Interview-to-Offer Ratio

It measures the yield of your interview loop, and by implication the accuracy of everything upstream of it. Count candidates who entered at least one interview in a period, then count how many of that same cohort received an offer. The result tells you how much interviewing the business buys per offer it makes. That is a real budget line: every extra loop consumes senior time that could go to product, customers or existing staff. When a ratio like this sits far below where the team was a quarter ago, the screen has stopped doing its job, and interviewers are absorbing the cost. Analytics inside an ATS built for recruiters will produce the cohort automatically, which matters because manual counts drift.

No, and this is where the metric gets misread. A ratio near the top of its range can mean the screen is excellent and only viable people reach the loop. It can equally mean interviewers are rubber-stamping: nobody wants to be the person who blocked a hire during a hiring push, so weak signals get rounded up. Tell the two apart by looking at what happens after the offer. If first-year attrition is climbing and new hires are ramping slowly, the loop is approving people it should have declined. If retention holds and managers report strong hires, the high ratio is genuine efficiency. A structured interview question set with written scorecards makes rubber-stamping visible, because a weak candidate leaves a thin evidence trail.

Counting rounds instead of candidates is the common one. A five-stage loop for one person can be logged as five interviews, which drags the ratio down and makes a well-run process look wasteful. Decide once whether the denominator is unique candidates who reached any interview, and hold to it. The second trap is period mismatch: offers made this month often belong to candidates interviewed last month, so an unmatched cohort produces noise that gets read as a trend. The third is dominance. One high-volume requisition can supply most of the interviews in a quarter and set the team number by itself, hiding a very different picture on every other role. Segment by job family before reporting anything.

Fix the screen before touching the loop. A falling ratio means more people are getting through to interviews who were never going to pass, so the first move is to review the last set of rejected-at-interview candidates and ask which signal was available earlier and went unused. Often it is something concrete: a licence, a specific system, willingness to work a shift pattern, a realistic salary expectation. Add that to the phone screen and the ratio recovers without anyone changing their standards. The second move is loop design. Cutting a redundant round rarely lowers quality, and it returns hours immediately. Track the change with the rest of your recruiting metrics so you can see whether quality held after the loop got shorter.