The Pay Conversation Across the Loop, Run to a Number-Intact Offer

You are in an active interview loop, and pay keeps coming up: at the recruiter screen, from the hiring manager, after the panel, and again on the pre-offer call. This guide is for a candidate who wants to run that whole conversation as one sequenced process rather than surviving each question in isolation. It gives you what to say and what to gather at every stage so you reach the offer with your target range uncommitted, their band surfaced, and a documented value case ready to use.

Most pay-during-interview advice treats the question as a single moment to survive. A real loop is not one moment. It is four or five, each with a different asker who wants a different thing, and the number you give early follows you to the end. The method below is the playbook you keep open through the live process.

Why the question is different at every stage

Each stage asks about pay for its own reason, and the right move changes with the asker. The recruiter wants to fill a field and avoid wasting rounds. The hiring manager wants expectations aligned against a band they cannot exceed. The pre-offer caller wants to converge. Reading the asker correctly is what lets you defer early and anchor late.

A recruiter phone screen is a filter, not a deep interview, and salary is the question that sinks most screens. The recruiter is often just trying to fill a field in their tracking system, so the fix is to decide your range before the phone rings. The hiring manager asks to ensure your expectations align with what can actually be delivered, and that manager is restricted to a set range with no strings to pull higher. By the pre-offer call, the people on the line are trying to close.

The timing principle that ties this together is simple: speed early, deliberation late. The ideal time to commit to a number is after both sides have expressed strong interest, typically toward the final stages, when you negotiate from strength. Everything before that is deferral and preparation.

Who you actually face

The people raising pay are a known population, and knowing their counts helps you calibrate who has what authority. In Refolk's index of professional profiles, recruiters outnumber the compensation specialists who approve bands by a wide margin, which matches the field reality that the person asking you early is rarely the person who sets the ceiling.

In Refolk's index, US recruiters outnumber comp approvers by roughly 1.67 to 1. The recruiter asking for your number on the screen is usually relaying it to someone else, which is exactly why the number you give becomes a note in a file rather than an offer. Treat it accordingly.

The three numbers you set before the phone rings

Before any screen, build your range from three numbers: floor, target, anchor. The floor is the figure below which you walk. The target is what you actually want. The anchor is the top of the band you quote. Quote target-to-anchor, with your target sitting at or near the bottom so the whole band sounds reasonable.

Keep the band tight, roughly $10,000 to $20,000 wide, and often under $10,000. If you are aiming for $100,000, quote $95,000 to $115,000 when you are still exploring, or tighten toward $95,000 to $102,000 once you want the job. Published guidance converges on this shape.

Notice the pattern across every row: the target sits at or near the bottom of the band. There are two reasons to express the number precisely rather than in round figures. First, a wide, round range gets read as its bottom, so flexibility you intended reads as a discount. Second, precise offers reduce counteroffers compared to round offers, so a precise figure holds its position better.

To set these three numbers defensibly, you need market data for the specific role, level, and location, not a national average. Public compensation data by level and company gives you the band shape. Refolk builds a baseline from your own history and the public market so your floor, target, and anchor are grounded before the first call rather than guessed on the spot.

The deflect, range, disclose rhythm

The documented rhythm is: deflect once to surface their band, then give a researched range, and never disclose current pay. That is the whole decision tree. You deflect at most once before you answer, and you answer with a range, not a single figure.

The deflection is a flip that turns their question into intelligence. When they ask your expectations, try: "I would love to hear what you have budgeted for this role so I can tell you if we are in the same range." If they give you their band, you have learned something and spent nothing. If they decline or need your number to advance the round, hand over the researched range without drama rather than trading silence.

The current-pay question is a separate case, and the answer is always to redirect to the role: "I'd rather focus on the value of this role than what I made somewhere else. For this position I'm targeting [range]." You name your target range; you never name your current number.

Where the current-pay question is barred

In a growing set of jurisdictions you can decline the current-pay question by citing the law and pivoting to your range. As of August 12, 2026, eighteen states plus Washington, D.C. have a statewide law restricting employers from asking applicants about pay history, relying on it to set an offer, or both.

The detail varies by state. Delaware allows confirming salary only after an offer is extended. Maine bars asking until after a job offer is negotiated. Oregon cannot ask until after an offer is extended. Vermont can confirm volunteered history only after an offer. In these places, at the screen stage, you can simply note the question is not permitted and pivot to your researched range.

Run the loop stage by stage

Here is the ordered procedure. Each stage has an actor, a rough duration, one move, and a definition of done. Run it in order and you arrive at the offer with your range intact.

At the hiring-manager screen, the pivot has two halves. First, deflect the number toward scope: "my focus is on finding the right organizational fit and understanding the full scope of responsibility." Then turn it into a question that gets you their band: "what is the approved target range you have budgeted for this role?" You leave pay open and leave with the information you need.

Sources disagree on one detail of the recruiter-screen step. Some say give a range immediately to confirm fit; others say deflect first and only range if pushed. Both agree you deflect at most once before answering, so the safe default is to try the single flip and answer without hesitation if it does not land.

Build leverage while you interview

Leverage is not a competing offer. It is an articulated, documented case for why you are uniquely valuable, assembled from the interviews themselves. Your job in the negotiation is to articulate and amplify what makes you uniquely valuable such that the company decided to offer you the job. Interviews are where you gather that material, which is why the middle of the loop is leverage-building, not just gatekeeping.

What you actually need by the pre-offer call is three things: market data showing what similar roles pay, quantified value demonstrating what you deliver, and alternatives you are considering. Build a value inventory as you go. By the second or third interview you should understand the job's expectations well enough to know which of your in-demand skills the role needs, and each match supports a figure toward the top of your range.

The mechanism is performance re-rating, and it is observable. In one documented case, a candidate deflected by saying they were still doing market research; after the technical interview the company increased the band based on performance, and the offer came in above the higher end of the band with room still to negotiate. The interviews moved the number. That only works if your range was still uncommitted.

Interviews are not just gatekeeping. Run well, they re-rate you, and the band moves to meet your performance.

The reason running this conversation to the offer intact is such an edge is that almost no one does it. Just 30% of Pew respondents recalled asking for higher pay the last time they were hired, yet among those who did ask, two in three improved on the offer.

Read the table as a combined improvement rate of roughly 66% for people who asked. For context, 78% of new hires who negotiated starting pay say they received a better offer, and people who negotiate get an average of 18.83% more than those who accept the first offer. The gap exists because over 58% of workers never negotiate at all.

If you want to understand who inside a company holds pay authority, or to study how recruiters at your target companies handle bands, you can search for them directly.

How this goes wrong

The failure modes below are the most valuable part of this guide, because each one feels right in the moment and costs you the offer or the money. Every one has a false positive: a reason it feels like good play when it is not.

• Over-deflecting past the recruiter. It feels principled and protected. In reality, getting screened out over a salary standoff before anyone has met you is the worst possible trade. Check: did you deflect more than once? Stop at one.
• Trusting a posted band as the real budget. A wide range on a posting looks generous. But a posted range can be soft, tiered, or wider than the real budget, so it is a guide, not a promise. Check: read a wide band as ambiguity and aim at the midpoint.
• Anchoring too high and killing likability. The research seems to say go high. But a $125k ask against a $100k band makes a later $100k offer feel inadequate to both sides, so they simply will not make it. Check: is your anchor still inside their plausible band?
• Volunteering current pay in a ban state. You assume the law protects you automatically. The ban stops them from asking, not you from telling. Check: never offer the number, even when it is legal to withhold.
• Assuming recruiters forget your screen number. "I'll just revise later." Recruiters keep very detailed notes and know exactly what you shared. Check: treat the first number you give as the floor of everything downstream.
• Leaning only on a competing offer. An offer in hand feels like total leverage. But in one survey, 78% of CFOs said they would not counter another offer to keep someone. Check: is your value case documented independently of any offer?
• Giving a round, soft-sounding wide range. "Flexibility is good." But a wide, round range just gets read as its bottom. Check: keep the band within about $10k to $15k and express it precisely.

The anchoring failure deserves one more note, because the research is more subtle than the headline. A $100,000 anchor in the Idaho study yielded an average offer of $35,385 against $32,463 for the control, so anchoring works. But across other experiments, the anchoring effect diminishes and higher demands reduce likability and hiring probability. The win is anchoring at the top of the plausible band, not above it.

Reach the offer with the run complete

The run succeeds when you reach the offer with your range uncommitted, their band surfaced, and a documented value case ready, before you make any counter. The counter itself is a separate job and out of scope here. Your stop point is the offer landing with headroom intact.

Before you treat the run as done, verify the state you are in. The checklist below is what a complete run looks like at the moment the offer arrives.

Keep the run current as the process moves. Two things change underneath you. First, the jurisdiction rules shift: Washington, D.C.'s ban expands to all employers June 30, 2026, and Virginia's statewide ban takes effect July 1, 2026, so re-check your state's rule when a process starts rather than relying on last year's answer. Second, the market rate for your exact role and level drifts, so refresh your three numbers from public compensation data before a new loop, not during one.

Based on everything I've learned about the scope, the team, and the impact expected, I'd be looking at a base in the $X to $Y range. I'm also very interested in understanding the equity and bonus structure before we talk about anything firm.

One last framing to hold through the whole loop. Employers price in a counter you may never make: since 52% of them intentionally offer below what they will pay, arriving with your range uncommitted preserves the exact headroom they already budgeted. Running the conversation stage by stage is not aggression. It is keeping the money that was always meant for the role from quietly being left behind, and the compounding is real: in one study, graduates who negotiated earned about $6,450 more up front, nearly $800,000 more by retirement.