You Instrument Your Website. Why Not the Software Your Business Runs On?
Most businesses that have invested in GA4 think of it as a marketing tool — something you put on a public-facing website to understand how visitors behave, where they come from, and whether they convert. That is a legitimate and valuable use. It is also only half the picture.
The same logic that makes analytics powerful on a public website applies equally to internal applications. Your CRM. Your ERP. Your project tracking system. Your custom workflow tools. Your content management interface. Every one of these is a software environment that real people use every day — and right now, your understanding of how they use it is almost certainly based on anecdote, not data.
Someone mentioned in a meeting that the approval workflow takes too long. A manager noticed that a particular report never gets pulled. Help desk tickets keep coming in about the same screen. These are signals, but they are incomplete ones — filtered through whoever happened to say something, on the day they said it, about the specific thing that frustrated them enough to mention it.
GA4 on an internal application gives you what the cameras analogy describes for a public website: a complete, uninterrupted view of every user, every workflow path, every friction point — not a sample, not a complaint, not an impression. Everyone. Every time.
Internal Feedback Is the Most Incomplete Data Source in Most Organizations.
With a public website, you have at least some external signals — customer feedback, support tickets, conversion rates, search performance — that give you a partial picture of what is working and what is not. Imperfect, but present. With an internal application, those external pressure valves largely do not exist.
Employees adapt. They find workarounds. They learn which screens to avoid and which sequences to follow to get through a process without triggering the confusing part. They stop reporting friction because reporting it never seemed to change anything. The institutional knowledge of how to navigate a broken workflow lives in people’s heads — and when those people leave, it leaves with them.
What gets reported is the tip of a very large iceberg. The duplicate data entry that adds ten minutes to every record update. The report screen that requires six clicks to reach and two to close. The approval step that creates a bottleneck every time a specific manager is out of office. The feature that was built at significant cost and is used by almost nobody. None of these surface reliably through normal feedback channels. All of them are immediately visible in properly instrumented analytics.
- Employees adapt around broken workflows rather than reporting them consistently
- Workarounds become institutional knowledge that lives in people’s heads, not documentation
- Help desk tickets capture acute failures — chronic inefficiencies go largely unreported
- Analytics captures what actually happens, not what people remember or choose to mention
Feature Usage, Workflow Drop-Off, Navigation Patterns, and Where Time Goes.
Properly configured GA4 on an internal application surfaces the same categories of insight it does on a public website — applied to your internal users and their workflows instead of your customers and their purchase paths.
Feature usage tells you which parts of the application are actually being used and which are being ignored. A module that was built to streamline a process but has near-zero engagement is either misunderstood, redundant with something else, or so difficult to use that people gave up. You cannot tell which without the data — and the fix for each is completely different.
Workflow drop-off shows you where multi-step processes fall apart. If a significant percentage of users start a workflow and abandon it at a specific step, that step has a problem — whether it is confusing, slow, requires information users do not have at that point, or simply asks too much. Every abandoned workflow is either an incomplete task or a workaround being used instead. Both cost time and introduce error.
Navigation patterns reveal whether users are moving through the application the way it was designed to be used or whether they have developed their own paths — which is almost always a sign that the designed path has friction the development team never saw. And session duration and task completion time show you, in aggregate, how long things actually take versus how long they should.
- Feature usage — what gets used, what gets ignored, and what was built but never adopted
- Workflow drop-off — exactly where multi-step processes fail and users abandon
- Navigation patterns — how users actually move through the application versus how it was designed
- Task completion time — aggregate data on how long things take and where time is being lost
- Error and dead-end rates — screens that consistently send users backward or out
Companies Spend Millions Implementing Systems Nobody Knows How to Use Efficiently.
Enterprise Resource Planning systems — platforms like JD Edwards, SAP, Oracle, and Microsoft Dynamics — represent some of the largest software investments a business makes. They are deeply customized, heavily integrated, and almost universally described by the people who use them daily as difficult, counterintuitive, and slower than the processes they replaced.
The implementation cost is scrutinized carefully. The ongoing user efficiency is almost never measured. Organizations spend years and significant resources getting these systems live and then operate them based entirely on anecdotal feedback about what is working and what is not. The assumption is that if the system is running and producing output, it is performing. That assumption misses an enormous amount.
Analytics instrumentation on an ERP or similarly complex internal system can identify where users are spending disproportionate time relative to task complexity, which workflows have the highest abandonment or error rates, which custom modules are delivering value and which are being bypassed, and where retraining or interface adjustments would produce measurable efficiency gains. That is not a small opportunity — for large organizations, the productivity impact of eliminating even modest friction across thousands of daily user interactions compounds into significant recoverable time and cost.
The question is not whether your ERP or internal system has inefficiencies. Every complex system does. The question is whether you are measuring them or just living with them.
This Is UX Research, Not Surveillance. The Distinction Matters.
The natural first reaction from HR and IT when analytics on internal applications comes up is some version of: are we monitoring employees? It is a fair question and it deserves a direct answer.
The goal of instrumenting an internal application is not to watch individuals — it is to understand where the software is failing the people using it. The data is aggregate behavioral data: how many users completed this workflow, where did drop-off occur, how long does this process take on average. It is the same kind of data a UX researcher collects when evaluating a product. The subject of the research is the application, not the employee.
Configured correctly — without personal identifiers, focused on workflow and feature interaction rather than individual sessions, and governed by a clear internal policy — analytics on internal applications is straightforwardly a tool for making the software better and the people using it more efficient. That framing, established clearly from the start, tends to address the concern. The alternative — continuing to operate complex internal systems with no behavioral data — is not a more respectful position. It is just a less informed one.
Your Internal Applications Are Running. The Question Is Whether They Are Running Well.
Every internal application your business depends on is a system that real people interact with for hours every day. The cumulative effect of that interaction — efficient or inefficient, intuitive or confusing, well-adopted or quietly bypassed — shows up in your operational performance whether you measure it or not.
The difference between measuring it and not measuring it is the same difference as having cameras in your store versus managing by what you happen to notice on the days you are there. The activity is happening either way. The question is whether you have the complete picture or just the parts that surfaced through complaint and coincidence.
GA4 is free. The instrumentation work is straightforward for applications built on standard web technologies and manageable for more complex environments. The data it produces — complete, continuous, aggregate behavioral data on how your internal users actually interact with your systems — is the kind of intelligence that used to require expensive UX research engagements to approximate. Now it is available as infrastructure.
You already know analytics belongs on your public website. The same argument, with the same logic, applies to every application your business runs internally. Complete data beats anecdote. Every time.