HiPo identification without bias: the 3-signal model

60-Second Summary

High performance ≠ high potential. The first is about today's job; the second is about the job two levels up.
The three signals: learning velocity, ambiguity tolerance, peer-multiplier effect. All three are needed.
Visibility bias is the #1 distortion — the loudest, most CEO-adjacent people get flagged HiPo even when their team is dragging.
Run a bias audit before the HiPo list goes anywhere: gender, tenure, function, manager. Patterns reveal what the process is actually measuring.

Most HiPo lists are popularity contests with a spreadsheet. The people who get flagged are usually loud, present, and managed by someone who's good at advocacy. That's not high potential — that's high visibility. Real HiPo identification asks a harder question: who can do a job that doesn't exist yet, with information they don't have, and pull a team along with them?

Why high performance isn't enough

The expensive mistake

Promote a high-performer with no learning velocity and you don't get a great manager — you get a frustrated former IC who's now blocking three careers. The Peter Principle is just bad HiPo identification with a fancy name.

The three signals

All three must be present

1
Learning velocity
How fast do they get to competent in a new domain? Concrete test: a project they took on this year where they started from zero — how long to first useful output?
2
Ambiguity tolerance
Do they make forward progress when the goal is unclear and the data is incomplete? Concrete test: an open-ended initiative — did they reduce the ambiguity for others, or wait for it to be reduced for them?
3
Peer-multiplier effect
Do the people around them get better? Concrete test: 360 from 3 peers — would they want to work with them again, and did their own output go up when they did?

Any two without the third is interesting but not HiPo. All three is rare — usually 5-10% of a team.

The HiPo identification process

Each manager nominates up to 15% of their reports with written evidence for each of the three signals.
Cross-functional panel (3-5 people) reviews nominations. Anyone the panel hasn't seen work directly gets challenged.
Reject nominations that lean on one signal heavily — those are usually performers, not HiPos.
Final list shouldn't exceed 10% of the org. If it does, you've labeled performance as potential.
Annual review. Anyone on the list 2 years running with no new stretch experiences comes off — it's a development list, not a status badge.

Bias audit checklist

Gender split on the HiPo list vs. the eligible population. >15 pt gap means investigate.
Tenure split — are new joiners under-represented because nobody knows them yet?
Function split — engineering and sales usually over-represented because output is visible.
Manager concentration — is one manager nominating 4 of 6 of their team? Either they have a unicorn team or they're inflating.
Visibility check — names that come up in CEO/exec rooms. Are they on the list because of the work or because of the seat?

The Sunday morning test

If you imagine reading the HiPo list out loud at an all-hands, would the people not on it understand why? If the answer is 'they'd be confused or upset', the list is reflecting visibility, not potential.

How to communicate (and not communicate) the list

Do / Don't

Tell each person on the list, privately, with specifics about why and what's expected.
Give them stretch assignments inside the next 90 days — otherwise the label means nothing.
Tell them: 'this is a 12-month label, not permanent.'

Don't

Publish the list — creates a caste system and torpedoes culture.
Tell their manager and not them — they figure it out from compensation drift, and trust collapses.
Use it as the only input to promotions or comp — it's one signal, not the only one.

Written by Pawan Joshi.Sources cited inline.

First published 16 Jun 2026See site changelog →