When Visibility Stopped Working as a Measure of Performance

A photograph went viral in Bengaluru in 2023. A woman on the pillion of a scooter, laptop open, working in traffic.

The reactions were predictable. Some were horrified. Many were admiring. A surprising number were simply recognising.

That recognition is the point worth examining. Not the image itself, but what it reveals about what workplaces have learned to reward.

The Proxy That Once Made Sense

Visibility as a performance measure was not irrational. In factories, offices, and trading floors, presence was production. If you were there, you were working. If you left early, you were not. The logic was direct and, for a long time, defensible.

Knowledge work broke that logic. Slowly enough that many leaders did not notice.

What changed was not the hours, but the relationship between hours and output. A consultant thinking through a problem on a walk is working. A manager attending a meeting out of obligation is, in a meaningful sense, not. Presence stopped predicting performance. The measurement did not change with it.

The proxy outlived the conditions that made it useful.

What the Measurement System Cannot See

Measurement systems have a familiar failure mode. They see what they are designed to see. Everything outside that frame either disappears or gets distorted.

India’s MoSPI Time Use Survey 2024, designed to capture household labour, found that women performed an average of 305 minutes of unpaid domestic and care work daily. Men, 97 minutes. This labour is real, skilled, and economically consequential. It also appears nowhere in most productivity statistics or performance records. As far as the system is concerned, it did not happen.

The survey is about households. The dynamic is not.

When a system cannot see a category of work, people who perform that work face a choice. Continue, and accept that it will go unrecognised. Or redirect effort toward what the system can see, whether or not it is more valuable.

Offices operate the same way. When leaders measure presence, people optimise for presence. When leaders measure responsiveness, people optimise for responsiveness. The midnight message, the Sunday email, the laptop open on a moving scooter—these are not signs of exceptional dedication. They are rational responses to what the system rewards.

Workers are not gaming the system. They are reading it correctly.

Workers are not gaming the system. They are reading it correctly.

The question is what the system is asking them to become.

The Scale of What Is Being Counted Wrong

A Slack–Qualtrics survey of 18,000 desk workers across nine countries found that productivity is still often measured through inputs—time spent, messages sent, status lights kept green—rather than outcomes such as goals met or value created.

Twenty-seven percent of executives reported relying on visibility and activity metrics. Sixty-three percent of workers said they make an effort to appear active online even when they are not working.

Those numbers belong together. A majority of workers are performing for a system that a significant minority of leaders openly admit they still use. Measurement and behaviour have aligned.

The system is stable. It is also producing the wrong thing—with increasing efficiency.

What AI Will Force Into the Open

Generative AI compresses execution. Tasks that took a day can take an hour. Drafting, analysis, structuring—faster, and getting faster still.

This creates a pressure that visibility-based systems cannot absorb. If someone completes in two hours what used to take two days, the system has no way to register that honestly. The rational response is to stretch the timeline. Submit on day two. The system records productivity. Nothing is learned.

AI does not create this problem. It exposes it—and removes the ambiguity that allowed it to persist.

AI does not create this problem. It exposes it.

More precisely, it removes the ambiguity that allowed it to persist. What was once inefficiency begins to look like performance theatre.

Leaders who continue to measure visibility will not just misread performance. They will increasingly institutionalise its simulation.

What Evolving the Measure Actually Requires

This is not an argument for output-only management. That has its own blind spots. Measuring only deliverables misses collaboration, mentorship, and the slower work of building institutional knowledge.

The question is not presence versus output. It is whether leaders have developed a measurement vocabulary that can see both.

That vocabulary begins with simple but uncomfortable questions. Who is in this meeting, and what would be lost if they were not? When someone finishes early, is the instinct to reward them or to refill their time? Is the team being evaluated on what it produces, or on how much it is seen to be trying?

These are not performance management questions. They are diagnostic ones. They reveal, more clearly than any values statement, what a leader actually believes about how work happens.

The photograph sparked a debate about corporate pressure and work–life balance. That debate will continue.

The more consequential question is simpler, and harder to avoid:

If visibility no longer signals value, what exactly are we measuring—and what, in the process, are we training people to perform?