Topic 9 — Tools and metrics

Goal: Know the tools recruiters use and the numbers that measure success.

Lesson 9.1 — The ATS is the recruiter's home base

On her first morning at Northwind Robotics, Yara Solis got a laptop, a badge, and a login to something called Greenhouse. She'd never heard of it. By the end of week one she had it open on a second monitor all day, the way she used to keep her sales dialer open. Every candidate she touched lived in there.

That tool is the ATS — the Applicant Tracking System — and it's the central piece of software in recruiting. It stores every candidate, tracks each one through the pipeline stages you learned in Topic 4 (applied, screen, interview, offer, hired), and holds the notes, scorecards, and interviewer feedback attached to each person. When Marcus, the VP of Engineering, asks "what happened with that backend candidate from two weeks ago?", the answer is in the ATS, timestamped, not in Yara's memory.

Think of it as the recruiter's equivalent of the financial analyst's spreadsheet, or the engineer's code editor. It's where the work actually lives.

If it's not in the ATS, it didn't happen. Notes in your head don't survive a busy Tuesday.

Comfort with an ATS is simply expected on the job, and job ads name it constantly: "experience with Greenhouse or Lever required." Imani, Yara's mentor on the talent-acquisition team, told her not to panic about that line. "You're not learning a hundred tools," she said. "You're learning what an ATS does. Once one clicks, the next one is just a different set of buttons."

Lesson 9.2 — The main ATS names (and why you only learn the category)

Yara kept seeing four or five product names and worried she'd have to study them all. She doesn't. But it helps to know where each one tends to live, because it tells you something about the company before you even interview.

ATS	Where it dominates
Greenhouse, Lever	startups and mid-market
Ashby	fast-growing startups (the fastest-growing choice)
Workday, iCIMS	large enterprises

Northwind, at ~90 people and Series B, runs Greenhouse, which fits its size exactly. If Yara ever moves to a 10,000-person company, she'll probably meet Workday instead, and it'll feel heavier and more corporate, because it's built for scale and compliance, not speed.

Here's the freeing part: you learn the category, not the catalog. Every ATS does the same core jobs, store candidates, move them through stages, capture feedback, report on the pipeline. Master one and you can pick up any company's specific stack in days. So when an ad lists an ATS you've never used, you're not unqualified. You're a few buttons away.

Lesson 9.3 — Sourcing platforms and the tools around them

Tracking candidates is one half of the toolkit. Finding them is the other, and that's a different shelf of tools.

The dominant one is LinkedIn Recruiter — the paid, professional tier of LinkedIn built for sourcing (Topic 5). It gives Yara advanced search (filter by title, skills, company, location, years of experience) and InMail, the messaging that lets her reach people she isn't connected to. This is where she found Devon Asante, a strong backend engineer who wasn't job-hunting at all, just open to the right conversation.

Around LinkedIn sits a newer layer of AI sourcing tools — names like HireEZ, SeekOut, and Gem. These aggregate candidate profiles and contact details from across many sites into one searchable place, so Yara isn't stitching together fragments by hand. Some come as browser extensions that surface a person's email or profile while she's looking at a page.

One caution Imani drilled into her: these tools find more people, faster, but a worse fit found faster is still a worse fit. And InMail isn't magic, strong candidates get more of it than they can read, so the message still has to be good (Topic 5). Every tool here supports the craft. None replaces it.

Lesson 9.4 — Communication and scheduling: where the day actually goes

A surprising amount of recruiting isn't sourcing or interviewing. It's coordination, and three plain tools carry most of it.

Email is still the spine of candidate communication: outreach, updates, the offer itself. Slack is where Yara coordinates internally with the hiring team, a quick "Marcus, can you do Thursday 2pm for the Devon onsite?" lands faster than five emails. And scheduling tools like Calendly or GoodTime kill the back-and-forth of booking interviews: she sends a link, the candidate picks a slot that already matches the interviewers' calendars, done.

Why does this matter so much? Because speed is a candidate-experience issue. Every day a strong candidate waits for a calendar invite is a day a competitor can move faster. Tools that shave the dead time aren't a luxury; they protect the pipeline.

And don't underestimate the humble spreadsheet. Plenty of recruiting still runs on one, especially at smaller companies, for tracking a search, planning a hiring push, or pulling a quick report for a hiring manager. Yara keeps a shared sheet for Marcus's open reqs even though Greenhouse could technically do it, because he likes seeing it his way.

Lesson 9.5 — The numbers that measure your work

Six weeks in, Marcus asked Yara a blunt question in their sync: "Are we actually getting faster?" She didn't have a number. She does now, because recruiting is increasingly measured, and a handful of core metrics are the recruiter's vocabulary for answering exactly that.

• Time-to-fill / time-to-hire — how long a role takes to fill. A common US anchor is ~44 days (SHRM's 2025 median time to fill); it swings higher for senior or specialized roles. Faster is generally better, and slow loses candidates (Topic 4). Time-to-fill counts from when the req opens; time-to-hire counts from when a candidate enters the pipeline.
• Pipeline / funnel conversion rates — the percentage of candidates who pass from one stage to the next (e.g., what share of phone screens become offers). This is the funnel from Topic 4, expressed as numbers.
• Offer-acceptance rate — the share of offers candidates say yes to. The average sits around 65–79% (US recruiters often see the lower end), and top teams reach 85%+.
• Source-of-hire — which channel (LinkedIn, referrals, inbound applicants) produces actual hires rather than mere applicants, so you invest where it pays off.
• Quality-of-hire — the hardest to measure and the most important: do your hires perform and stay? A common proxy is the percentage rated "meets or exceeds" at 12 months. A pile of fast placements who quit in six months is not a win.

Yara doesn't need to compute these by hand. Greenhouse reports most of them. What she needs is to read them, which is the next lesson.

Worked example — Reading the funnel to fix a stuck search

Marcus's hardest open req, a senior backend engineer, has been open eight weeks. He's frustrated. Yara opens Greenhouse instead of getting defensive.

The funnel tells a story. She sourced and screened 30 candidates. Of those, 12 reached the onsite, 3 got offers, and only 1 accepted — and Devon, the one she most wanted, declined. Time-to-fill is already past 55 days, well over the ~44-day benchmark. The numbers point at specific leaks, not vague failure.

First leak: 30 screened, only 3 offers. That's a low pass-through, and the interviewer notes show why, every rejection cites the same fifth requirement (a niche framework almost nobody has). The funnel is telling her the bar, not her sourcing, is the problem.

Second leak: 3 offers, 1 acceptance. A low offer-acceptance rate. Devon's decline note says a competitor came in higher on comp.

So Yara walks into the sync with evidence, not apologies: "We've screened 30 strong candidates and none clear all five requirements, because requirement five is rare in this market. And we lost Devon on comp. If we drop that requirement to 'nice-to-have' and move the band up 8%, I think this closes in three weeks." Marcus, who started out skeptical of recruiters, agrees on the spot, because she brought him a diagnosis backed by his own funnel. She loops in Pri from People Ops to confirm the new comp band is legal and inside the company's range. Two weeks later, a re-engaged Devon accepts.

The metrics didn't fill the role. Reading them did.