What Readiness Costs
The stretch assignment, the word nobody defines, and the rule that says not to look.
On this page
Somewhere right now, someone is months into what they were told was a stretch assignment. They're doing the larger job. They're being paid for the one they had, on the understanding that this is how a person shows they're ready. Everyone's heard it: dress for the job you want.
Behind that decision there's usually something written: a scale that runs from one to five, a list of qualities the company says it looks for, a sentence in a handbook about meeting expectations. At a small company, there's often nothing written at all, just an owner who says they know it when they see it.
The question isn't whether the standard is written down. The question is what stands behind it.
Take a rating form. It can look complete: five levels, a description for each, a box for comments. It can't show how the person filling it in got from your work to the review. A form isn't a standard because it looks like one. It's a standard only if something stands behind it.
Where the standard is written down, the words get published. How the person writing your review applies them doesn't. Where nothing is written, there's only that person's idea of excellence. Either way, that person holds the power. They decide what excellence means, measure you against that meaning, and don't have to show it.
It can be made visible, and it doesn't take authority. It takes a request.
The same sentence, three years running
Someone's been told the same thing three years in a row by three different people. The first year it was executive presence. The second, gravitas. The third, how they show up.
Whoever is deciding reads that as confirmation. Three people send one signal.
It's one signal because it's one conversation. The word moved between them on conference calls, at offsite meetings and round tables, and in feedback passed from one manager to the next. None of it was written down, so none of it can be measured, and it still follows the person.
Ask those three to define the term, and you get three different definitions, or an apologetic silence. The agreement isn't evidence about the person. It's evidence that a word traveled and its meaning didn't.
None of the three invented anything. They inherited a term with a meaning already inside it. A definition can be read, and there's still an argument over how it applies to one person in one situation. That part was never written down anywhere the person could find it and work against it.
Even the people writing the reviews rarely call the criteria clear. Of 338 managers and directors surveyed at a Fortune 500 technology company, 24 percent of the men and 16 percent of the women said the criteria for promotion were clear (Correll et al., 2020).
That's the whole technique. Publish the standard. Withhold the interpretation.
One version of this has been measured. A large retail chain rated 29,809 management-track employees on their performance and their "potential," and the ratings were meant to be "a standardized, universal reference for assessing readiness for higher positions." The employees weren't formally told their ratings. Women were rated lower than men on potential despite higher performance ratings, and that difference accounted for about half of the gap in who was promoted. Women who went on to outperform men with the same ratings were rated lower on potential again the next year (Benson et al., 2026).
You can't prepare for a word that means whatever the person writing your review decides it means. You can only resemble it, or fail to.
Resembling it doesn't cost everyone the same. When the person holding the pen uses a word like presence, they picture one way of being in a room: one pace, one amount of eye contact, one length of small talk before the point. A person who is autistic, who has ADHD, who stutters, or who works through an interpreter can do every part of the job and still not match, and the gap gets filed under presence.
Closing that gap is work, and it has a name: the masking tax — the hours spent rehearsing, editing, and auditing every exchange so it lands the way the room expects. It's not paid. It's the same unpaid labor as those months, done every day against a standard nobody has written down.
In one study, people formed less favorable first impressions of autistic people within seconds, and the difference disappeared when they judged only a transcript of what was said (Sasson et al., 2017). In a U.S. national sample of adults in their late twenties and early thirties, people who stutter earned more than $7,000 a year less than people who don't, even after accounting for background (Gerlach et al., 2018).
And it doesn't stay in one decision. The same unexamined word sets the raise. It sets the schedule. It decides which account, which job, which shift. It decides who gets sent to the training and who gets kept when the payroll has to shrink. One word that nobody has defined travels into every one of those decisions. It carries the same interpretation each time, and no one is ever asked to show it.
An interpretation that was never written down can still be asked for.
Two prices for one job
When the role finally posts, the company has months of documented evidence that the internal candidate can do it, and a résumé and three conversations for the external one.
The external candidate is priced at market. The internal candidate is priced off their existing salary, set before any of this began.
The same anchoring shows up in raises. In one experiment, M.B.A. students playing bank managers set raises for four employees. Handed the budget as a percentage, they gave raises that tracked each person's existing salary, so the gap already there carried forward (Gunnell et al., 2026).
One study went through seven years of personnel records at the U.S. investment banking arm of a financial services firm. Outside hires started out paid around 18 percent more than people promoted into similar jobs, and the people promoted got significantly better performance evaluations for their first two years. By the study's estimates, their salaries would take seven years to catch up, and their total pay never would (Bidwell, 2011).
Two methods run at the same time, inside one hiring decision. Neither one is published. There's no line in any handbook that reads, "We price strangers at market and employees at history."
Staying has a price too. At the same retail chain, managers correctly judged women less likely than men to leave. Among employees with the same ratings, the ones judged likely to leave, who were more often men, got higher potential ratings the next year, more promotions, and higher pay. They didn't perform any better. The authors write that firms "essentially reward the threat of exit" (Benson et al., 2026).
More documented evidence produces the lower number. That's not a rounding error, and it's not a rhetorical flourish. Either those months were evidence of capability, in which case they should have moved the price, or they weren't, and nobody should have asked for them. There's no third possibility.
The unpaid months become the record that the work is worth nothing. The person did it for free. Then the offer comes: "We keep promotion increases within a set range of current pay, and this is your number." It arrives as policy, so there's nothing to negotiate. The larger job gets priced as a raise.
The price came from a method. A method can be asked for.
The rule that says do not look
The federal government published a method once, and the method was a rule about when not to check.
The Uniform Guidelines on Employee Selection Procedures took effect in 1978 (U.S. Equal Employment Opportunity Commission, 1979). Paragraph 4(C) is headed "Evaluation of selection rates. The 'bottom line.'" If a hiring or promotion process shows no adverse impact overall, federal agencies usually "will not expect a user to evaluate the individual components for adverse impact." There are two exceptions, and a clause for unusual circumstances.
As of October 2026, that paragraph hasn't been amended since December 31, 1981. The Equal Employment Opportunity Commission is now finalizing a rule to rescind its interpretive portions of the Guidelines. The Department of Labor has already published a final rule removing its own copy from the rules for federal contractors, effective October 26, 2026 (U.S. Department of Labor, 2026).
It doesn't say the components are fair. It says that if the total looks acceptable, nobody is expected to open it.
The components are the steps a person actually goes through. The Guidelines' definition of a selection procedure includes "training programs, or probationary periods." Unpaid months spent proving readiness for a promotion fit that description. So does the rating. If the total looks acceptable, neither one gets opened.
Paragraph 4(A) puts the recordkeeping on the employer. The party whose selection procedures are in question is the party that maintains the data that would disclose their impact. Paragraph 4(D) supplies the only correction. It applies where the employer hasn't kept that data and a group is underrepresented in the job, compared with the labor market or, for jobs filled from within, with the workforce. In that case, the agencies may draw an inference of adverse impact from the failure to keep it. The word is may, not must.
An aggregate is published. The components stay closed. The employer holds the ledger.
Forty-eight years later, the ink hasn't moved.
No agency is going to open the components. The person inside the process can ask to see them.
The method that was published and filed as vocabulary
Rating people against named qualities was a remedy. Somebody identified a real failure: raises and promotions ran on charisma and familiarity. Then they built a fix: name what the job actually requires, write it down, and decide on that, not on who reminded the person deciding of themselves.
It hasn't gone away. In a study of hiring at elite professional service firms, built on 120 interviews with the people doing the hiring, evaluators sought candidates who were "not only competent but also culturally similar to themselves" (Rivera, 2012).
The fix came with a second half: definitions written before anyone is rated against them, and evidence that what you're measuring is what you meant to measure.
The second half has evidence behind it. A 2026 review of 970 published articles on the causes of the gender pay gap reports that policies encouraging transparency, formalization, and accountability in pay decisions have been shown to reduce the gap by reducing discrimination (Livingston et al., 2026).
Even the length of the scale changes how people rate. When a university professional school cut its students' rating scale for instructors from 10 points to 6, the gap between men's and women's ratings in the most male-dominated fields largely disappeared (Rivera & Tilcsik, 2019).
What traveled from company to company was the first half: "executive presence," "strategic thinking," "team player," "ready for the next level." What stayed behind was the checking.
You can copy a phrase into a handbook in an afternoon. You can't copy the evidence behind it, and under the bottom-line rule, no federal agency is going to ask you for it.
So the remedy arrived everywhere, and the checking arrived nowhere. A set of words built to make judgment more accurate became the most reliable way on earth to write down a preference and have it read as a finding.
The method was published. It was filed under vocabulary.
The second half is still there to ask for: the definition, and the evidence.
What these have in common is not the subject
The common thread isn't promotion, or pay, or one person being talked about in a room.
Each one invites the affected person to look at an output and withholds the method that produced it: the rating without the reasoning, the offer without the pricing rule, the stretch assignment without the finish line, the word without the definition. Where the definition was published, everyone agreed to treat it as paperwork.
Every one of those is a form with nothing behind it. It's not exactly a lie. It only has to look complete to the person it's about.
That person is the one being measured. From where they sit, it looks like a process.
Next: Define the Goalpost is the other half: how to ask, for yourself and for someone else.
If you recognized yourself in this, or someone on your team, The Naming Work is where to start.
You already know what they call it when you do it. Nobody names the version being done to you.
Your work, your effort, and your discomfort get renamed one at a time, until the record reads like consent.
The Naming Work isn't a read. It's a scored diagnostic: fifteen questions, measured against a rights floor that you didn't set and that they can't move, not against how reasonable you're being.
You get a verdict. What you do with it is the actual work, through coaching or the app.
Name it before you change it.
Start at https://app.kovachconsultants.com/the-naming-work
References
Benson, A., Li, D., & Shue, K. (2026). "Potential" and the gender promotion gap. American Economic Review, 116(2), 375–417. doi.org/10.1257/aer.20220831
Bidwell, M. (2011). Paying more to get less: The effects of external hiring versus internal mobility. Administrative Science Quarterly, 56(3), 369–407. doi.org/10.1177/0001839211433562
Correll, S. J., Weisshaar, K. R., Wynn, A. T., & Wehner, J. D. (2020). Inside the black box of organizational life: The gendered language of performance assessment. American Sociological Review, 85(6), 1022–1050. doi.org/10.1177/0003122420962080
Gerlach, H., Totty, E., Subramanian, A., & Zebrowski, P. (2018). Stuttering and labor market outcomes in the United States. Journal of Speech, Language, and Hearing Research, 61(7), 1649–1663. doi.org/10.1044/2018_JSLHR-S-17-0353
Gunnell, H. T., Schuhmacher, K., & Towry, K. L. (2026). Un-nudging pay gaps: The role of pay raise budget framing. The Accounting Review, 101(2), 281–311. doi.org/10.2308/TAR-2024-0105
Livingston, B. A., Jeong, S., Klein, F. B., & Chakroborty, A. (2026). Mind the gap: An integrative review of the causes of the gender pay gap. Journal of Organizational Behavior, 47(2), 261–287. doi.org/10.1002/job.70034
Rivera, L. A. (2012). Hiring as cultural matching: The case of elite professional service firms. American Sociological Review, 77(6), 999–1022. doi.org/10.1177/0003122412463213
Rivera, L. A., & Tilcsik, A. (2019). Scaling down inequality: Rating scales, gender bias, and the architecture of evaluation. American Sociological Review, 84(2), 248–274. doi.org/10.1177/0003122419833601
Sasson, N. J., Faso, D. J., Nugent, J., Lovell, S., Kennedy, D. P., & Grossman, R. B. (2017). Neurotypical peers are less willing to interact with those with autism based on thin slice judgments. Scientific Reports, 7, Article 40700. doi.org/10.1038/srep40700
Uniform Guidelines on Employee Selection Procedures, 29 C.F.R. §§ 1607.4, 1607.16 (1978, as amended 1981). ecfr.gov
U.S. Department of Labor. (2026, August 21). Rescission of Executive Order 11246 implementing regulations [Final rule]. Federal Register, 91, 54444. federalregister.gov
U.S. Equal Employment Opportunity Commission. (1979). Questions and answers to clarify and provide a common interpretation of the Uniform Guidelines on Employee Selection Procedures. eeoc.gov
U.S. Equal Employment Opportunity Commission. (2026). Rescission of Uniform Guidelines on Employee Selection Procedures (RIN 3046-AB43) [Unified Agenda entry]. Office of Information and Regulatory Affairs. reginfo.gov
This post is general information, not legal advice.
