The intake meeting decides the placement before the first CV

Seven questions, put to the client before any search begins: why the role is opening now, who decides, which criteria genuinely rule a candidate out, what the budget is, what the interview schedule looks like, what failure looks like, and how many agencies are working the same need. The last one decides whether the role is worth taking. The other six decide whether it can be filled, and no amount of sourcing talent makes up for an answer that is missing.

Yet the intake meeting is the step the trade rushes most. It happens on the phone between two appointments, someone writes down the title, three technologies and a salary band, and the search starts. Three weeks later the client turns down the fifth candidate for a reason they had never mentioned, and everyone discovers that the search was for a job that only ever existed in the call notes.

Why does the intake matter more than sourcing?

Because it is the only moment when the probability of placement can still be changed, and everything that follows merely carries it out.

The orders of magnitude are well known. A role worked on contingency against other agencies fills 40 to 55% of the time, against 70 to 80% on exclusivity and 75 to 90% on a paid retainer, and we have shown that the real price of a contingency role is the fee multiplied by that rate: at a 20% fee and a 45% fill rate, an open role earns 9% of salary. Those ranges are averages, and they blend well-qualified needs with the rest. The intake is where you find out which end of the range the role you just received belongs to.

Misalignment is not a recruiter’s hunch. In the survey Metaview ran in 2026 among 505 recruiting leaders and hiring managers, only half of searches, 49%, start from requirements that both sides clearly share. The same respondents rate the relationship good or excellent 90% of the time, and 58% of them still contemplate working around their counterpart. Everyone gets along, and nobody is talking about the same job.

The gap is starker still when each side is asked whether they communicate enough. In a survey published by HRO Today, 31% of hiring managers and 13% of recruiters completely agreed that the level of communication between them was appropriate. Those are figures from inside a single company, between colleagues; between an agency and its client, there is no reason to think things are better.

What are the seven questions that qualify a need?

They fit on one page, and their order matters, because the first ones make the later ones easier to ask.

Why is this role opening now?

A replacement, a new position, a project starting on a fixed date: the answer tells you where the urgency comes from and whether it is real. A replacement after a difficult departure often hides the actual criterion, which is not to hire the previous person again. A dated project, on the other hand, gives you a deadline the client will keep, because it does not depend on them.

Who decides, and who can say no?

The person giving the brief is rarely the only one who signs off. You need the name of everyone who will meet the candidate, the order of the interviews, and above all the person who can veto without having attended the intake. A CTO who discovers the role at the third interview is the most common cause of feedback that contradicts everything agreed.

Which criteria genuinely rule a candidate out?

The test is simple and it changes the conversation: would you reject a candidate who is perfect on everything else but does not meet this criterion? If the answer is no, it is a preference, and it gets written down as one. Beyond three deal-breakers the list no longer describes a job but an ideal, and the search that follows looks for candidates no agency will ever find.

What budget, measured against today’s market?

A salary or a day rate gets tested against the market during the intake, not after the third rejected candidate. If the budget sits below what the requested profile commands elsewhere, say so while the client can still choose between raising it and dropping a criterion. Said three weeks later, the same observation sounds like an excuse.

Which interview slots, booked right now?

Ask for the slots before you have a candidate, along with a committed feedback time, forty-eight hours for instance. We have shown that a placement’s lead time is lost in the waiting rather than in the work, with a 45-day median, 18 days in the top decile and 81 in the bottom one. Slots blocked during the intake are the only part of that waiting a recruiter can settle before it starts.

What does failure look like?

The most useful question in an intake is the one least often asked: which candidate have you already seen and turned down, and why? Two or three real profiles, rejected or shortlisted, say more than a list of skills, because they force the client to put into words what they had not written down. It is also the best way to surface a criterion nobody dares put in a job description.

How many agencies are working the same need?

The answer moves the probability of placement more than all the others combined, and it is still the one people avoid asking. It opens the only negotiation that matters: what exclusivity would buy the client in time, in a guaranteed number of profiles by a set date, and in a search reaching candidates who never apply.

When should you turn a role down at the end of the intake?

When several answers are missing at once, because the weaknesses of a need compound rather than cancel out.

A role open to four agencies, with a decision-maker you will never meet and a budget below market, is not a hard role: it is a role whose odds sit at the bottom of the range, worked at the same cost as the others. Declining is not final. You decline by offering the terms on which you would take the same need, exclusivity over six weeks for example, or a revised budget, and by asking to be called back if the position is still open.

The most common outcome, in fact, is neither yes nor no but a rewrite. The intake ends with a need reworded together with the client, one criterion dropped, a schedule set, and that version is the one you work. The client gains as much as the agency, since they discover their own contradictions before seeing ten candidates.

The same reasoning applies to IT consultancies staffing a project rather than a permanent job. The need comes from a buyer or a project lead on the client side, with a target day rate and a start date, and the same questions apply almost word for word. Cobalt describes what changes when you recruit mission by mission rather than profile by profile, and the intake is the first place that difference shows.

What can an agent prepare before the intake?

The client’s history, at the moment it is useful, rather than in a report nobody opens before the call.

How many roles this client has opened with you over two years, how many were filled, how long they usually took to give feedback, which candidates they rejected and for what written reason. All of that already sits in your software, and it changes the conversation: you do not raise the question of timing the same way with a client who answers in two days and one who takes twelve. It is a task whose result can be checked at a glance, which makes it a good first candidate for delegation.

It is worth saying straight away what this does not do. An agent does not raise a fill rate, and it does not speed up a client’s decision. It makes available, at the time of the intake, the calculation you would have done if you had an hour to spare, which is already a lot when four calls are lined up for the afternoon.

What should it record afterwards, and who approves?

The need as it was agreed, in the system of record, and the summary to send the client, as a draft.

The first part stays internal and needs no approval: the seven answers, deal-breakers kept apart from preferences, the budget and the slots, written once on the role’s record. Every piece of client feedback on a candidate is then compared against that version. When a rejection cites a criterion that was never in the intake, the agent flags it with the intake sentence and the feedback sentence side by side, instead of producing a match score nobody could challenge, which is our argument about the match score that ranks without explaining.

The second part leaves the company, so it follows a different rule. The summary sent to the client commits the agency to a budget, a timeline and a scope: the agent drafts it, a person reads it and sends it. It is the same distinction that separates a meeting recap from its follow-up, and it holds for the same reason: what reaches a client carries your name.

This is exactly the work Balt does around an intake for us: the client’s history in the conversation before the call, the need recorded afterwards, the message to the client ready for review. The intake itself stays yours.

What if the need changes mid-search?

It will, and the written intake exists precisely so that you see it change rather than to stop it from changing.

A client who discovers what they want by meeting candidates is acting in good faith, and that is often how a need becomes precise. The trouble is the silent change, the one you piece together afterwards from five successive rejections. With a recorded intake, drift shows up at the first piece of feedback that departs from it, and the conversation it calls for happens with the original version in front of everyone.

That conversation is a new intake, a shorter one, and it deserves the same questions. If the added criterion pushes the role above budget, or a new decision-maker appears, the probability of placement has just changed, and the decision to keep working the role gets made again along with it.

What remains is what happens once the intake is done well and the search is underway: candidates waiting for an answer while the client thinks it over. That is where lead time leaks away, and silence there costs more than bad news.