Engineering pipelines reward speed over substance. The candidates who apply in the first hour get read closely. The ones who apply on day four are assessed by whoever is left with time, against a standard that has quietly drifted since Monday.

Hexlid hires across four engineering teams, and the roles that took longest to fill were consistently the ones that stayed open past the first burst of applications.

A queue that was really a clock

Hexlid ran the funnel most engineering teams run. A role opened, applications arrived in a burst, and the first thirty were read properly. The remainder were skimmed for a familiar employer or a familiar stack.

That is not a hiring standard, it is a reading schedule. It also selected hard for candidates who happened to be job-hunting on the day the post went up, which has little to do with whether they could do the job.

Recruiters could see the pattern in the data but not fix it by hand. Reading every application to the same depth simply took more hours than a hiring manager had in a week. Nobody was willing to cut the depth to buy back the hours.

Evidence before order of arrival

Hexlid uses Yardstick Sourcing to surface what a candidate has actually built and shipped: repositories, contributions, the shape of the work rather than the shape of the CV. The same evidence is weighed for everyone, regardless of when they applied.

The brief is written in plain language: the systems the team runs, the problems the role will own, the level of ownership expected. Sourcing works from that, and the evidence it returns is attached to the candidate so a reviewer can check it rather than take it on trust.

A late applicant now arrives with the same evidence package as an early one, so the shortlist stops rewarding whoever refreshed the job board first. The reviewer reads the work, not the timestamp on the application.

What the team still owns

The scorecard stays Hexlid’s own: engineers agree what good looks like before the role opens, and every shortlist is read by a named engineer who can overrule anything on it. The evidence makes the conversation shorter, not shorter-handed.

What the evidence changed

Engineers are hired on evidence, not on who applied first. Since the brief-first approach went live, 3× more shortlisted candidates have applied after the first three days of a role being open, up from a small handful before.

Every shortlist is still 100% reviewed by a named engineer before an interview is offered. Reviewing one now takes -40% less time, because the evidence arrives already organised rather than needing to be dug out of a CV or a GitHub profile.

Late applicants are read to the same standard as early ones, which is the only part of the funnel that was ever really broken. Reference checks and the final culture conversation are still done the old way, by a person, on a call.

What they are doing next

Hexlid is now testing the same brief-first evidence against its internal transfer process, where a similar first-hour advantage was quietly shaping who moved teams.