Hire twelve engineers by the end of Q3. Fill the sales team before the January kickoff. These are task lists with a deadline rather than goals, and they say nothing about whether the twelve were the right people, how long good candidates waited for an offer, or how many accepted. A team can hit the number and still have had a poor quarter.
OKRs, objectives and key results, were built for the other kind of goal. The objective describes the change you want, in plain words. Two to four key results state what would measurably prove it, each with a starting value and a target, tracked in a weekly check-in across a quarter. If you have only seen the framework applied to product teams, it is worth browsing examples of OKRs written for different teams to see how the same structure bends to different work. Below are three for talent acquisition.
Example one: speed without sloppiness
Objective: candidates we want get an offer before a competitor gets there first. The starting position is a median of 41 days from first interview to offer, long enough to lose people in a tight market. The first key result is to bring that median down from 41 days to 25. The second is to raise the share of interview loops completed within ten working days from 30 per cent to 70 per cent, because the median hides the tail. The third is to lift offer acceptance from 68 per cent to 85 per cent, which keeps speed honest: rushing candidates shortens the clock and lowers acceptance, so the two have to move together.
Example two: quality the business can feel
Objective: hiring managers trust the shortlist we send them. Here the key results lean on leading indicators, the kind the NetSuite guide to KPIs distinguishes from lagging ones. Ninety-day retention is lagging; by the time it moves, the quarter is over. So the first key result is to raise hiring-manager satisfaction with shortlist quality, surveyed after each loop, from 3.4 to 4.3 on a five-point scale. The second is to cut the share of candidates rejected at final stage for reasons that should have surfaced at screening from 24 per cent to 10 per cent. The third is to increase the proportion of new hires who pass their probation review without concerns from 81 per cent to 92 per cent, accepting that this one only partly lands within the quarter.
Example three: the experience and the pipeline it builds
Objective: people who go through our process recommend it, whether or not we hired them. The first key result is to move candidate NPS, measured after final decision and including rejected candidates, from plus 12 to plus 40. The second is to grow the share of hires that come through employee referrals from 15 per cent to 30 per cent, which is both a cost measure and a proxy for how staff talk about the company. The third is to ensure 95 per cent of rejected finalists receive personal feedback within five working days, up from a baseline that turned out to be around 40 per cent.
What makes these work
Each objective could be said out loud in a leadership meeting without a slide. Each key result has a from and a to, so the weekly check-in is about a number moving or not, rather than a status update. And each set contains a counterweight, so that pushing hard on one measure cannot be done by quietly wrecking another. The CIPD's resourcing and talent planning survey is a useful place to sanity-check baselines for time-to-hire, cost and retention across UK employers. Speed without acceptance, or referrals without quality, is the trap a single narrow target invites.
The targets are also ambitious. Reaching 70 per cent of an OKR that aimed at 25 days from 41 is a good quarter. If the team hits every key result comfortably, the goals were probably KPIs in disguise.
Where OKRs do not fit in recruiting
Not all hiring work belongs in an OKR. Steady-state volume hiring, such as keeping a contact centre or warehouse fully staffed with predictable churn, is an operation to be monitored, and a KPI with a threshold does that job better than a quarterly stretch goal. The same goes for compliance measures like right-to-work checks completed on time, where 100 per cent is the only acceptable value and there is nothing to stretch towards. And a team with no say over its own process, because agencies and hiring managers dictate every step, cannot commit to outcomes it does not influence.
The sensible split is to keep the operational measures on a dashboard and reserve OKRs for the two or three things the recruiting team genuinely wants to change this quarter. Write those with numbers, check them weekly, and let the headcount plan be what it is: a plan.
