Find the days before you try to save them

The days lost to a vague brief do not appear as a delay. They appear as rework: candidates submitted and rejected, a shortlist rebuilt in week three, a salary band that was never real. Every one of those cycles costs a week, and the cause is upstream of anything a recruiter does afterwards.

A brief is complete when you can predict rejections from it. That means the must-haves are separated from the nice-to-haves and both are testable, the salary band is what the client will actually pay rather than what they hoped for, the interview process is written down with the number of stages and who is in each, and somebody has said what the last person in this role did all day.

The hard part is not knowing this, it is getting a busy client to sit through it. The lever that works is evidence rather than persuasion: showing an account that nine of their last eleven rejections were on salary converts an argument about process into a conversation about the band.

On most agency desks the longest wait in the process is between a submittal going out and the client saying anything about it. It is also the interval the agency has least direct control over, which is why it tends to be excluded from internal reporting and therefore never addressed.

You cannot make a client respond faster by wanting it. You can change the conditions. Agree a response window at the point of taking the brief and put it in the terms, so chasing is a reference to an agreement rather than a nuisance. Submit in batches the client asked for rather than one candidate at a time, because a single CV is easy to leave in an inbox and a shortlist of three is a decision. Give the hiring manager a way to respond that takes thirty seconds, which usually means a shared shortlist view with a yes, no or interview control rather than a reply-all thread.

And measure it per account. A client who averages nine days is not the same commercial proposition as one who averages two, and that difference belongs in how you prioritise the desk.

Interview coordination rarely appears as a problem because no single step is slow. A day to get panel availability, a day for the candidate to reply, a day to discover the second interviewer is travelling, a reschedule that costs four more. Across three stages that is a fortnight nobody decided to spend.

The fix is structural rather than diligence. Hold candidate-selectable slots against real calendar availability so the negotiation happens once. Book the next stage before the current one ends, while everyone is in the room and engaged. Have a named substitute for every panel member, agreed at intake, so one person's travel does not stop the process.

Reminders matter more than they should. A confirmed interview that a candidate forgets costs the same days as one that was never booked, and a reminder through the channel the candidate actually reads is the cheapest intervention available.

The gap between a decision to hire and an offer in the candidate's hands is often days, and it is nearly always internal to the client: a compensation approval, a signature, a finance sign-off nobody flagged at intake. Ask during the brief who approves an offer and how long it takes. Clients rarely volunteer this and frequently do not know until asked.

This interval matters out of proportion to its length because it is where candidates are lost rather than merely delayed. A strong candidate at offer stage is usually in a process elsewhere, and a week of silence is where a counter-offer lands. Time to fill and offer acceptance rate are the same conversation at this point in the process.

The agency-side work here is keeping the candidate warm with something real to say, and having discussed resignation and counter-offer before the offer exists rather than after it arrives.

A three-month notice period is not inefficiency, and reporting it inside time to fill produces a number that punishes desks for the seniority of the roles they work. This is the strongest argument for measuring time to hire, which ends at offer acceptance, alongside time to fill, which ends at start date.

What you can do about notice is factual. Establish it at first conversation and carry it on the record. Tell the client at submittal rather than at offer, because a start date discovered late is where a placement collapses. Where a client genuinely needs someone sooner, that is a brief-level decision about seniority and buy-out, and it belongs in week one rather than week seven.

Then exclude notice from any internal comparison between desks. Otherwise your executive search desk looks slow next to a temp desk that fills in three days, and you will end up managing the wrong thing.

Software helps with the parts that are measurement and coordination. It can timestamp every stage transition so the intervals are computable without anyone maintaining a spreadsheet. It can flag a submittal that has had no client response beyond that client's own normal window. It can hold interview slots and send reminders through the channel a candidate agreed to. It can surface an offer past its response deadline as a task with the context attached.

It cannot make a hiring manager decide, shorten a notice period, or repair a brief that was never taken properly. Any product that implies otherwise is describing a marketing position rather than a mechanism.

The honest claim is narrower and still worth having: most of the days in a slow search are visible in data you are already generating, and the work of finding them is where the improvement comes from.