Nobody asks this question at the start of a search. They ask it with the offer letter drafted, the approval chain finally cleared, and a quiet feeling that something is off. That feeling is usually right, and it is usually specific.
A declined offer almost never looks surprising in hindsight. The process took five weeks. The salary range was never confirmed. The candidate met four engineers and nobody they would report to. Each of those was visible days or weeks before the offer went out, and each was fixable then. The reason declines feel like bad luck is that nobody wrote them down while there was still time to act.
So here is the list. Nine things worth checking before you press send, why each one matters, and what to do about it if the answer is wrong.
First, what the market average will and will not tell you
Offer acceptance runs around 84 per cent globally in 2026, with Australia a few points below at roughly 79 per cent. Those figures are useful for calibrating expectations across a year of hiring and useless for the offer in front of you. An average of a market is not a probability for a case, and treating it as one is how teams end up comfortable with a process that is quietly losing them people.
What travels better than the average is the reasons. Across the market, candidates who decline at the final stage cite the same handful: a counter-offer from their current employer, compensation that misaligns late, a process that drags, and culture or flexibility surfacing at the end. Cutting across all of them, poor communication is the single most cited complaint, named by around a quarter of those who declined. We set those out in full in what drives offer acceptance, and how they differ by company stage in why engineers decline offers by company stage.
The checklist below is those reasons, turned into things you can look at today.
The money: three checks
1. Where the offer sits in the market band
Not whether it is "competitive", which is a word that means nothing to the person receiving it. Where it actually falls between the 25th and 75th percentile for that discipline, level and city. An offer below the floor of the band will be found out, because candidates check, and the ones you most want to hire check hardest. An offer in the bottom quarter is in band but has nothing held in reserve, which matters enormously if a counter-offer is coming.
The salary checker will give you the band in about ten seconds. If the offer is below the floor and the budget will not move, that is a real answer, but then the offer needs something else that is worth the gap and that you can name out loud.
2. The uplift on what they earn now
A candidate on 160 looking at 168 is being asked to take on all the risk of changing jobs for eight thousand dollars, which their current employer can find in an afternoon. Under about 10 per cent, assume a counter-offer is coming and start building conviction now rather than after they resign.
If you do not know what they currently earn, that is its own risk: you cannot tell whether the offer is worth moving for, or how cheaply it could be matched. It is a normal question and the answer changes what you should offer.
3. Whether a range was agreed in writing before the technical interview
This is the most preventable decline in hiring and it still happens constantly. If the candidate expected 220 and the offer arrives at 190, the conversation ended weeks ago and everyone has been wasting time since. Confirming a range in writing before any technical stage feels blunt. It saves weeks, and in a market where strong engineers have options, transparency reads as confidence rather than weakness.
The pace: three checks
Three of the nine are about speed and contact, and that is deliberate. Communication is the most cited failure and it does not show up as a single number. It shows up as elapsed weeks, as round count, and as silence at the end.
4. Days from first call to offer
Inside three weeks, enthusiasm has not had time to decay. Past four, a competing process has time to start and finish. Past five, you are relying on the candidate to stay interested through a gap you have given them no reason to sit through.
The fix is almost never the interviews themselves. It is the gaps between them. Pre-block interviewer time and pre-agree the approval chain before the role opens, because that is where the weeks actually go.
5. Number of interview stages
Three is proportionate. Four is defensible if each round is clearly doing different work. Five is where strong engineers start reading the loop as a preview of how decisions get made once they are inside, and six is losing you candidates before the offer rather than at it.
If you are at five, one of those rounds is almost certainly there because nobody removed it. Check whether two rounds are asking the same question of the same person.
6. The delay between the final interview and the offer landing
Two days and the candidate never had to wonder. A week of silence after the last round is where doubt sets in, and past that they have generally assumed the answer is no and re-engaged elsewhere. Even a holding message resets the clock. Silence is not read as deliberation, it is read as a lack of interest.
Conviction and fit: three checks
7. Have they met the engineering leader they would report to?
Every interview is a sell as well as an assessment. A candidate who can picture their first ninety days and who they would work for has something concrete to weigh a counter-offer against. A candidate who has met four peers and no leader has an abstraction. Half an hour with the right person before the offer goes out is one of the cheapest interventions available.
8. Has the working pattern actually been agreed?
Not assumed, agreed, and ideally in writing. Flexibility is the reason that most reliably surfaces at the very end, when it is least fixable and most expensive. Unagreed is not the same as fine. Ten minutes now or a declined offer later, and if it genuinely cannot move, say so before the offer rather than discovering it during the notice period.
9. Are they holding a competing offer?
Worth asking again immediately before the offer goes out, because it changes late and it changes everything. A competing offer turns every soft item above into a hard one: each is now being compared side by side against somebody else's version rather than judged on its own. Speed is the only lever you fully control at that point.
Run it, rather than reading it
We built the Offer Acceptance Check to make this list something you do in two minutes rather than something you agree with and forget. It applies all nine checks, scores the offer out of 100 with fixed weights, ranks what is costing you most, and gives each item a fix. It runs entirely in your browser and nothing you enter is uploaded.
One thing it deliberately does not do is give you a probability. The weights are our judgement about relative importance, not a regression, and we hold no dataset that would support a calibrated likelihood. Anyone offering you one for a single offer is guessing with a decimal point attached. What fixed weights buy is comparability: the same offer always scores the same, so a fixed offer can be measured against the original and two offers can be compared honestly.
The two failure modes people miss
Assuming money is the problem. It is the most visible lever and the one everyone reaches for, but six of the nine checks above cost nothing. A team with a fast, well-communicated process and in-band pay will beat a team paying five per cent more over five weeks, and does so regularly.
Fixing it at the offer. Every item on this list is cheaper to fix on day three than on day thirty. By the time the offer letter is drafted, the process length is already a fact, the rounds have already happened and the leader meeting either took place or did not. The checklist is most useful run at the intake conversation and again at the halfway point, which is why we run a structured intake as standard.
You rarely lose an offer at the offer. You lose it somewhere in the four weeks before it, and find out at the offer.
FAQ
How do I know if my offer will be accepted?
You cannot know, but you can check the things that reliably cause declines before you send. Nine are worth checking: where the offer sits in the market band, the uplift on the candidate's current pay, whether a range was agreed in writing before the technical interview, elapsed time from first call to offer, the number of interview stages, the delay between final interview and offer landing, whether they have met the engineering leader they would report to, whether the working pattern has been agreed, and whether they are holding a competing offer.
What is a good offer acceptance rate for engineering roles?
Offer acceptance runs around 84 per cent globally in 2026, with Australia a few points below at roughly 79 per cent. That is a market average rather than a target: it says nothing about a specific offer, and a team running a fast, well-communicated process with in-band pay should expect to sit above it.
Does paying more guarantee the offer is accepted?
No. Money removes one reason to decline, and it is the most visible one, but three of the nine checks here are about pace rather than pay. Poor communication is the single most cited complaint from candidates who declined, and no amount of base salary fixes a five-week process with a week of silence at the end.
How long should an engineering hiring process take?
Aim to be inside three weeks from first call to offer, with no more than four interview stages and no more than two days between the final interview and the offer landing. Past four weeks a competing process has time to start and finish, and a five-round loop reads to a strong engineer as a preview of how decisions get made once they join.