Lever

Lever exposes a wide set of objects: opportunities, postings, applications, offers, interviews, feedback, users, and stages. This integration focuses on two objects — opportunities and offers — which together span the full recruiting lifecycle from initial candidacy to accepted hire. Postings, interviews, and feedback forms were not included in this first version: their data is either too distal from individual recruiter action or too dependent on organizational schema consistency to produce reliable per-user indicators. Five KPIs were retained, selected against three criteria: ability to attribute to a named owner, resistance to gaming, and balance between leading and lagging indicators.

Hiring outcomes: reading the funnel from pipeline activity to confirmed hire

Candidates Hired counts the number of opportunities archived with a hired status in the period. It is the primary outcome of the recruiting function and carries minimal gaming risk: a hire involves offer signature, background check, and start date confirmation, all of which are externally verified. Offers Extended counts the number of offers formally sent to candidates. It functions as a leading proxy for Candidates Hired, providing visibility on outcomes two to four weeks before a hire is confirmed. When read together, the two indicators reveal pipeline throughput: a sustained gap between Offers Extended and Candidates Hired signals acceptance problems before they are visible in the hired count alone.

Offer Acceptance Rate measures the proportion of sent offers that are ultimately signed, attributed to the hiring manager rather than the recruiter. The distinction is deliberate: compensation structure, role clarity, and stakeholder alignment during the interview process are the hiring manager's domain. A low acceptance rate that co-occurs with a high Offers Extended count indicates a structurally mis-calibrated offer process, not a volume problem. This pairing prevents the false conclusion that extending more offers will improve outcomes when the underlying issue is offer quality.

Recruiter activity and efficiency: the daily work made measurable

Candidates Advanced counts the number of stage progression events recorded on opportunities within the period. It is the only leading indicator in this integration and represents the recruiter's core daily contribution: moving candidates through the pipeline toward an outcome. This KPI carries a medium gaming risk — a recruiter under pressure can advance candidates who are not genuinely qualified — which is why it is systematically read alongside Candidates Hired and Offer Acceptance Rate. If stage advancement volume is high but downstream indicators stagnate, the quality of advances is insufficient.

Median Time to Hire measures the median number of days between opportunity creation and hire archival, attributed to the recruiter who owns the opportunity. The median is used instead of the average because hiring duration distributions are right-skewed: a small number of unusually long processes would distort an average and mask the typical experience. This indicator is in tension with Offer Acceptance Rate: compressing hiring timelines to improve time to hire can degrade acceptance rates if candidates are rushed through evaluation stages before alignment on compensation and expectations is secured.

Scope and limits of the integration

Lever records status transitions and ownership assignments, but does not measure the quality of hiring decisions. A candidate hired in fourteen days is not a better hire than one hired in forty-five days; the API data does not track post-hire performance, retention, or manager satisfaction. The Candidates Advanced indicator is subject to the quality of stage configuration: teams that define stages inconsistently across job postings produce advancement counts that are not comparable across recruiters or time periods.

Additionally, the integration relies on the opportunity owner field being actively maintained. When recruiters reassign ownership mid-process or operate under a pooled team model without assigned owners, attribution breaks down and per-recruiter KPIs lose interpretive value. The reliability of these indicators depends directly on the team's discipline in maintaining accurate ownership records throughout the recruiting lifecycle.