Managing up against an underrating boss means doing three things consistently: making the work visible in the boss's own success metrics, asking directly what would change the boss's assessment, and building a second audience for the record. The problem is common and it is gendered. Women are less likely than men to say a manager has advocated for them, and women leaders are more likely to have their judgment questioned, according to the McKinsey and LeanIn.Org Women in the Workplace 2024 report. For every 100 men promoted to manager, 81 women were promoted, per the same report, and the difference is often decided in one person's head.
This guide covers the practical side of correcting an underrating relationship. It is general career information, not a promise of promotion and not legal advice.
Why do bosses underrate people?
Mostly information, sometimes bias, usually both. A manager sees a fraction of the work, whatever surfaces in meetings and escalations, and fills in the rest from proxies: visibility, similarity to people the manager has promoted before, and confidence of presentation. The research on performance evaluation is blunt about the similarity part: the same behaviors are read differently depending on who performs them, a pattern documented in Catalyst's work on gender bias in assessment and in academic performance-review studies that found personality language applied to women where achievement language was applied to men.
The actionable conclusion is that an underrating is a hypothesis the boss holds on incomplete data. Hypotheses respond to data. What they do not respond to is resentment, which is invisible, or overwork, which the boss already counts as baseline.
What is the first move?
A single direct conversation, scheduled and prepared: what would you need to see over the next six months to rate my work as excellent and support me for the next level? The question forces the boss to state criteria out loud, which does three things at once. It surfaces any standard the employee has been guessing at, it creates a written target if the answer is captured in a follow-up email, and it distinguishes the fixable cases from the unfixable ones. A boss who answers with concrete criteria has, willingly or not, signed up to be measured. A boss who answers with vagueness twice has supplied information of a different kind.
The follow-up email matters more than the meeting. A short note restating the agreed criteria, sent the same day, converts a conversation into a record, and records are what get compared to outcomes.
How does someone make work visible without bragging?
By reporting in the boss's metric system, on a schedule, in writing. A monthly five-line note, what was delivered, what it moved against the agreed criteria, what is next, does the work of self-advocacy without a single adjective. The tone is bookkeeping, not broadcasting.
Two habits sharpen it. Numbers before narrative: shipped four days early, error rate down, revenue retained, then one sentence of context. And credit distributed accurately, naming the team's part, which reads as authority rather than inflation and, per the research on gender and self-promotion, lowers the social penalty women pay for stating results plainly.
Related stories: Operations roles are an underrated path to the top for women · How women managers run hybrid teams without burning out.
When should the record go around the boss?
Not around, alongside. The legitimate second audiences are skip-level meetings where the format exists, cross-functional projects where the work touches other leaders, and written artifacts, postmortems, planning documents, dashboards, that travel without their author. Visibility built this way is earned rather than political, and it cannot be accused of going around anyone because it is the work itself moving.
The caution is real, though. A direct campaign to higher-ups against the boss reads as disloyalty and usually costs more than it wins. The sequencing that survives office politics: fix the data problem first, get criteria in writing second, build the parallel audience third, and treat the promotion conversation as the fourth step, not the first.
How does someone know when to leave instead?
When the evidence is complete. The boss has stated no criteria, or has stated them and then ignored them at review time. The visibility work has run for two or more cycles with no movement in assessments, assignments, or pay. And the pattern does not exist elsewhere: peers and other leaders rate the work highly, which rules out the comforting alternative explanation that the work is the problem.
Pay data sharpens the decision. Women working full-time were paid 83.6 percent of what men were paid in median weekly earnings, per the Bureau of Labor Statistics for 2023, and under-rating is one of the mechanisms that produces that number quietly, one review at a time. Staying in a role where the record is capped has a measurable cost; the job market pays for the documented record, not the invisible one.
| Situation | Best next step |
|---|---|
| No stated criteria | Criteria conversation plus follow-up email |
| Criteria stated, then ignored | One documented recalibration meeting; then broaden audience |
| Low visibility of the work | Monthly five-line metric note |
| Pattern persists two cycles | Internal transfer or external search |
What is the mindset that holds up?
Managing up is not flattery and it is not combat; it is the maintenance of an accurate record against an organization that loses information constantly. The boss who underrates an employee is frequently not malicious but busy and biased toward whatever data arrives loudest. The employee who supplies cleaner data, on a schedule, in writing, with criteria in the file, either corrects the assessment or documents its refusal. Both outcomes are usable. That is what managing up is for.