Blog
From Excel and email to an application: the pattern behind digital transition
· 4 min read
A spreadsheet with a tab for every month. An inbox with hundreds of notifications that someone retypes by hand into that spreadsheet. A colleague who knows exactly which formula must never be touched. Sound familiar? At many SMEs, a process that once started small now runs on a patchwork of Excel and email, and that process has long outgrown the system underneath it.
What stands out is that the problem rarely lies with Excel itself. Excel is a perfectly good calculator. The problem starts the moment a spreadsheet has to do something it was never designed for: letting several people enter data at the same time, keeping track of a status that has to be the same for everyone, or repeating an action that happens hundreds of times a year. Email then becomes the vehicle for data that really belongs in a system, and that creates exactly the kind of chaos that eventually calls for a different approach.
Why Excel and email break down as you grow
Three things go wrong once a process grows while still running on Excel and email. First, there is no single source of truth: everyone works with their own copy of the file, and nobody is quite sure whether they have the latest version. Second, every handover between people costs time and invites mistakes: someone has to read an email, pick out the right data and enter it somewhere by hand. Third, there is no control: no system enforces a mandatory field, flags a duplicate entry or automatically triggers the next step.
As long as the volume is small, discipline keeps this manageable. As soon as the number of people or transactions grows, that same discipline becomes the bottleneck. That is precisely the point at which it becomes worth asking whether a process needs an application rather than yet another version of the same spreadsheet.
The pattern behind a successful move
From our own build work with OutSystems, the low-code platform which, in OutSystems' own words, lets you build applications with minimal hand-written code, the same pattern keeps coming back whenever a process moves from Excel and email to an application.
It starts with taking stock before anything is built: which screens are needed, which roles exist, and which other systems need to be connected. Skip this step and you often discover halfway through the build that an integration or a role is missing, one that could have been known from the start.
Next comes a clickable sketch with realistic sample data, in both phone and desktop formats, before any actual building begins. It shows a client straight away whether the direction is right, instead of finding out after weeks of build time that a screen does not match how people actually work.
The riskiest part is then proven first, not last. An integration with an external system, or a process that a lot depends on, is made to work on its own first. That way you run into any setbacks early, rather than in the final week before go-live.
After that, nothing is delivered in one go. Components are built and measured one at a time before the next one starts. Smaller pieces are easier to check and adjust than one big whole that only gets tested at the end. And roles and permissions are set up from day one, not bolted on afterwards, because a screen without a clear role otherwise turns into a forgotten loose end later on.
A concrete example
This pattern is not just theory. At MEA Beveiliging, the installation process ran entirely through email: every installation notification was processed by hand, with 50+ dealers and hundreds of vehicles a year. That led to chaos and delays. The solution was a fully automated dealer and certification platform in which dealers manage their own installations, payment flows run automatically and certificates are generated instantly. The results are on our projects page: 50+ dealers, 768 vehicles, and the process is now 100% automated.
What this case shows is not that an application is automatically better than a spreadsheet. It shows that a process that depends on manual handovers between people, once the volume is high enough, benefits from a system that takes over those handovers while also providing a verifiable, single record.
What does this mean for your business?
If you recognise the pattern from this article in your own business, these are the steps to start with.
First, map out where in your organisation a process runs through Excel and email while the volume has become too large for discipline alone. Look at the number of people involved, the number of times a week it is repeated, and how often an error creeps in through manual retyping.
Then, before you even think about building, take stock of the screens, roles and integrations that process needs. This stops you from discovering halfway through that something is missing.
Next, start with a clickable sketch instead of building straight away. A sketch with realistic sample data shows whether the direction is right before any build time goes into it, and that is cheaper than rebuilding afterwards.
Finally, decide which part carries the most risk, and prove that part first, on its own. That prevents a setback from coming to light just before go-live, when changing course is most expensive.
Would you like to know whether your process lends itself to this approach, or first see what such a move looks like in practice? Get in touch and we will look at the pattern in your situation together.