The Pay-Talk Readiness Standard, Graded Before the Question Lands

Before a recruiter or interviewer asks about money, your compensation prep should already be done, consistent, and checkable. This guide is for candidates heading into any interview stage who want to grade their own pay-talk preparation pass or fail against explicit criteria, so they never improvise a number live. It gives you the pre-flight standard: the researched range, the target placement, the private floor, the base-versus-total split, the no-disclose boundaries, and the checklist that grades the prep itself before the question lands.

Other guides in this library script the answer, grade a single transition statement, or decode an offer already in hand. This one does something earlier and narrower: it defines when your preparation is actually finished, so two people assessing the same prep reach the same verdict. If you cannot pass this standard quietly at your desk, you are not ready to pass it live on a call.

What "pay-talk ready" means, stated as five separate items

Pay-talk ready means five distinct pieces of preparation exist, are internally consistent, and each has a source behind it. Published guidance converges on preparing a researched market range, a stated target inside it, a private walk-away floor, a base-versus-total-comp split, and a scripted set of no-disclose and redirect lines. Miss any one and the prep is incomplete, because the gap surfaces at exactly the moment you cannot afford it.

These are not one blurry number. Coursera is explicit that you need two self-defined figures: "when conducting research, it's helpful to determine two key pieces of information for yourself: a specific pay rate and a corresponding pay range." Robert Half frames the range as one "you've researched that you would accept." The standard below treats each item as a pass/fail line, not a topic you have "thought about."

The reason to separate them is that they fail independently. A candidate can have a beautifully researched range and still leak their current salary. Another can have a disciplined no-disclose line and a floor they would actually decline. The checklist at the end grades all five, and any one miss is a fail.

What to research before the pay conversation, and what counts as backing

Acceptable backing is a federal employer survey, posted-range data where the law mandates it, and named aggregators, triangulated so no single source carries the number alone. The BLS Occupational Employment and Wage Statistics program is the authoritative floor: it "produces employment and wage estimates for approximately 830 occupations based on a survey of business establishments." On top of that, cross-check against posted ranges and against aggregators like Levels.fyi, Glassdoor, and Payscale.

BLS OEWS is broad and credible. It covers roughly 830 occupations across about 530 metro and nonmetro areas, drawn from a six-panel sample of roughly 1.1 million establishments. Use it for the median and the spread, at both national and metro level, because a national figure alone misprices you in an expensive or a cheap market.

The weakness of BLS is time. OEWS data "are published annually with a May reference date," and the May 2025 estimates were released May 15, 2026. That lag is why the baseline is necessary but not sufficient. Your prep passes this section only when every number you quote carries both a source and a date, and when the BLS figure has been cross-checked against several current postings at your level.

How wide the range should be, and where to place your target

Keep the spread tight, roughly 10 to 20 percent, because the spread is what gets graded, not the midpoint. A tight band reads as researched; a wide one reads as guessing and gets discounted. Robert Half states it plainly: "a spread of $85,000 to $95,000 tells the employer you know the market; $70,000 to $110,000 tells them you're guessing. Keep the gap tight enough to look deliberate." CNBC adds the mechanism: "the wider the range, the softer it is - it almost seems like you're not saying much."

Published examples cluster in a narrow band, which gives you a reference set to calibrate against.

The pattern is clear: everything a guide presents as credible sits at or below about 21 percent, and the single "guessing" example sits at 57 percent. Stay under 20 percent and you look deliberate.

Target placement is where honest sources disagree, so the standard cannot mandate one rule. CNBC advises the floor and the target both sit low: "Lares says to be clear about your absolute minimum and put that number at the bottom of your range. Your target number should be on the lower end too." The opposing view is that hiring managers seat hires mid-band on purpose, because "a manager doesn't want their staff hitting the ceiling in year one." A third heuristic, when a posted range exists, targets "around 75 to 90 percent of the range." Your prep passes when the placement is documented and defensible, not when it matches any one school.

The spread, not the midpoint, is what the employer is grading. Keep it under twenty percent and it reads as deliberate.

The disclosures that fail the standard, and the legal boundary behind them

Naming your current or past salary fails the standard, and in many US jurisdictions the employer asking is also illegal. But the two facts are not the same, and confusing them is dangerous. Salary-history bans "do not forbid workers from volunteering their salaries," which means the law protects you only if you keep your mouth shut. The load-bearing control is your own no-disclose discipline, not the statute.

The legal landscape is real but uneven, and the counts conflict by date and definition. A recent tracker narrows the statewide picture: "as of August 12, 2026, eighteen states plus Washington, D.C. have a statewide law that restricts employers from asking job applicants about their pay history." An older Wikipedia count records "twenty-two American states and twenty-three American municipalities" with some form of ban as of May 2024, with the first passed in Massachusetts in August 2016. The exact number is not settled across sources, so treat it as a range and confirm your own jurisdiction locally rather than relying on any single count.

Separately, posted-range laws are spreading, which changes what you can research before you ever talk to anyone.

Roughly 14 states plus D.C. now require pay ranges in postings, and per Payscale data about one in four US workers was projected to be covered by a pay-transparency law. That does not mean the posted range is trustworthy. Without a width cap, employers comply with bands like the reported Netflix software-engineer listing of "$90,000 to $900,000, a difference of $810,000." The posting verifies almost nothing at that width; the BLS and aggregator triangulation still carries the weight.

The redirect when they press first, written to copy

When a recruiter will not share their range and presses you to name one, offer to talk pay, ask for their budgeted range first, and only if they refuse give a researched range on the higher end while stressing flexibility. The sequence matters because the first number spoken anchors the whole process, and practitioners report recruiters keep detailed notes. That is precisely why this standard grades prep before the call rather than hoping you improvise well during it.

Robert Half scripts the turn-around directly. U.S. News adds the fallback for when they keep pressing: "give a range based on your market research that's on the higher end, but emphasize you're flexible, otherwise the interviewer may screen you out as too expensive." Note that sources split on ordering. Robert Half runs research, then range, then turn-around; Coursera and Indeed advise reversing the question before ever stating a number. Both are defensible. What is not defensible is naming a figure before you have tried the redirect at least once.

Redirect: "I'm happy to talk about pay. Could you share the range you've budgeted for this role? That way we can make sure we're in the right ballpark before we go further."

If they refuse and press: "Based on my research for this role and market, I'm targeting the [X to Y] range, and I'm flexible depending on the full picture of the role and comp."

No-disclose (current salary): "I'd rather base this on the market and the scope of this role than on what I've made before, so I'm comfortable sharing my target range instead."

The single-number question comes later. Robert Half: "later in the process, you may be asked to provide a single salary number, not a range. By a second or third interview, you should understand the job expectations and how success will be judged." Where that number lands is a fit call: a specialized skill the employer wants supports a figure toward the top of your range, while limited experience makes a point closer to or below the midpoint more realistic. Prepare that single number as a separate output, not as an afterthought you compute on the spot.

When the hard part is building a band-consistent set of numbers and materials quickly across many applications, Refolk tailors your resume and positioning to each posting from your own history, which keeps the figures you present consistent with the band you are targeting.

The procedure: from market baseline to a graded pass

Run these seven steps in order, treating each "done" condition as a gate. The output is a documented prep packet and a pass/fail verdict two reviewers would agree on.

Use a real search to see how peers in your exact situation are priced and placed before you lock your numbers.

How this goes wrong: failure modes and false positives

Most pay-talk prep fails in one of eight predictable ways, and each has a specific check that catches it. Treat this section as the part of the standard that does the real work, because a prep that looks complete can still be quietly broken.

Two of these deserve extra weight. The floor that is not a real floor is the quiet killer: Robert Half warns the low figure must still be a yes, not a lowball set to look modest. If you would decline the number at the bottom of your range, your range is a lie you told yourself, and you will be held to it. The fix is a single blunt question: would I sign at this exact number?

The second is the screen lowball. Blind posters describe naming a figure 10 to 15 percent under their real floor to stay in the running, then being held to it because recruiters keep detailed notes. The number you say first anchors the entire process. This is the structural reason the standard grades prep before the call. You do not want to be discovering your floor while someone is waiting for an answer.

A note on the Software Engineer market, because it illustrates why triangulation is non-negotiable. In Refolk's index, the US and UK pools are very different sizes, and a thinner market offers fewer public comps to validate a stale federal median against.

The US-to-UK ratio is about 8.1x. A seniority split was attempted but did not return reliably in the index, so I will not claim a senior-versus-entry breakdown; that comparison is not established. The point stands without it: where there are fewer profiles and fewer live postings, lagging data does more damage, and you should lean harder on the postings you can find.

The checklist to grade a pass before the question lands

Run this before any interview stage where money might come up. Any unchecked item is a fail, and two reviewers running the same prep should reach the same verdict.

How to keep this current when you reuse it

Re-verify the prep every time the market or the law could have moved, because both the data and the statutes shift under you. The two things that go stale fastest are your market baseline and your legal boundary, and each has a clear re-check mechanism.

For the data, remember OEWS is annual: the May 2025 estimates landed May 15, 2026, so the published median is up to a year behind the market. Re-pull the BLS figure when a new annual release drops, and re-scrape three to five live postings every time you reactivate a search, since the live band is the real-time check and the only one that reflects right now.

For the law, treat the counts as moving. Posting-law thresholds and salary-history bans are being added state by state, and the published totals already disagree by date. Before relying on a ban, confirm your specific state or metro against a current legal tracker rather than any number quoted here. When you move markets or levels, rebuild the whole packet rather than editing the old one, because a thin market like the UK gives you fewer comps and a stale number costs more there. The standard does not expire, but the figures inside it do, so date everything and grade again before each stage.