move from execution to oversight, headcount stops scaling linearly with revenue, and the shadow operating system goes back into the actual operating system, where it belongs.

If the signs above are familiar, the question isn't whether to change. It's whether your business can absorb another year of the gap between what the system does and what the work actually requires.

Most can't. They just don't realize it until the shadow system finally gives way.