The metric that changed how we hire wasn’t time-to-fill or cost-per-hire. It was acceptance rate. We were filling roles fast, but candidates were declining offers at a rate that masked a sourcing problem: we were pulling from a pool that looked right on paper but wasn’t actually a good match.

Once we started tracking acceptance rate as the front-line health signal, everything upstream got clearer. Roles with sub-70% acceptance almost always had the same issue: the job description and the sourcing criteria weren’t aligned. We’d describe the job one way, then search for candidates using a different mental model. Closing that gap got our acceptance rate to around 83%, and shortened the search cycle because we were chasing fewer people who weren’t actually interested.

For a long time we tracked what most companies track, time-to-hire, cost-per-hire. Easy to measure, easy to report. But we kept having moments three months into someone’s tenure where the team would think “how did we miss this?” Interviews went well, resume was strong, something still wasn’t adding up.

When we started correlating assessment scores against actual 90-day performance reviews, the pattern was uncomfortable. We were consistently overweighting how someone performed in interviews and underweighting what the structured assessment was already telling us. We had the right signal the whole time. We just weren’t trusting it.

Once we rebalanced that weighting, the quality of hires improved noticeably. Not the volume, the quality. The metric that unlocked it wasn’t a new data point. It was closing the loop between the hiring decision and what actually happened after.

Most companies measure the process. Very few measure whether the process produced the right person. That gap is where almost every hiring problem lives.

When I launched Epigen Bio Talent, one of the advantages was that I didn’t have to inherit old habits. I could build the recruiting process with analytics in mind from the start rather than layering measurement on after the fact.

Earlier in my career, like many recruiters, I relied heavily on intuition and a handful of lagging indicators—time to fill, placements made, and general client satisfaction. Those are still important, but they don’t tell you much about why a search succeeded or where quality was being lost along the way.

Today, at EBT, we track the process much more closely. We look at where candidates enter the funnel, how they progress, where they drop out, and which sourcing channels consistently produce people who make it through final interviews and perform well after placement.

The single most useful metric for us has been interview-to-offer ratio by source. In life sciences recruiting, the candidate pools are specialized and often quite small. If a particular channel is generating a lot of resumes but very few candidates who reach the offer stage, that tells us quickly that the issue is quality, not quantity. Conversely, a smaller source that reliably produces finalists deserves more attention.

What’s changed most is that decisions feel less anecdotal. We still use judgment, of course, but the data helps us focus our effort where it has the highest probability of producing strong, durable placements. In a field as technical and relationship-driven as life sciences, that combination of analytics and human insight has been extremely valuable.

The turning point for us came when we stopped measuring hiring speed and started measuring hiring success a year later. I’ve seen companies celebrate fast recruiting numbers, only to realize months later they hired the wrong people. A bad long-term hire costs far more than an open role ever will.

The metric that changed our approach was first-year retention tied to performance. Were new hires still with the company after a year, and were they actually succeeding? That changed the interview questions we asked and the backgrounds we prioritized.

I worked with a manufacturing company that struggled to retain leadership talent. On paper, the candidates looked ideal: polished interviews, recognizable companies, impressive resumes. But the analytics exposed a pattern: many came from highly structured environments and struggled in a culture where leaders had to make fast operational decisions with limited support.

Once we shifted the hiring process toward adaptability and decision-making under pressure, retention improved within months.

“One bad hire can ace every interview. Analytics matter because they expose patterns experience alone can miss.”

The best hiring decisions still involve human judgment but good analytics keep that judgment honest.

Time-to-fill stopped being useful for us about a year ago. We still report it because someone asks, but the number does not change how we hire anymore.

We tracked it for 2 years and the number kept improving while the actual hires got worse. Faster fills meant we were saying yes to candidates who looked good on paper because the role was bleeding. The metric we replaced it with is 90 day retention split by source. It is slower to read and it tells you which channel sends people who stay.

The single insight that changed candidate quality was noticing that referrals from our 2nd year employees outperformed referrals from senior hires by a wide margin. I do not have a clean theory for why. We doubled the referral bonus for that band and the next 6 hires were the strongest cohort we had run.

I use HR analytics to be more intentional in the ways we recruit people. Instead of just reacting to vacancies as they come up, we make recruitment a key focus of our business plan. We analyze application flow and how many people move from one stage to the next in the hiring process. This helps us identify bottlenecks and where the best people lose interest.

The Important Thing We Measure

The thing that helps us find the best candidates is our First-Year Performance Rating by Source metric.

We look at how new employees do after one year and where they originally came from. This gives us an idea of which sources give us great employees and which ones just give us a lot of people.

What We Learned From This

When we looked at the data we found out that people who came from employee referrals and special job websites did better in their first year than people from larger job boards. This prompted us to make some changes:

* We Changed Our Budget: We stopped spending so much money on larger, aggregated job boards and started paying employees more to refer their friends.

* We Focused On The Right Places: We started looking for people on websites where our best employees came from.

* We Got Rid Of The Noise: By focusing on the best sources, we got fewer people applying, which meant our team could spend more time with the higher quality candidates.

By looking at this data we make sure we are spending our time and money on strategies that will add the most value in the long run.

The best metric I tracked at Performance One Data Solutions was speed by source. We spent six months watching how candidates moved, and referrals were clearly faster and stuck around longer. It showed us exactly where to focus our recruiting efforts. You really should run these numbers every few months. It keeps your hiring honest and tells you where to spend your time.

When I’ve been working on HR analytics, I’ve found that it makes the most sense to use it to get the most out of your hiring strategy, not just keep an eye on how many people you’re hiring. I’ve built Power BI dashboards for recruitment that pull in data from CRM systems, and they display the basics like cost per applicant, time to get a candidate in the door, and how a particular group of candidates is doing – but its the finer details that really make the difference.

By drilling down into the specifics, we were able to figure out which channels were the best value for money when it came to specific jobs, and which ones were just bringing in a load of applicants without actually getting anyone good. That let us shift our budget over to the channels that consistently delivered the good stuff and got people hired a lot faster who were a better fit.

One metric stood out as being absolutely priceless in this process – cost-effectiveness of recruitment channels, broken down by job role or skill set. What it gave us was a clear sense of which candidates were any good – because the channels that were cheapest in the long run were also the ones that were producing people who moved further along in the hiring process and ended up performing better on the job.

Once we started focusing on that metric, our whole hiring process got a lot smarter, and the quality of the candidates shot up, without us having to spend more money on hiring.