Skip to content

When Excel is no longer enough: five signs you need your own application

When an Excel spreadsheet reaches its limits: five signs it should become an application, when Excel remains the right choice and how the switch works.

David KreutzerManaging Director, Viewconic GmbH5 min read

Why Excel works for so long

In many companies, Excel is the unofficial operating system. Quotes, shift planning, inspection records, commissions, stock lists: what began as a quick overview controls a core process years later. That is no argument against Excel. It shows how flexible the tool is and how quickly a department can build a solution with it on its own.

At some point the balance tips. The spreadsheet is no longer an aid but a bottleneck. The following five signs help you judge whether that point has been reached. The more of them apply, the more likely it is that the spreadsheet should become an application.

Sign 1: several people work in the same file

As long as one person maintains the spreadsheet, it works. As soon as several people enter data at the same time, the problems begin: the file is locked, someone carries on in a copy, and in the end there is “Planning_final_v3_new.xlsx” next to “Planning_final_v3_new_MK.xlsx”. Which version applies is known only to whoever saved last.

Collaborative editing in the cloud eases this but does not solve it. There is still no clear responsibility per record and no workflow that defines who may change what and when.

What an application does differently: Everyone works on the same data, at the same time and without copies. Every record has a status and an owner.

Sign 2: only one person understands the spreadsheet

Nested formulas, hidden columns, macros that someone wrote years ago: in many companies there is exactly one person who knows why the spreadsheet works. When that person is on holiday, the process waits. When they leave the company, the knowledge goes with them.

This risk appears on no balance sheet until it materialises. Often it is only noticed when a macro stops running after an Office update.

What an application does differently: The rules are in the code and in the documentation, not in one person’s head. Tests ensure that a calculation is still correct after a change.

Sign 3: errors are noticed late

An overwritten formula, a date in the wrong format, a typo in an article number: Excel accepts almost any input. Errors therefore often show up only where they cost money, in a wrong invoice, a duplicate order or a report that does not add up. Who changed what and when can hardly be traced afterwards.

What an application does differently: Inputs are checked when they are entered. Required fields, value ranges and dependencies between fields are defined. Every change is logged with time and person.

Sign 4: data is retyped or copied

An order arrives by email, is transferred to the spreadsheet, from there to the ERP and at month end to accounting. Every transfer costs time and is a source of errors. If a considerable part of the work consists of moving data from one system to another, the problem is not the diligence of the staff but the solution.

What an application does differently: It talks directly to ERP, CRM or accounting through interfaces. Data is captured once and used where it is needed.

Sign 5: permissions and evidence become important

As soon as personal data, prices or terms are in the spreadsheet, the question arises of who may see what. In a file, as a rule everyone who can open it sees everything. When customers, auditors or your own data protection officer ask who had access and what was changed, it becomes difficult at the latest.

What an application does differently: Roles define who sees and edits which data. A change log provides the evidence that a spreadsheet cannot.

When Excel remains the right choice

Not every spreadsheet has to become an application. Excel remains the better tool when

  • one person maintains and analyses the data,
  • the task is a one-off analysis or a what-if exercise,
  • the structure is still changing constantly because the process is only taking shape,
  • errors have no expensive consequences.

Excel also keeps its place after a replacement: for analyses based on an export from the new application. What is replaced is the process, not the program.

How the switch works

A spreadsheet that has grown over the years is the best specification there is. It shows which data is needed, which rules apply and where the special cases are. That is why the switch with us begins with a joint analysis of the spreadsheet and the process behind it, together with the people who work with it every day.

The familiar steps follow: we build in an AI-assisted way in milestones, test automatically and present a working version for acceptance after every iteration. The data from the spreadsheet is cleaned and transferred. For a transition period, the old and new solutions can run side by side until your team has switched over with confidence. You can find more about the process and typical starting points under Replace Excel, and for Access databases under Replace Access.

What you can clarify before the first conversation

You do not need a requirements specification to talk about a replacement. Answers to a few simple questions help, and they are usually already known in the department:

  • Who works with the spreadsheet, in what role and how often?
  • Where does the data come from, and where does it go afterwards? Which systems are involved?
  • Where do errors or queries occur today, and what do they cost in everyday work?
  • Which reports do management or the customer need regularly?
  • Which special cases are known only to the person who maintains the spreadsheet?

From these answers, the analysis produces a shared picture of what the application has to do and what deliberately stays out. Often it also turns out that not the whole spreadsheet has to be replaced, but first the part where most errors arise.

What the switch costs

It depends on the same factors as any custom software: business rules, interfaces, data quality and permission requirements. How we estimate effort and what AI changes about that estimate is described in the article What does custom software cost – and what does AI change about estimating? You can find an overview of what we offer under AI-assisted software development.

If you thought of a particular spreadsheet while reading, it is usually a good starting point for a first conversation. Just bring the file with you.

Frequently asked questions. Briefly answered.

Is your own application worth it instead of Excel for small teams too?

It depends less on team size than on the process. If several people work on the same data at the same time, errors become expensive or data is retyped between systems, an application is worth considering even for a few users.

Can we transfer our existing data?

As a rule, yes. The data is cleaned, mapped to the new structures and checked before the switch. The effort depends less on the volume than on how consistently the spreadsheet has been maintained over the years.

Do we have to give up Excel completely afterwards?

No. For reports and one-off analyses, Excel remains a good tool. An application can export data as a spreadsheet at any time. What is replaced is the process, not the program.

How long does it take to replace an Excel solution?

The analysis clarifies that, because rules, interfaces and data quality determine the effort. We plan in milestones, so you can review a working version early and replace the old spreadsheet step by step.

Does this also apply to Access databases?

Yes, often even more so. Access solutions frequently depend on the one person who built them and on a specific Office version. The signs in this article apply to them just as much. What Access adds is covered on the page “Replace Access”.

This article was produced with the support of automated research and AI.

Discuss a project

Do you have a project in mind? Let us talk about your process.

In the initial conversation, we clarify whether your project suits us. After the joint analysis, you receive a clear framework with milestones – and after that you see a working version in every iteration.

Further reading

More from the blog. Articles on related topics.