Job Search Funnel Calculator

A job search is a funnel with four observable stages: applications sent, applications that reached a first interview, applications that reached a final round, and applications that produced an offer. Each stage counts applications, not events — one job that involved five conversations still counts once.

Counting this way is what makes the conversion rates comparable to each other. It also means the numbers can only fall as you move down the funnel, which is why the calculator rejects a stage larger than the one above it rather than quietly reporting a conversion rate above 100%.

Ties resolve to the earliest step. That is a judgement, and worth stating: an early bottleneck constrains every stage beneath it, so when two steps are equally bad the earlier one is usually where attention is better spent.

Most job-search frustration is described in aggregate — the search is going badly — when the actual problem sits at one specific step. The three shapes below call for different responses, and telling them apart is most of the value here.

• The drop is at applications → first interview. Almost nothing reaches a human conversation. The lever is what you are sending and where you are sending it, not how you perform once invited.
• The drop is at first interview → final round. You are credible enough to be met but not to be shortlisted. This is usually about how the early conversation is handled, or about a genuine mismatch between the roles targeted and the record.
• The drop is at final round → offer. The most expensive place to lose, and the most competitive. Reaching finals repeatedly is itself evidence that the top of the funnel is working.

Funnels get thin fast. A search with 60 applications, 6 interviews and 2 finals is measuring the second and third steps on 6 and 2 cases respectively. At that size a single outcome can move which step looks worst, and the calculator says so when the stage it is measuring from is small.

The right use of a thin funnel is directional: it tells you where to look first, not what is wrong. Treat the top of the funnel, where the counts are largest, as the most trustworthy part of the picture.