Recruitment metrics that actually matter for sourcing (2026)

You can recite your time-to-fill from memory. But ask which sourcing channel actually produced your last five hires, or how many profiles you reviewed to build last week’s shortlist, and the answer gets fuzzy. That gap is the problem with most recruiting dashboards: they measure the end of the funnel and ignore the part you spend the most hours on.

If you source candidates yourself, the metrics that change your week are upstream of time-to-fill. This is a practical list of recruiting KPIs worth tracking when sourcing is the bulk of your job: what each one means, why it matters, and a realistic range to aim at. No fabricated benchmarks. Where a hard number does not exist, you get the qualitative version instead.

Want early access? SourceLens is pausing new sign-ups for a short while. Join the waitlist and you’ll be first in when we reopen.

Start with the lagging numbers (but don’t stop there)

Two metrics get quoted in every hiring review. They matter, but they are results, not levers.

Time-to-fill

The days from a role opening to an accepted offer. It is the number your hiring managers care about, and the one most likely to land on a leadership slide.

Why it matters: it tells you how long roles stay open, which maps directly to lost productivity and frustrated managers. But it is a lagging number. By the time time-to-fill is bad, the cause is usually three steps back: a weak search, a slow shortlist, a channel that doesn’t convert.

Benchmark: it swings hard by role. A volume sales role might fill in two to three weeks. A niche engineering or specialist role can take two to three months, and that is normal. Compare a role against the same role last quarter, not against a company-wide average that mixes everything together.

Cost-per-hire

Total recruiting spend for a period (job ads, tool subscriptions, agency fees, your time) divided by hires made.

Why it matters: it keeps budget conversations honest, especially when someone wants to add another tool or post on another job board. On its own it is blunt. A low cost-per-hire on a role you filled with a bad candidate isn’t a win.

Benchmark: there is no universal figure worth quoting, because it depends entirely on role seniority, region, and whether you used agencies. The useful move is to track it per channel, not as one company average. Then you can see which sources drive it up.

The sourcing metrics nobody puts on a dashboard

Here is where sourcers actually live. These rarely show up in ATS reports, but they are the numbers that tell you whether your sourcing is working.

Time-to-source

The days from picking up a role to handing over a usable shortlist. It is the sourcing-specific slice of time-to-fill, and the part you control most directly.

Why it matters: if a shortlist takes you a week instead of two days, every downstream stage starts late. Time-to-source is the lever; time-to-fill is the result. Track it and you’ll spot a stuck search long before it shows up as an open req that’s been live for six weeks.

Benchmark: for a focused search, a working shortlist in a few days is reasonable. If it routinely takes longer, the bottleneck is almost always evaluation: reviewing profiles, not finding them.

Sourcing channel effectiveness

Which channels produce candidates who get hired. LinkedIn search, X-ray via Google, referrals, your own talent pool, inbound. Each one has a different hit rate.

Why it matters: most recruiters spread effort evenly across channels out of habit, not data. When you track hires by source, the picture usually surprises people. One or two channels carry most of the weight, and a couple eat time while producing almost nothing.

Benchmark: don’t chase a target percentage. The signal is the spread. If your hires cluster in one or two channels, double down there and cut the rest. Our talent sourcing strategy framework covers how to log this per search without building a reporting system.

Profiles-reviewed-to-shortlist ratio

How many profiles you open and assess to shortlist one candidate. Review 100 to shortlist 5, and your ratio is 20 to 1.

Why it matters: this is the single best read on how sharp your search and your evaluation are. A bloated ratio means you’re wading through irrelevant profiles, either because the search is too loose or because you can’t tell quickly whether a candidate fits. Both cost hours.

Benchmark: there’s no universal right number. It shifts with role and how tight your search is. Watch the trend instead. A search that needs 100 profiles for 5 shortlisted is doing worse than one that needs 100 for 15. When the ratio improves, your week gets shorter.

InMail and outreach response rate

The share of people who reply when you reach out, whether by InMail, connection request, or email.

Why it matters: a low response rate means you’re either targeting the wrong people or your messages read as templates. It also burns through limited InMail credits fast.

Benchmark: response rates vary by seniority, market, and how personalised your message is, so a single industry figure isn’t worth much. The honest framing: generic blasts sit low, and messages that prove you actually read the profile and understood where someone worked pull noticeably better. The recruiters with the best response rates aren’t better writers. They evaluated each candidate properly before reaching out.

Qualified-candidate rate

Of the candidates who reach the hiring manager, how many are judged genuinely qualified rather than waved through.

Why it matters: this is your quality check on everything upstream. A high reject rate at the hiring-manager stage means your shortlist looks right on paper but misses on substance, often because relevance was judged on job titles alone instead of what someone actually did.

Benchmark: aim for most of your shortlist to survive the hiring-manager review. If half your submissions get bounced, the problem isn’t the manager being picky; it’s the shortlist.

Shortlist quality is the metric hiding behind the others

Most of the numbers above point back to one thing nobody measures directly: how good your shortlist is. A clean profiles-to-shortlist ratio, a high qualified-candidate rate, a short time-to-source: they’re all symptoms of strong shortlist quality.

And shortlist quality comes down to evaluation. You can find 500 profiles in an afternoon. Deciding which 10 are worth a hiring manager’s time is the hard part, and it’s where the hours go. The thing that slows it down is rarely the search. It’s that you don’t recognise most of the employers in your results. “Senior Engineer at Lumera Labs” tells you nothing until you research the company (size, industry, growth stage), and you can’t do that for 240 unknown companies per search.

So you fall back on the logos you recognise and skim past the rest, which quietly drops strong candidates at companies you’ve never heard of. That hurts shortlist quality and inflates the two numbers you do care about: time-to-source and the profiles-to-shortlist ratio.

The under-measured metric: time-per-profile

Here’s the number almost no one tracks: how long you spend evaluating each profile.

It matters more than most KPIs on your dashboard because it multiplies. Three to five minutes per unfamiliar profile across a few hundred profiles is most of your week, gone to research you barely register doing. Cut it to under a minute and the whole funnel speeds up: faster time-to-source, more profiles reviewed, a sharper shortlist.

This is what SourceLens was built to fix. It’s a Chrome extension that adds employer context next to every LinkedIn profile. For each candidate it analyses their last 8 employers on the signals that decide fit (company size, funding stage, industry, B2B or B2C, growth phase, tech stack, region) so you can judge relevance in seconds instead of opening a dozen company pages. It runs in SAFE MODE (only URLs are processed, no scraping) and works on every LinkedIn tier, from Basic to full Recruiter.

The effect shows up in exactly the metrics this post is about:

• Time-per-profile drops because the employer research is already done.
• Profiles-to-shortlist ratio improves because you stop skipping unfamiliar companies and start catching the strong candidates hiding behind unknown logos.
• Time-to-source shortens as a direct result of the two above.

It won’t write your boolean strings or your outreach. That’s still on you. For the search side, our LinkedIn sourcing tips for 2026 covers the techniques that pair well with faster evaluation.

What to actually track this quarter

Don’t build a 15-metric dashboard you’ll abandon by week two. Pick a handful and log them per search in a spreadsheet:

• Time-to-source per role
• Profiles reviewed to shortlist size, per search
• Hires by sourcing channel
• Response rate by outreach type
• Qualified-candidate rate at the hiring-manager stage

Run them for a month and the patterns show themselves. You’ll see which channel to cut, which search to kill, and where your hours actually go. For most sourcers, the answer to that last one is evaluation, which is exactly the metric to attack first.

Pick one open role this week, log the numbers above, and check back next week. To see how the employer-context piece works in practice, read how SourceLens works.