When you’ve run enough searches you begin to see patterns in the ones which fail. At one company, nobody could describe what the hire would actually work on. One founder said he wanted a senior operator and then turned down every candidate who disagreed with him. At another company, the interview process tested candidates thoroughly and folks were reluctant to “sell” anyone on the role. If they weren’t already excited from the first contact, why bother?
The pre-mortem
It’s six months from now, the search failed, and you’re writing the retro. What does it say?
Below is our list, built from real searches which went poorly. Some of these will apply to your search and some won’t.
Some of these are caused by us, others by you. We take a search we should have pushed back on. We see a problem in week two and soften it instead of saying it. What makes the list useful is agreeing in advance on what the early signals are and what happens when one shows up. If we agree at the start that something might go wrong and what we’ll do about it, then when it happens in week six we do that thing. If we never talked about it, week six gets spent arguing about whose fault it is.
Failure modes we work on with you
These are the failure modes Beluga exists to fix. If any of these are true, that’s fine. This is what you’re hiring us for.
You have a story for investors and customers, and no story for candidates
You have a version of the company you’ve told a hundred times. It’s tuned for investors, or customers, or the board. It works. But nobody has written the version for the candidate deciding whether to leave a job that already pays them well.
Candidates have different objections, and they research you differently. They’ll look up who left in the last year, ask what happened to the last person in this role, and check whether headcount grew or shrank. Whatever the fuzzy part of the business or the opportunity is, a strong candidate will find it, and if your answer isn’t clear they’ll assume you don’t have one.
Compensation is where this shows up most often. If you can’t tell a candidate what the equity could be worth and on what basis, “competitive equity” reads as “we haven’t decided.” If the cash is fine but the move is lateral, you need a reason the move is worth making that isn’t the salary.
And candidates aren’t asking these questions in a vacuum. They’re comparing your answers with answers from other companies they are considering. You increase your chances when the answers are given before they have to ask.
Our first job together is building the candidate-facing version of your story, the one that answers the hard questions before they’re asked. Why did the last person leave? Where is the business weak and what are you doing about it? What does the equity or the growth path actually look like? If you don’t have answers, we’ll build them with you. If you refuse to build them, see Part 2.
More outbound won’t fix outbound that isn’t working
If you’ve sent a thousand messages and gotten three replies, the instinct is to send five thousand, add more agencies, and widen the funnel.
That makes it worse. Three replies from a thousand messages means something about the targeting or the message is wrong, and scaling it just tells more people about it. Before we add any volume, we debug: who exactly are we reaching, what are we saying, and why would this specific person care? Usually the answer traces back to the story problem above, or to the mismatch problem below.
The candidate you want isn’t attracted to the offer you’re making
Your ideal candidate profile and your employer value proposition have to agree with each other. Often they don’t. You want someone who has already done this job at a company twice your size, and you’re offering cash ten percent under what they make now, a five-day in-office policy, and a team of two where they currently manage twelve.
There are two fixes: change who you’re targeting, or change what you’re offering. There is no third option where the mismatch resolves itself through persistence. We’ll show you the market data and lay out both paths. The decision is yours to make, and we’ll flag the mismatch early enough that you still have time to make it.
Your interviews thoroughly test candidates while giving them no reason to say yes
Interviews have two jobs: figure out whether the candidate is great, and make the great ones want the job. Most of the loops we inherit do the first one badly and approach the second without a plan, if at all.
We’ll help you design a loop that’s tight: four stages, time-boxed exercises, clear ownership of what each stage evaluates, and that treats every conversation as partly a sales call. The best candidate in your pipeline is also in three other pipelines, and they’re experiencing your process as a preview of what working with you is like.
“Not a fit” is not feedback
If your feedback on candidates is “didn’t feel senior” or “no spark,” we can’t calibrate, and neither can you. Vague feedback usually means one of two things: the evaluation criteria were never defined, or there’s a real criterion nobody wants to state. Both are fixable. We’ll give you a feedback structure that forces specificity, and we’ll notice when your stated criteria and your actual decisions diverge.
We’ll test what you actually want, not what you say you want
We do this deliberately. We’ll calibrate early by looking at real profiles and ensure that the list includes a range, including some ‘wildcard’ profiles: folks who maybe don’t have the stated experience, the logo, the title, but where there’s alternative evidence they could be a strong fit for the role. We do this to ensure we’re all crisp on what exactly we’re looking for, and what’s a red flag versus a yellow flag.
Proxies are useful. Top logos and fast promotions are real signals. But they’re shortcuts, and shortcuts are how you hire someone whose background looks right and who can’t do the job. When you rank the batch, the gap between what you said you wanted and who you actually picked is the most useful data we get. We’d rather find your automatic rejections in week one than discover them in week eight.
Failure modes only you can fix
We can point all of these out but fixing them is something only you can do. If they stay unfixed, the search fails no matter how well we execute, and some of these should stop the search before it starts.
You can’t say what this person ships in their first 90 days
If the role description is all trajectory, “someone who can take us to the next level,” and no deliverables, you’re not ready. You’re hiring the five-year partner and forgetting you have six months of work that has to get done, at minimum, to hit the targets you already reported to investors.
We ask what this person will have shipped by day 90 and the answer comes back as an opaque vision statement. So do the work of splitting the five-year bet from the six-month job. They’re usually two different people. Pick which one this search is for. If you want both, what happens when you can’t have both?
The five-year bet
- What you’re buying: someone who can build the org you’ll need three rounds from now
- How the role reads: trajectory, scope, ceiling
- When you find out you were wrong: eighteen months in
- What it costs you: this half’s targets slip while they build for later
The six-month job
- What you’re buying: specific work that has to ship this half
- How the role reads: named deliverables with dates
- When you find out you were wrong: ninety days in
- What it costs you: you ship now and outgrow them in two years
You haven’t decided what this person actually owns
Related, and just as common: you know where the company needs to be in a year, but you haven’t written down who’s doing what today and what moves off whose plate when this person lands.
Hiring someone to figure out the plan is a legitimate reason to hire. A strong senior candidate will map the gaps and pick the sequence better than you would, and if you’ve already decided all of it, you may be hiring for capacity instead. What they can’t do is decide what they’re allowed to own. That part is yours, and it has to be settled before they arrive, because otherwise they spend their first quarter negotiating it with people who were there first and who never agreed to give anything up.
The uncomfortable part is that giving a new hire real ownership means taking something away from someone who currently has it. Founders avoid this. Do it anyway, before the search kicks off. One page is enough: who owns what today, and what this hire takes over.
| Who owns what today | What isn’t working today | What this hire takes over |
|---|---|---|
| Founder runs outbound and every first call | No repeatable qualification, nothing written down | Owns pipeline end to end from week 3 |
| VP Eng approves every architecture decision across four teams | Decisions wait days for one person, and the teams have stopped proposing anything | Owns platform architecture, with the VP reviewing only cross-team changes |
You want senior wisdom but you’ve built a culture that repels it
Some teams are intense in a way that reads, from the outside, like a cult. That’s a legitimate choice, and intensity attracts certain people. But if you’re hiring for seniority, ask yourself one question honestly: what is a senior person allowed to disagree with you about?
If the answer is nothing, what you want is capacity, and you should hire for that instead, or change the answer. We’ve watched companies burn months rejecting every experienced candidate who wouldn’t perform enthusiasm, without ever noticing the pattern. We’ll show you the pattern, but changing it is on you.
You’re hiring skills nobody on your team can judge
Hiring a CTO with no technical person in the loop does not work, for you or the candidate. The interviews will make you feel confident without telling you anything about whether the person can do the job.
If the role requires expertise you don’t have, get an assessor before we submit the first candidate: an advisor, an investor’s technical partner, a paid expert interviewer. This is one of the few hard gates in our process. We won’t put candidates into a loop that can’t evaluate them. It wastes their time, yours, and ours.
The hire is standing in for a decision you haven’t made
There are two reasons to hire someone before you know what you want. One is that you need expertise you don’t have. The other is that you don’t want to make the call. They look the same from the outside, and only you know which one you’re doing.
If it’s expertise you’re missing, a full-time hire is usually an expensive, risky way to buy it. An advisor, a fractional operator, or three paid conversations with people who have done it will tell you what you need to know in a few weeks, and you’ll write a much better role description afterward.
If it’s a decision you need to take that you’re avoiding, hiring someone doesn’t make it, it just moves it to someone with less context and less authority than you. Hiring a COO because a board member said you need one often means nobody has decided what the COO does. Writing a job description that quietly routes an underperformer’s work to a new person means the new person arrives to do half of someone else’s job, without being told that’s what’s happening. In both cases you get an expensive, confused employee and the decision still doesn’t get made.
The math doesn’t work and nobody’s said it
If the comp band was set by what employee number three makes rather than what the market pays today, the search will confirm the market rate the hard way, one rejection at a time. If the title is a level above what the scope justifies because you can’t pay market cash, you’re attracting title-optimizers. And if your runway funds this salary for seven months, the responsible move might be to not run this search at all.
Failure modes that show up after sourcing starts
These don’t exist at kickoff. They develop while the search is running. Naming them now is how they get caught early. For each one, what matters is agreeing on the tripwire and the response before it happens.
| Signal | What it looks like | Tripwire and response |
|---|---|---|
| The spec grows one candidate at a time | You meet a strong candidate who has a skill you hadn’t asked for, and you start wanting that skill in everyone. It happens a bit at a time and never gets written down, so six weeks in you’re hiring for a different role than the one we scoped. | Any requirement that wasn’t in the kickoff doc gets written down and re-scoped, or it doesn’t exist. |
| Feedback slows, then stops | Candidate feedback taking more than 48 hours, twice in a row, means the search has slipped down your priority list, and candidates notice. | Pause sourcing and talk, rather than pouring more people into a stalled pipeline. |
| The loop grows a stage every time you get nervous | Every past bad hire tempts you to add a step. Seven-stage loops produce dropout among exactly the candidates with the most options, without producing better decisions. | The loop designed at the start is the one we run. Changes go through the same conversation as spec changes. |
| One hesitation becomes a veto | If one interviewer’s doubts are enough to reject a candidate, then every close call becomes a rejection. Nobody decided to run it that way, but that’s how it works. | Decide the tiebreak mechanism up front, while it’s abstract and easy. |
| You say X, you approve Y | Sometimes your decisions and your stated criteria drift apart. That’s normal, and it means the search is teaching you what you actually want. | That only helps if we notice and rewrite the spec. Otherwise we both keep sourcing against a stale role description. |
| The search stops without being paused | Fundraise heats up, product breaks, priorities shift. | Fine, but tell us, and we’ll pause properly. Candidates left hanging mid-process experience it as being ghosted, and they talk. |
| The offer has no closing plan | An offer sent by email with a one-week deadline is not a closing plan. | By the time we’re extending an offer, we should know what this candidate cares about, who from your team calls them, and what happens when the counteroffer lands. |
Three questions worth answering before you start a search
You don’t need much preparation to start a search. But if you want to set yourself up for success, reflect and answer these three questions.
What does this person ship or solve in their first 90 days, and what does the company miss if they don’t?
Look at the other roles on the market right now. Why would you choose yours?
It’s six months from now and the search failed. What’s your best guess why?
The answer to question three is usually right. Founders almost always know their own failure mode, and the hard part is telling us early enough that we can do something about it.
The roles that get filled on time are the ones where everyone named the risks while there was still time to do something about them.