Not every hiring metric deserves attention.
Kristina Golovko
MindDesign
The metric hierarchy
Signal
Tells you something true about hiring quality
Decision
Changes or confirms what you would do
Action
Drives a specific improvement
Outcome
Better hiring as a result
A metric that doesn't change a decision or drive an action is reporting theatre. Track less. Act on what you track.
Why this matters
There is no shortage of things you could measure in hiring: applications per role, source of hire, interview completion rate, offer acceptance rate, diversity of pipeline, time-per-stage, candidate NPS, cost-per-hire, time-to-productivity, 90-day retention... The list is long and the temptation to track everything is real.
But metrics that don't change decisions are not analytics — they're reporting. They consume time to collect, time to review, and produce the impression of data-driven hiring without the reality. The companies with the best hiring outcomes are rarely the ones with the most comprehensive dashboards. They're the ones that track a small number of high-signal metrics and act on them consistently.
Choosing which metrics matter requires asking one question about each: does this metric, when it changes, tell me something that would change what I do? If yes, track it. If no, stop.
Founder reality
Audit the metrics you're currently tracking before adding more:
For each metric you currently track: when did it last cause you to make a different decision?
Are there decisions you're making in hiring without a metric to inform them — where a simple measurement would help?
Is there a metric you track primarily because it was always tracked — without evidence that it improves anything?
If you could track only three hiring metrics, which three would give you the most useful signal about whether hiring is working?
Are the people who collect the metrics the same people who act on them — or is there a disconnect between data production and data use?
The three-metric question is a useful forcing function. The answers usually reveal what actually matters — and what is being tracked out of habit.
The signal set
This is not a minimum dashboard. It is a sufficient signal set for most founder-stage and Series A/B hiring programmes.
Quality of hire at 90 days — does the process produce strong performers?
The most important metric. Measured as: percentage of hires rated strong at 90-day review against a pre-defined contribution standard. If this is declining, something in the evaluation process needs attention. If it's improving, the process is working. Nothing else tells you this directly.
Time-to-fill by role tier — how long does each type of search take?
Segmented by role complexity. Useful for: identifying process friction (elongating searches without improving quality), setting realistic expectations with hiring managers, and calculating vacancy cost accurately. Not useful as a universal target.
Strong-candidate conversion rate — what proportion of strong candidates in the pipeline convert to hires?
Separate from total pipeline conversion. Measures: how well the process retains the candidates worth retaining. A declining strong-candidate conversion rate signals process friction, slow decisions, or offer competitiveness issues. A stable or improving rate signals the process is working for the right candidates.
Offer acceptance rate — what proportion of extended offers are accepted?
A declining offer acceptance rate signals: misaligned compensation, slow offer process, or a candidate experience that erodes confidence during the search. Investigate declines by asking candidates who declined why they did so. The data is almost always available and almost never collected.
Source of quality hire — which sourcing channels produce strong performers?
Track quality-of-hire by source — not just hire by source. The sourcing channel that produces the most hires is not necessarily the one that produces the best hires. Allocate sourcing investment to channels that produce quality, not volume.
Early attrition rate — what proportion of hires leave within 12 months?
Early attrition is a lagging indicator of hiring quality and onboarding quality simultaneously. Track it separately from voluntary and involuntary. A high voluntary early attrition rate signals a role reality mismatch — the hire expected something different from what they found. A high involuntary early attrition rate signals an evaluation failure.
Common mistakes
01
Tracking cost-per-hire as a primary efficiency metric
Cost-per-hire optimised in isolation produces cheap searches that produce poor hires. It is only useful as a metric when paired with quality-of-hire — which reveals whether the cost was well-spent.
02
Dashboard-first analytics
Building a hiring dashboard before deciding which decisions it should inform is common and produces expensive data infrastructure with low decision impact. Define the decisions first. Then build only the data capability needed to inform them.
03
Measuring activity rather than outcome
Number of interviews conducted, number of sourcing messages sent, number of applications received — these are activity metrics. They measure effort, not effectiveness. Outcome metrics (quality of hire, conversion rate, offer acceptance) are more useful because they measure whether the effort is working.
04
No metric ownership
Metrics that nobody owns are metrics that nobody acts on. Assign each metric to a specific person who is responsible for monitoring it, interpreting changes, and recommending actions when it moves in the wrong direction.
Example scenario
A 70-person B2B SaaS company with a hiring dashboard tracking 22 metrics. Quarterly hiring review: 90 minutes reviewing data, 15 minutes discussing actions. Leadership felt informed but couldn't articulate what the data was telling them to do differently.
The metric audit
22 metrics reviewed: 14 classified as activity metrics (measuring effort). 5 classified as outcome metrics (measuring results). 3 classified as vanity metrics (no decision linkage identified).
Of the 5 outcome metrics: 3 had not been reviewed in the previous quarter because the data wasn't up to date. 1 had declined significantly with no action taken.
The one declining metric: strong-candidate conversion rate, down from 74% to 51% over 6 months.
The redesign
Dashboard reduced to 6 metrics (per the framework above). 16 metrics discontinued.
Each remaining metric assigned an owner and a review cadence.
Strong-candidate conversion rate investigated: identified offer speed as primary cause (average 18 days from decision to offer; candidates accepting alternatives in the interim).
The outcome
Offer preparation moved to parallel with final round. Average decision-to-offer time: 3 days. Strong-candidate conversion rate recovered to 78% within 2 months. Quarterly hiring review reduced to 35 minutes — all decision-driven.
Track less. Act on what you track.
Six outcome-oriented metrics, consistently tracked and consistently acted on, produce more hiring improvement than 22 activity metrics reviewed quarterly. The question for every metric: does this change a decision? If not, stop tracking it.
Continue reading
What Good Hiring Looks Like →