Ask any hiring manager who's run enough searches which candidates drop out mid-process, and you'll get the same answer: it's usually the strongest ones. That's not bad luck, and it's not a coincidence worth shrugging off. It's the predictable output of a specific mechanism: your best candidates have the most alternatives and the least patience for a long process, so a slow funnel doesn't just delay your hiring, it actively filters for the candidates with the fewest other options. If your process is slow, you're not failing to hire your best candidate. You're running a filter specifically engineered to select against them.
The mechanism: your funnel filters on patience, and patience correlates with weaker options
A hiring funnel is, whether anyone designed it this way or not, a filter. Every additional round, every week of waiting for feedback, every scheduling delay is a point where a candidate can either stay in the funnel or leave it. Candidates don't decide to stay or leave based on how much they like your company in the abstract, they decide based on what else is available to them right now. A candidate with no other live conversations has every incentive to wait out your process, because waiting costs them nothing. A candidate juggling two or three other serious conversations is weighing your process against real, moving alternatives, and every week you take is a week those alternatives get to move ahead of you. The result isn't random attrition. It's a filter that keeps the candidates with fewer options and loses the ones with more, which are disproportionately your strongest candidates, because strength in this market is exactly what generates other options.
The numbers: your process length vs. how fast the market moves
The typical hiring process for a senior AI or engineering role runs somewhere between four and eight weeks from first interview to signed offer, once you account for scheduling gaps, multiple rounds, feedback consolidation, and offer approval. Meanwhile, a genuinely strong candidate in an active search is often fielding a first offer within two to three weeks of starting to look, sometimes faster if they're being actively recruited rather than applying cold. That gap, your process taking twice as long as the market takes to produce a competing offer, is the entire mechanism in numbers. You don't need every candidate to have a faster alternative for this to hurt you consistently, you only need your strongest ones to.
| Stage | Typical company timeline | What a strong candidate experiences in parallel |
|---|---|---|
| Initial screen to first interview | 1-2 weeks (scheduling delay) | Already in first interviews with 1-2 other companies |
| Full interview loop (4-6 rounds) | 2-4 weeks | A faster competitor's loop (2-3 rounds) completes and extends an offer |
| Internal feedback + committee decision | 1-2 weeks | Candidate receives a first offer, sets a decision deadline |
| Offer approval + negotiation | 1-2 weeks | Candidate accepts elsewhere before your offer is finalized |
This isn't a candidate-experience problem, it's a selection problem
Companies that notice this pattern often respond by trying to make the slow process feel better, warmer emails, more frequent check-ins, a faster response to questions. That helps at the margins, but it doesn't fix the underlying mechanism, because the issue was never that candidates felt bad about waiting. The issue is that the strongest candidates have somewhere else to be, and no amount of friendly communication changes the fact that a six-week process is still six weeks. Better candidate experience is worth doing, but treating it as the fix mistakes a symptom for the disease. The actual funnel length is the disease.
Why the candidates who ghost are usually the ones you wanted most
This is the part hiring teams find counterintuitive until they see the pattern a few times: the candidate who goes quiet after round three, who stops responding to scheduling emails, who takes two weeks to reply and then declines, is disproportionately likely to have been your strongest option, not your weakest. A weaker candidate with fewer alternatives has more incentive to stay engaged through a slow process, because staying engaged costs them little and the downside of dropping out is real. A stronger candidate weighing your process against a faster, equally attractive alternative elsewhere has every incentive to take the option that resolves first. The silence isn't disinterest in your company, it's often the opposite: enough interest elsewhere, moving faster, to make waiting on you not worth it.
The actual fix is shortening the funnel, not softening it
If the mechanism is process length, the fix has to target process length directly: fewer rounds, faster scheduling, decision-makers involved earlier so feedback doesn't need a second committee pass, and offers that go out within days rather than weeks of a final round. None of that requires lowering the bar on who you hire, it requires removing the dead time between decisions, most of which was never adding evaluative value in the first place. A shorter funnel keeps the same filter for quality and removes the filter for patience, which is exactly the filter working against you right now.
