5 signs your ERP has a shadow system problem

Your ERP went live. The training happened, the old system was switched off, and the project was closed with a good report.

A few years later, you walk through the plant, or open a management pack, and notice something. The numbers people actually trust don’t come from the ERP. They come from a file, a message thread, or the one person who knows how to put it all together.

That is a shadow system. It’s the set of spreadsheets, side apps, chat groups, and personal workarounds that grow at the edges of an ERP to do the work the ERP doesn’t. Researchers who study ERP use in smaller companies call these “feral systems”: data held in tools nobody officially sanctioned, even though it belongs in the ERP.

5 signs your ERP has a shadow system problem
Xaas genie consultant
Written by

Xaas genie consultant

Shadow systems aren’t a sign that your people are careless. They are a sign that real work has more steps than the system was set up for. But left alone, they cost time, hide risk, and make every decision slower.

Here are five signs that yours has grown larger than you think.

Sign 1: Your most trusted numbers come from a file, not the ERP

Before the weekly review, which report does everyone open? If the answer is a spreadsheet that someone prepared, and not the ERP report, that’s the clearest sign.

This is common, and it’s not new. A manufacturing survey by Sikich found that 53% of the manufacturers polled used manual processes to measure their key performance indicators, compared with 26% who used applications such as an ERP. That survey dates from 2015, so it’s an old reference point. More recent finance research points the same way. PwC’s 2026 CFO survey for South-East Europe found that 56% of organisations still use spreadsheets to reconcile and consolidate financial and management data, even after investing in ERP.

The quick test: Pick your last management review. Count how many numbers on the slides came straight from the ERP and how many came from a file someone built.

Sign 2: The same data lives in more than one place

An order is in the ERP. It’s also in the dispatch tracker, the sales team’s sheet, and the finance team’s reconciliation file. Each has a slightly different version.

Research on ERP use in smaller companies describes the pattern clearly: even where the ERP has built-in reporting tools, users export data into spreadsheets to reshape it, and then keep working from the copy. Once the copy exists, it starts to drift from the source.

You can spot this by listening. If the first thing said in a meeting is “which number is right?”, you have duplicate data.

The quick test: Choose one figure, such as yesterday’s dispatches. How many files hold it? Do they all agree?

Sign 3: Decisions travel by WhatsApp and email

Approvals sit in someone’s inbox. The status of an urgent order lives in a group chat. A quality hold is agreed on a call, and then typed into the ERP the next morning, if at all.

This is the version of the problem that’s hardest to see, because it leaves no file behind. But it shows up in a simple way: when something goes wrong, someone says “let me check the thread.”

In our own work with Radiant Anodisers, an aluminium manufacturer, the daily reconciliation meant cross-checking ERP data, Excel sheets and WhatsApp messages, and it still didn’t give one reliable view of the operation. That is a shadow system in its purest form: three places that each held part of the truth.

The quick test: Ask for the status of one live order. Count how many people, files, or chats you need to consult to answer.

Sign 4: The whole process depends on one or two people

Every plant has a version of this. “Only Rajesh understands that sheet.” When he’s on leave, month-end gets nervous.

This is the risk side of shadow systems. A spreadsheet that has quietly become a process has no audit trail, no access control, and no backup of how it works. It also carries the usual spreadsheet risk. Studies of spreadsheets in real use, many of them audits by commercial firms, found errors in about 94% of those examined, and research on human error suggests that cell error rates of 1% to 5% are normal. A file that feeds a decision will almost certainly contain a mistake somewhere. It just won’t announce it.

The quick test: Who could rebuild your most important file from scratch if the person who maintains it left tomorrow?

Sign 5: Every new need becomes a new workaround

A manager needs a report that the ERP doesn’t produce. Asking for it means a formal request, a developer, and a wait. Building it in Excel takes an afternoon. So the sheet wins, and then it becomes part of someone’s weekly routine, and then it’s too embedded to retire, even if the ERP equivalent is eventually built.

This is how a shadow system grows. It isn’t one big decision. It is a long series of small, reasonable ones. The giveaway is that the number of files keeps rising, years after go-live, while the ERP stays the same.

The quick test: How many new spreadsheets, trackers, or chat groups appeared in the last twelve months? Does anyone keep a list?

How many of the five sounded familiar?

This is a rough count, not a formal score.

  • 0 or 1: you probably have a few workarounds, and they’re manageable.
  • 2 or 3: shadow systems are doing a meaningful share of the work. It’s worth mapping them.
  • 4 or 5: the real operating process sits outside the ERP. Decisions are being made on data you can’t easily audit.

Most manufacturers we speak to are somewhere in the middle. That isn’t a failure. It’s the normal result of a system that has been stretched.

What to do about it

You don’t need a transformation programme to begin.

  1. List what exists. Ask each function for the files, trackers, and chat groups the business really runs on. The working ones, not the official ones.
  2. Map where information stops. For each one, note which system the data should have stayed in, who feeds it by hand, and who relies on it.
  3. Start with the one that costs the most. Not the biggest file, but the one with the most hand-offs, the most rework, or the most risk.
  4. Fix it around the ERP, not instead of it. The aim is to connect what you already have, so information stops stopping.

That last point matters. At Radiant, the ERP stayed in place. It was connected, through APIs, to one operating layer built on Zoho Creator, and the manual reconciliation across the ERP, Excel, and WhatsApp ended. The result was one real-time view of production, stock, quality, and dispatch.

Find out where your information stops

At XAAS Genie, we built the ERP Friction Index to do the first two steps for you. We study how companies like yours operate, and we come back with a score out of 100 and a one-page card showing three likely operational gaps, two integration opportunities, one AI opportunity, one quick win, and an estimate of the impact. Everything in it is our best reading from the outside, so we check it with you before you act on it.

Request Your ERP Friction Assessment