When Employees Deliver Faster With AI, Should Management Raise Targets?
Your team finishes in two days what used to take five, thanks to AI. The obvious move is to raise targets. Here is why that can backfire, and what to do instead.
Management should not simply raise targets every time AI makes work faster. Doing so punishes the people who adopted AI first, encourages them to hide gains and often lowers quality. A fairer approach is to measure the real gain over time, share part of it with the team and reinvest saved time in better work, learning or reduced overload.
Picture a support team that used to close 40 tickets a day each. After adopting AI tools, the best performers close 60. Within a month, the new target is 60 for everyone, including those still learning the tools. The fast adopters get no reward, the slower ones are now "underperforming", and next quarter nobody mentions the new shortcut they found. It is a story that repeats in many offices, and it is avoidable.
How much faster does AI actually make people?
It varies a lot. One of the best-known studies, by economists Erik Brynjolfsson, Danielle Li and Lindsey Raymond, looked at more than 5,000 customer support agents using an AI assistant. On average, issues resolved per hour rose by about 14%, with the biggest gains, around 34%, among novice and lower-skilled workers, and little effect on the most experienced (NBER Working Paper 31161).
That finding matters for targets. AI gains are uneven across people and tasks. Setting one new number based on the fastest employee, or the first exciting week, rarely reflects reality.
Why can raising targets backfire?
What happens when targets jump too fast
- It punishes early adopters. If being efficient only brings more work, people stop being visibly efficient.
- It ignores checking time. AI drafts quickly, but verifying its work still takes human time. Cut that time and errors rise.
- It turns a metric into the goal. Goodhart's law, often summarised as "when a measure becomes a target, it ceases to be a good measure", applies strongly here.
- It erodes trust. Teams quickly learn whether productivity gains are shared or simply taken.
Also readBlameless Postmortems: How Good Leaders Turn Mistakes Into Team Learning→What is a fairer way to handle AI productivity gains?
A fair AI-gain policy
One practical split many teams can use is a three-way one: some of the saved time goes to higher output, some to better quality and checking, and some back to people, whether as learning time, fewer late nights or recognition. The exact proportions matter less than the fact that they are agreed openly.
- Set targets from team-wide data over several weeks, not from the best performer
- Include quality and error rates alongside speed
- Reward people who share AI shortcuts with the team
- Protect time for reviewing AI-generated work
- Explain clearly how productivity gains will be used
This connects to a bigger shift in how managers judge work, which we covered in AI-era leadership: evaluating AI-assisted work. It also links to an older H View theme: when every task becomes urgent and every gain is taken, leadership has already lost control.
Her View
Many employees already carry hidden extra load: covering for absent colleagues, handling family responsibilities, staying late quietly. When AI finally gives them a little breathing room, taking all of it away immediately sends a harsh message. Letting people keep some of that time is not softness. It is how you keep your most reliable people.
Also readLeadership Is a Daily Practice: 7 Habits That Build Trust at Work→His Insight
From a business view, the question is not whether to capture AI gains, but how to capture them sustainably. A target that rises 10% with team agreement and stable quality beats one that rises 50% and leads to errors, rework and resignations. Measure the full cost, including mistakes and attrition, before deciding a gain is real.
H View Verdict
When employees deliver faster with AI, raising targets overnight is the easy move and usually the wrong one. Measure gains honestly over time, include quality, share the benefit openly and adjust targets with the team. That way AI becomes a reason to stay, not a reason to burn out.
Frequently Asked Questions
Should companies raise targets because of AI?
Possibly, but gradually and fairly. Targets should be based on measured team-wide gains over time, include quality, and leave part of the gain for learning, checking and wellbeing.
How much does AI improve productivity at work?
It depends on the task and person. A study of over 5,000 support agents found about a 14% average rise in issues resolved per hour, with roughly 34% for less experienced workers.
What is Goodhart's law?
It is the idea that when a measure becomes a target, it stops being a good measure, because people start optimising for the number rather than the real goal.
How should managers share AI productivity gains?
Agree openly how saved time will be used, for example split between higher output, better quality and benefits for employees such as learning time or less overtime.
H View stories follow our editorial policy. Spotted an error? Tell us and we will correct it.
Rate this article
One tap is enough. No comment needed.
Was this article helpful?
Your answer goes straight to our editors.
Comments0
No comments yet. Start the conversation.