Employers

Improving Candidate Quality

Too few applicants and too few good applicants are opposite problems with opposite fixes. Most teams treat both by generating more volume, which only helps one of them.

The short answer

  • Diagnose first. Few applications is a reach problem. Many applications and few good ones is a posting and screening problem, and adding volume makes it worse.
  • Most quality problems are created in the job description, one stage earlier than where they are felt.
  • Scoring applicants against the posting turns a stack into a ranking, which is the difference between reviewing everyone shallowly and reviewing the right ones properly.
  • Quality is also lost after the application: a slow or opaque process loses exactly the candidates who have other options.

Diagnose before you act

Write down two numbers: how many applications you received, and how many you would have been happy to interview. The ratio tells you which problem you have.

Few applications overall. A reach problem. The posting is not being seen by the right people, the pay is out of line with the market, or the role is genuinely scarce. More channels and better targeting help here.

Plenty of applications, few worth interviewing. A definition and screening problem. More volume adds review work without adding candidates you want. The fix is upstream, in what the posting says and who it invites.

Teams reliably misdiagnose this, because both feel the same from the inside: an empty shortlist.

Fix it upstream

Most quality problems start in the job description. A posting that describes responsibilities generically attracts generic applications, because there is nothing in it for a strong candidate to recognise or a weak one to be deterred by.

Three changes do most of the work. Describe the actual work concretely enough that someone can picture their week. Publish the pay range, which removes a whole category of mismatch before it reaches you. And separate day-one requirements from learnable ones, so capable people are not filtered out by a list nobody meant literally.

Covered in full in Creating Great Job Descriptions.

Make screening consistent

Unstructured screening degrades as volume rises. The first ten applications get read carefully, the next forty get scanned, and the last of those is judged by a different standard than the first.

Consistency is the fix: the same criteria, applied to everyone, recorded. That is dull to do by hand, which is why it is usually the first thing to slip and the first thing worth automating.

On GigFinder applicants are scored against the posting, so review starts from a ranking rather than a stack. The value is not that a score decides anything. It is that attention goes to the right end of the list, and the criteria stay the same for the fiftieth applicant as for the first.

Quality lost after the application

The candidates you most want have other options, which means they are the most sensitive to a slow, silent or confusing process. Every day of delay filters your shortlist in exactly the wrong direction.

Three cheap fixes: tell candidates what the stages are and roughly how long each takes; reply to everyone, even briefly; and compress the gap between the final interview and the decision, which is where most late losses happen.

Frequently Asked Questions

First work out which problem you have. Few applications overall is a reach problem, fixed with better targeting and competitive pay. Plenty of applications with few worth interviewing is a definition and screening problem, and adding volume makes it worse. For the second, fix the job description, publish the pay range, and screen consistently against stated criteria.
Usually because the posting is generic. A description of responsibilities in abstract terms gives strong candidates nothing to recognise and weak candidates nothing to be deterred by, so the applications that arrive are equally generic. Describing the actual work concretely, and publishing the pay range, filters far more effectively than a longer requirements list.
Its value is consistency and attention, rather than the score deciding anything. Unstructured review degrades as volume rises: the first ten applications are read carefully and the last forty are scanned against a drifting standard. Scoring against the posting keeps the criteria identical for the fiftieth applicant and puts attention at the right end of the list.
Probably, if the process is slow or silent. The strongest candidates are the ones most likely to have other offers, so delay filters your shortlist in the wrong direction. Telling candidates what the stages are, replying to everyone, and compressing the gap between the final interview and the decision are the three cheapest fixes.

Comparisons

By industry

  • Information Technology
  • Healthcare
  • Sales

Start from a ranking, not a stack

Applicants on GigFinder are scored against your posting, so the same criteria apply to the fiftieth application as to the first.