When ROI Exists on Paper But Not in the Operation

The business case was approved. The numbers were solid. The projected ROI was over 200%. Six months after implementation, the process still takes the same time, the team has created workarounds, and nobody can explain exactly why the savings never materialized.

This scenario is more common than it should be. And the problem is almost never the original analysis. The problem is everything that happens between the spreadsheet and reality.

Why business cases fail

A business case is built under controlled assumptions: a certain volume of cases, a certain execution time, a certain percentage of automation. Those assumptions are reasonable at the moment of making them. They are also fragile.

Volume changes. People adapt the process to their own logic. Edge cases appear that were not in the model. The tool behaves differently in real use than in the pilot. And the expected savings, which existed on paper, do not materialize because reality is not a spreadsheet.

The business case is not wrong. It is incomplete. It models what should happen but does not anticipate how execution will respond to friction. ⚠️

The gap between design and execution

There is a gap between how a process is designed and how it runs in practice. That gap has a name in industrial engineering: it is called the implementation gap. In automation projects, it is the space where expected ROI evaporates.

The implementation gap is produced by several converging factors: lack of clear ownership of the automated process, poorly defined exceptional cases, insufficient training for the team that will use the tool, and absence of metrics that make the gap visible in real time.

The most damaging part is that this gap is often invisible for months. The tool is running. No one says it is failing. But the results are not there. And when someone finally asks why, there is no clear answer, just a mixture of 'we are still adapting' and 'the process was more complex than expected.'

Lack of adoption and follow-up

Automation without adoption is not automation. It is software running alongside a parallel manual process that the team prefers because it feels safer, because they understand it, or because nobody trained them on the new one well enough.

Adoption is not a training problem. It is a design problem. A process that requires the team to change their behavior without understanding why, without seeing an immediate benefit, and without clear support, will not be adopted, regardless of how good the tool is.

Follow-up is the other missing ingredient. Most automation projects have a launch but no systematic review at 30, 60, or 90 days. By the time someone looks at the numbers, the implementation gap is already established and normalizing. 🌀

Poorly defined indicators

'It works better' is not an indicator. 'Approval cycle reduced from 5 to 1.5 days' is.

One of the most frequent causes of the gap between paper ROI and operational ROI is that the success metrics were never defined with enough precision. When there is no measurement, there is no feedback. When there is no feedback, there is no correction. And when there is no correction, the initial drift normalizes as the new standard.

Defining metrics before implementation (not after) is what allows deviations to be caught when they are still small and correctable.

👉 APM is not just another methodology: it’s the system that makes automation work.

How to close the gap, automation project management

The gap between paper ROI and operational ROI is a management problem, not a technical one. The answer is not better software or more sophisticated analysis. It is applying rigorous project management to each automation initiative, from the correct diagnosis of the problem to be solved through to post-implementation measurement.

In practice this means: defining a clear process owner, establishing measurable metrics before the first line of automation, scheduling systematic reviews, and treating each automation as a living project rather than a one-time delivery.

The variables that generate this gap, and how to account for them in the analysis, are explained in detail in the complete automation ROI variables. Understanding them in advance is what prevents the gap from appearing in the first place.

The ROI that exists on paper is a possibility, not a promise. The difference between the two is execution. And execution is never managed by itself. ⏳