In a job search, why fix your walk-away number before the final interview round?
answer
- Two numbers, not one
- Written while you still can
- The floor, not the wish
- Below this you decline and keep looking
basics
~20 sA walk-away number is the lowest package you would genuinely decline below, chosen from benchmark data before any offer exists. Fixing it early keeps a late, pressured decision from being made by adrenaline rather than by evidence.
solid answer
~50 sMarket research exists to produce two numbers on a benchmark sheet: a target and a walk-away. The target is a defensible point inside the pay band for the level, geography and company stage you are interviewing at. The walk-away is the floor below which you would decline and keep searching. Both are written before the loop ends, because a floor chosen after you hear a number is anchored by that number, and because an offer usually arrives with a response window and pressure is a poor time to invent a position. The private form of it is a sentence you say only to yourself: 'My target is set against senior-band data for this market, and I have a walk-away number.' It is working data, not a script, and it is only worth anything if you would actually honour it.
go deeper
Know that market research ends in two written numbers, a target and a walk-away, and that both are set before any offer arrives. Be able to say what each one is for without confusing the floor with the figure you hope for.
Be ready to explain the mechanics: how a percentile position inside a level band becomes each number, why a position travels better than a bare figure, and why a floor derived after hearing an offer is anchored by it.
Show the judgment of holding a pre-written floor when a response window is closing, and of distinguishing new information about the role's level or scope — which legitimately moves both numbers — from pressure, which does not.
Own the harder call of what the floor is even denominated in when packages are structured differently, and accept the cost: a floor set too high wastes processes, one set too low quietly caps every search you run.
## The two numbers a benchmark sheet exists to produce Market research is not an end in itself. It exists to produce a written position: two numbers on your own benchmark sheet, each attached to a level, a geography and a company stage, and both fixed before the final round of a loop. - **The target** is where you would be pleased to land — a defensible point inside the pay band for the level you are being hired at, not the best figure you can imagine. - **The walk-away** is the floor — the point below which you would decline and continue searching. The sheet itself is small: one row per data point, with columns for level, geography, company stage, what the figure includes, and where it came from, plus two lines at the bottom for the target and the walk-away. ## Express both as positions in a band, not as bare figures A bare figure carries no frame. It dates, it does not travel between markets, and it silently smuggles in a level you may not have been offered. A position inside a band survives all three. Suppose you gather entries for a senior-level security engineering role in one large US metro, and after discarding the ones you cannot tag to a level you hold 23 usable entries whose totals spread from roughly the 35th to the 85th percentile of that sample. A target at the 68th percentile is ambitious and still inside your own evidence. A floor at the 41st percentile says you would rather stay in the process elsewhere than accept the lower two-fifths of what that level pays. These are illustrative percentiles for one level in one US market — substitute your own benchmark, and note that bands, and what parts of a package are negotiable at all, vary by employer, by market and by jurisdiction. Eventually you do write an absolute figure, because offers arrive in currency. Write it beside the percentile it came from, so that later you can tell whether it was the number or the level mapping that moved. ## Why the numbers are fixed before the loop ends **Anchoring.** A floor invented after you hear a number is a function of that number. Every study-worthy thing about negotiation preparation reduces to this: the first figure you encounter reshapes what feels reasonable, and the only defence is a position that predates it. **Time pressure.** An offer commonly arrives with a response window. Deciding your floor inside that window means deciding it while tired, flattered and afraid of losing the process. **Honesty with yourself.** Before an offer exists, a floor is cheap to set truthfully. Afterwards it is expensive, because admitting the offer is below your floor now costs you something concrete. ## A floor you would not honour is not a floor The walk-away is not a bluff and it is not a line you deliver. Its whole function is private: it converts a hard decision under pressure into a lookup. If you would in fact accept a package below the number you wrote down, you did not write a floor, you wrote a wish — and you will discover that at the worst possible moment. A useful test before you commit it to the sheet: if this exact package arrived tomorrow, could you say no and go back to the other conversations without regret? If not, the floor is too high; lower it until it is true. ## What legitimately moves the numbers, and what does not New information about the job moves them. If the role turns out to be opened a rung higher than you mapped, the whole band shifts and both numbers move with it. If your level mapping was wrong, redo the mapping and re-derive, rather than nudging the floor. Pressure is not new information. A closing deadline, a warm conversation with a compensation partner, or the sunk cost of five completed interviews change nothing about what the level pays. When you feel one of them tugging on the floor, that is exactly the moment the pre-written number is earning its place. ## Where the floor sits for a first search With no history to lean on, derive the floor the same way: pick the level the role is opened at, pick the geography and the company stage, gather what you can, and set the floor low in that band rather than anchoring it to your living costs. Living costs tell you what you need; they say nothing about what the level pays, and a floor built from them will either be embarrassingly low or unmeetably high by accident.
- How would you set a walk-away number for your first job, with no pay history to anchor to?Derive it the same way anyone does: fix the level the role is opened at, the geography and the company stage, gather what level-tagged entries you can find, and put the floor low inside that band. Do not build it from living costs — those tell you what you need, not what the level pays, and a floor built that way lands high or low by accident.
- What would make you revise a walk-away number you had already written down?New information about the job, not pressure from the process. If the role is actually opened a rung above where I mapped it, the whole band shifts and I re-derive both numbers. A closing deadline, a friendly conversation with a compensation partner, or five interviews already invested change nothing about what the level pays.
- If every band you find sits above the floor you wrote, what does that tell you?Usually that the floor was anchored to something other than the level — past pay, expenses, or a rung below the one you are interviewing at. Re-check the level mapping first, since a floor set against the wrong rung is wrong by a whole band, then decide whether the number is genuinely a floor or just a comfortable one.
A walk-away number is a reserve price at an auction: set before bidding opens, never announced, and worth nothing unless you would really let the item go.
saying these in an interview costs you the question
- Setting a floor only after hearing the first number
- Writing a walk-away you would not actually honour
- Using one number as both target and floor
- Anchoring the floor to current pay rather than the level's band
- Deriving a floor from living costs instead of benchmark data