Stop measuring clicks, start measuring time

# A new metrics playbook for enterprise UX

The absurd theatre of bad metrics

I've spent the last few weeks trying to get our design team's playbook in order on our internal wiki. It's a soul-destroying task, mostly because you quickly realise that half of the 'rules' you've all been following are based on nothing more than outdated habits and imported nonsense.


And nowhere is this more obvious than in how we measure success.

Every month, the entire company gathers for the all-hands meeting, a ritual led from the very top. And every month, we are shown a series of charts. They all point, as they must, up and to the right. We all nod along as the latest meaningless metrics are paraded before us: page views per module, the general number of support tickets, and other figures that bear only a passing resemblance to actual user success.

It is a magnificent piece of corporate theatre. And it is almost completely useless.

"Engagement" is a dangerously wrong idea

This plague of "success theatre" is fuelled by a single, seductive, and utterly useless word: "engagement."

We are told to celebrate when users spend more time in our app. We are told to cheer when they click more things. This is the conventional wisdom, handed down from the consumer-app cathedrals of Silicon Valley. And for enterprise software, it is dangerously wrong.

Celebrating that an employee is "engaging" more with your mandatory internal software is like a hammer manufacturer celebrating that it now takes twice as long to hammer in a nail. It's a sign that you have failed, spectacularly.

The user of a social media app is a customer. The user of enterprise software is an employee. They are a captive audience. They are not there to be "delighted" or "engaged." They are there to do a job, a job that is often stressful, complex, and time-sensitive. Our software is a tool, and the only measure of a good tool is how efficiently it gets the job done.

Stop measuring clicks.png

In the world of enterprise software, there is only one metric that truly matters: Time.
Every decision we make should be judged against a single, brutal question: does this give the user back their time, or does it steal more of it? If you want to build a metrics strategy that actually means something, you need to throw out the consumer playbook and start measuring what matters.
 

"Celebrating that an employee is 'engaging' more with your mandatory internal software is like a hammer manufacturer celebrating that it now takes twice as long to hammer in a nail."

 

A playbook for enterprise UX

Here’s a new one.

1. Measure time on task, not session duration

Stop measuring how long people spend in your app. Start measuring, with a stopwatch if you have to, how quickly they can get in, do the specific task they came to do, and get out. A 20% reduction in the time it takes to complete a core task is a real, quantifiable productivity gain that the business will actually understand.

 

2. Measure User Error Rate, Not Clicks

Stop celebrating clicks. Start counting the mistakes. An error in a stowage plan or a miscalculated off-hire claim isn't a minor inconvenience; it's a potentially multi-million-pound problem. A high click count is often a symptom of a user desperately clicking around a confusing interface, trying to figure out what to do. A low error rate is a sign of a clear, intuitive design that is actively preventing costly failures.

 

3. Measure Task Success Rate, Not "Engagement"

This is the simplest and most important metric of all. Forget vague notions of "engagement." Ask a more direct question: could the user actually do the thing they came to do? Yes or no. If a Chief Officer tries to update a safety compliance form and gives up halfway through, that is a 0% task success rate. That is failure. It doesn't matter how long they "engaged" with the pop-ups.

Speaking the language of business value

When you start measuring time, errors, and success, something magical happens. The conversation with your stakeholders changes. You are no longer having subjective arguments about the colour of a button. You are having objective, evidence-based discussions about efficiency, productivity, and risk. You are, in other words, finally speaking their language.

So, let's stop importing useless metrics from the consumer world. Let's stop celebrating engagement and start celebrating efficiency. Let's give our users back their most valuable asset. Their time.

Other writing

Before I designed software, I designed physical things. In that world, a mistake meant a furnace full of money. A look at the 'one-way door' decisions…
The design world is collectively losing its mind over a great, job-destroying monster called Artificial Intelligence. The debate is a tedious…
In shipping, a simple change in the weather can light the fuse on a multi-million-pound logistical bomb. A single, delayed ETA triggers a cascade of…