Define the question first
Identify whether you need to understand slow feedback, stalled applications or channel relevance. A chart without a decision attached often becomes decoration.
Document the start event, end event, population, exclusions and review frequency. Prefer a few understandable measures to many ratios with shifting definitions.
Separate two recruitment timelines
One working definition measures vacancy opening to offer acceptance; another measures application to acceptance. Document your own conventions consistently.
If a vacancy opens on 1 March, a person applies on 15 March and accepts on 31 March, the date differences are 30 and 16 days. Neither is automatically the time until employment begins. Explain whether paused periods are included.
Seven measures to discuss
These are suggested definitions, not a claim that every measure is automatically available in aiKip or another ATS.
| Measure | Calculation | Useful question |
|---|---|---|
| Applications | Count within the defined scope. | Is the volume relevant? |
| Interview progression | Applications reaching interview ÷ applications in the same cohort. | Does targeting match the role? |
| Offer acceptance | Accepted offers ÷ offers with a final response. | Are expectations aligned? |
| Time to fill | Acceptance date − opening date. | Where is time spent? |
| Applicant journey | Acceptance date − application date. | How long does the person wait? |
| Response waiting time | Time since an action requiring a reply. | Which commitments need action? |
| Results by source | Volume and progression by known source. | Which channels deserve investigation? |
Use coherent cohorts
A cohort groups applications entering in the same period or recruitment. Dividing September interviews by September applications can mix older applicants with people who applied yesterday.
For a completed fictional process with 100 applications, 20 interviews, 5 offers and 4 acceptances, the ratios are 20%, 25% and 80% respectively. The four acceptances are 4% of initial applications. If records remain under review, final-looking rates are premature.
-
100 applications
The fictional starting population.
-
20 interviews
20 ÷ 100 = 20%.
-
5 offers
5 ÷ 20 = 25%.
-
4 acceptances
4 ÷ 5 = 80%.
Completed fictional cohort: 100 → 20 → 5 → 4. Bars compare counts on a common 0–100 scale. Stage conversions are 20%, 25% and 80%; the overall conversion is 4%.
View chart data
| Stage | Count |
|---|---|
| Applications received | 100 |
| Interviewed | 20 |
| Offers | 5 |
| Acceptances | 4 |
Read delay distributions
For durations of 10, 12, 15, 18 and 60 days, the median is 15 and the mean is 23. Both are valid but describe different aspects.
Include sample size and vacancy context. Animation cannot make different roles or periods comparable; definitions and data quality remain essential.
In this small example, the 60-day case represents 52% of the combined 115 days. Removing it would lower the mean to 13.75 days, but change the population. Do not discard a value merely because it is inconvenient; check for errors, pauses or a genuinely longer search.
Fictional durations sorted in ascending order. The dashed marker shows the mean. These five hires are not a benchmark.
Mean : 23 days
View chart data
| Hire | Duration |
|---|---|
| Hire 1 | 10 |
| Hire 2 | 12 |
| Hire 3 | 15 |
| Hire 4 | 18 |
| Hire 5 | 60 |
Assess channels beyond volume
A high-volume channel may produce fewer relevant applications than a smaller one. Examine progression while recognizing small samples.
Explain attribution: a link or declared source simplifies a journey that may have several touchpoints. Never use source as a shortcut for an individual applicant’s suitability.
Share reports with appropriate access
Leaders may need recruitment totals without every CV; managers may need only their vacancies. Share information proportionate to the decision.
GDPR minimization and security requirements still matter. Small-group aggregate figures can sometimes identify people. Avoid ranking recruiters on timelines without considering their assigned roles and constraints.
Turn reporting into action in aiKip
Reliable reports depend on current application records. Check whether decisions made elsewhere are reflected in the ATS.
Choose one change, such as clarifying a stage or improving feedback timing, and assess it with stable definitions. Review aiKip’s Reports page for actual capabilities rather than assuming every formula here is automatically implemented.
Frequently asked questions
What is a good recruitment timeline?
There is no universal figure. Define the measure, compare similar roles and examine the steps explaining the delay.
Can open applications be included?
Yes, but disclose that the result is provisional and explain the observation period.
Are these aiKip customer results?
No. Every numerical example is fictional and illustrates calculation and interpretation.