Excel and shadow systems: when are they symptoms, and when do they reveal operational gaps?

When a company's formal system and the way work actually gets done persistently diverge, that is an operational signal worth investigating.

A shadow system is therefore not automatically a “bad user habit”.

Nor is it automatically an ERP fault.

It is a diagnostic clue.

What is a shadow system?

In Kyntrax terminology, a shadow system is a solution outside the formal business system that persistently handles data, rules, calculations or workflows needed for the company's actual operations.

Examples include:

Excel or Google Sheets;
a personal Access database;
approvals managed by email;
a Teams or WhatsApp conversation;
a separate stock list;
personal notes;
a manually maintained production or delivery schedule;
a custom report;
or simply a rule held in an employee's head.

A workaround is a slightly different concept. It occurs when an employee departs from the formally prescribed way of working to complete a task at all or achieve a particular objective.

The two often overlap.

An occasional workaround can become a permanent spreadsheet. Over a few years, that spreadsheet may even grow into critical business infrastructure.

Why do these solutions emerge?

There is no single universal answer.

The cause may be technological:

the ERP genuinely does not support a required function.

It may also be operational:

the process is not properly defined.

It may be an information problem:

the employee does not trust the data shown in the ERP.

It may be an access problem:

the necessary information is not accessible in time.

It may be a speed problem:

the official procedure is too slow to handle a real business situation.

And it may be an entirely legitimate local solution.

Modern process research therefore increasingly treats workarounds as more than simple “rule-breaking”. A 2026 Business & Information Systems Engineering study finds that investigating recurring workarounds can reveal persistent misalignments between processes, information systems and organisational goals. The research also stresses that not every workaround is harmful and that a one-off technical workaround should not automatically be treated as a structural problem.

This is very close to how Kyntrax examines operations:

our first question is not who departed from the system, but why they needed to.

When does Excel become a meaningful operational signal?

It is especially worth looking deeper when the same business information exists in several places.

For example:

ERP: 430 units in stock.

Excel: 390 units.

Warehouse list: 410 units.

Actually available to sell, according to the salesperson: 350 units.

The most important question here is not which program is “better”.

The first question is:

what do the four numbers actually mean?

One may represent physical stock.

Another may represent book stock.

A third may already deduct reservations.

The fourth may account for a quality or customer restriction that none of the systems handles.

In a 2016 multiple-case ERP study, 64% of the shadow systems identified had some data or functional connection with the ERP system. This is not a general percentage for all businesses, but a finding from that particular study. It illustrates how formal ERP and local systems often handle different parts of the same operation.

A key question about shadow systems: where is the source of truth?

Suppose a salesperson uses Excel to promise a delivery date to a customer.

Purchasing works from the ERP.

The warehouse keeps its own list.

The manager looks at a fourth report.

In this situation, it is not enough to say:

“Everyone must use the ERP.”

The company needs to know:

which information is authoritative for which business question?

The same system is not necessarily the authoritative source for every question.

The problem starts when nobody can answer that clearly.

This connects directly to the Kyntrax Insights topic Information flow blind spots: information may exist somewhere without being available at the next decision point with the right content, at the right time and under clear responsibility.

What needs to be investigated?

Investigating a shadow system should not begin by deleting the spreadsheet.

First, trace:

what triggers it → who uses it → what data goes into it → where that data comes from → who changes it → what it is used for → which decisions depend on it → where the result goes.

Then ask:

Why does this not happen in the official system?

Is the information available there?

Is it reliable?

Is it available at the right time?

Does the official process support the real situation?

Are we handling an exception or everyday operations?

What would happen if we removed the spreadsheet tomorrow?

That last question is often particularly revealing.

If a critical process would stop, the spreadsheet is no longer just a personal aid.

It is part of the actual operation.

An illustrative example

A trading company's ERP contains its orders.

However, the sales team keeps a separate spreadsheet recording which orders are “truly urgent”.

The easy first reaction would be:

The salespeople are not using the ERP properly.

But investigation might reveal that the company has no consistent business rule defining urgency.

The ERP is therefore not malfunctioning.

Nor are the employees necessarily making mistakes.

A business decision rule is missing.

In this case, the spreadsheet is not itself the cause.

It makes the cause visible.

This is an illustrative example, not a customer result.

When is a separate spreadsheet not a problem?

When its role is clear.

For example, it is used for a one-off analysis, scenario planning or an ad hoc calculation, and its output does not become an uncontrolled second source of business data.

A spreadsheet is not a shadow system simply because it exists.

It may become an operational risk because of the role it plays in the company's actual operations.

What does the research say?

The literature does not treat workarounds as uniformly negative. Recent research suggests that recurring workarounds may indicate process or system fit problems as well as local innovation. Investigating the underlying cause therefore matters more than automatically eliminating the deviation itself.

A 2026 action research study covering 80 sites observed an association between parallel systems and certain governance deficiencies. The author explicitly recommends against automatically interpreting parallel records as an implementation failure. However, the research comes from a single organisation and has no control group, so its findings cannot automatically be generalised to every company.

How does this connect to Kyntrax?

Kyntrax Mirror does not begin by deciding which systems should be replaced.

It investigates how the business actually operates.

A spreadsheet, personal list or manual correction used alongside the ERP is therefore more than simply “improper system use”.

It is evidence of a point that needs to be understood.

Mirror's task is to establish whether it represents:

a process problem, an information gap, a system limitation, incorrect configuration, an access issue, exception handling or a justified local solution.

Only then is it worth deciding whether to eliminate it, integrate it, change it — or make it official.

What does this article establish?

The presence of a shadow system alone does not prove an ERP fault, a process problem or user error.

It establishes a connection between formal operations and actual work that is worth investigating.

Professional background

Outmazgin, Soffer & Hadar: Leveraging Workarounds for a Problem-Focused Improvement of Business Processes, Business & Information Systems Engineering, 2026. Study

The Relation of Shadow Systems and ERP Systems — Insights from a Multiple-Case Study, Systems, 2016. Study

Sid Ahmed El Hebri Harid: The visibility–protection gap in enterprise resource planning governance: shadow systems and information quality degradation, Journal of Enterprise Information Management, 2026. Study