The 85% Exit Time Reduction Nobody Asked For

Everyone wanted their checklist item in our offboarding product. Nobody wanted a faster exit. The second one made us $800K. Here's how we prioritized it.

Here's a product decision that almost didn't happen.

We were building Keka's exit management module. The PRD was 40 pages of checklist items—every stakeholder wanted their task included. IT access. Asset recovery. Full and final settlement. Exit interviews. Knowledge transfer.

Everyone had opinions about what to add. Nobody asked: "What if the exit itself took 85% less time?"

How we found the real problem

I spent two weeks shadowing HR teams during actual employee exits. Not asking what features they wanted—watching what actually happened.

What I saw: The checklist wasn't the problem. The waiting was. An exit that should take 3 days took 11. Not because tasks were hard—because they depended on each other, and no one could see the critical path.

IT couldn't revoke access until finance approved termination. Finance was waiting for the manager's sign-off. The manager was waiting for HR to confirm the last working day. Everyone was waiting for someone else.

The solution nobody asked for

We rebuilt the entire flow around dependencies, not features. Instead of "here's your checklist," we showed "here's what's blocking everything else, fix this first."

We added: - Automatic dependency detection - Real-time bottleneck visualization - SLA tracking by stakeholder (this made some people nervous) - Smart nudges when one person was holding up the chain

The results

  • Average exit time: 11 days → 1.5 days
  • CSAT for HR managers: +80%
  • Module generated $800K in new ARR within 9 months

The lesson

The best features are often the ones nobody asks for—because customers describe problems in terms of what they already understand. They ask for better checklists because that's the mental model. They don't ask for "dependency-aware workflow orchestration" because that's not how they think.

Your job is to hear the surface request and understand the underlying need.