A search that fails in sourcing is a visible failure. Everyone sees the empty pipeline and the missed deadline. A search that fails at the offer looks like bad luck — a finalist who "went another direction," a counteroffer nobody saw coming. It is almost never bad luck. It is a number that was set months earlier by a process that never consulted the market.
Here is the sequence that plays out in most searches that die late. Finance sets a band from internal equity — what the last person in the seat made, plus a merit percentage, benchmarked against people who already work there. Recruiting runs the search against that band. Three finalists emerge. The offer goes out. The finalist declines, or accepts and is gone in a year.
The postmortem usually blames the finalist. The real failure happened before sourcing started: the band was priced internally and the candidate was priced externally. Those two numbers are set by different systems and they do not automatically agree.
What the market is actually paying in 2026
The national median base for VP-and-above roles now sits at $234,000, with VP-level bands running roughly $210,000 to $260,000 depending on function and scale, according to ExecSignals' 2026 compensation data. That is base alone.
In go-to-market roles the spread widens sharply. Pavilion's GTM Compensation Benchmarks, drawn from roughly 1,200 go-to-market leaders, put VP of Sales at growth-stage companies between $350,000 and $450,000 OTE, and CROs between $600,000 and $800,000 OTE before equity.
Finance leadership prices differently — it tracks company revenue more tightly than title. Robert Half's 2026 guide puts median CFO compensation at roughly $195,000 at companies doing $10–29M, $227,500 at $37–99M, and $241,500 at $80–149M. The same title, three very different numbers, entirely determined by the size of the business the person is joining.
of hiring managers say they will offer higher salaries for candidates with in-demand skills. The band you set last year is competing against employers who have already decided to break theirs.
Robert Half · 2026 Salary Guide
Why the gap is structural, not a negotiation failure
Internal equity is a real constraint and a legitimate one. Paying a new VP forty percent more than the two VPs who already report to the same executive creates problems that outlast the hire. Nobody should pretend that isn't true.
But internal equity is a backward-looking instrument. It prices the seat against decisions your company made in prior years, under prior conditions, for people hired from a prior market. The candidate is priced by what other employers are willing to pay them this quarter. When those two numbers diverge, the divergence is not a negotiating position. It is information.
A budget set from internal equity tells you what the seat used to be worth. It tells you nothing about what it costs to fill.
This is why "we'll stretch if we find the right person" so rarely works. By the time you have found the right person, you have spent the leverage. The candidate now knows they are the finalist. The stretch reads as a concession extracted under pressure rather than a considered market position — and a candidate who negotiated hard to get in tends to keep negotiating once they are there.
What the gap actually costs
The instinct is to treat a blown offer as a reset — you go back to the pipeline and try the next candidate. The arithmetic says otherwise.
- The seat stays empty longer. An open leadership role carries roughly $98 per day in measurable productivity loss, and senior searches routinely run past 90 days. A failed offer adds weeks before sourcing even restarts.
- The second-choice candidate is a different bet. You are no longer hiring the person you ran the search to find. You are hiring the person still available after the person you wanted said no.
- The stretch you refused is smaller than the failure you bought. Leadership IQ's study of more than 20,000 hires found 46% of new hires fail within 18 months, and 89% of those failures are about fit rather than skill. Replacement cost at senior levels reaches 213% of salary. A $25,000 gap on a $250,000 role is 10% of base — against a downside measured in multiples of it.
Put plainly: the money saved by holding the band is almost always smaller than the money spent by holding it. That is not an argument for overpaying. It is an argument for pricing the seat honestly before you go to market, so that the number you defend is one you actually believe.
Price the seat before the search opens
The fix is unglamorous and it happens early.
- Benchmark externally before the band is approved. Not against your own org chart. Against what companies of your size, in your geography, competing for the same people, actually paid in the last two quarters.
- Price the whole package, not the base. In GTM roles especially, OTE structure, variable mix, and equity often matter more to the candidate than base — and are frequently cheaper for you to move.
- Name the ceiling internally at kickoff. Not the target. The ceiling. If the answer is "we don't know yet," you do not have a search — you have an aspiration.
- Test compensation early with candidates, not late. A serious search surfaces the number in the first conversation. Discovering a $60,000 gap at offer stage means you spent eight weeks interviewing someone you were never going to hire.
A brief that hasn't been priced is not a finished brief. It is a job description with a search attached to it — which is the same failure we described in a long candidate list is a sign nobody read the brief, arriving at a different stage.
The question worth asking at kickoff
Before a search opens, one question separates the engagements that close from the ones that stall at offer: if the best candidate in the market says yes at the top of our range, can we sign it this week without another approval?
If the answer is no, the band is not real. The gap has already been created, and everything that follows — the sourcing, the interviews, the finalist — is being built on top of a number that will not hold. Fix it before the search opens, not after the finalist says no.