Published November 8, 2026.
A source-of-hire review usually counts where applicants came from. A useful review goes further and asks which sources produced hires who stayed, performed, and were fairly treated. That question changes the channel mix and the conversation with the client. It also keeps the review from rewarding a channel that produces volume without a hire.
Define the source before the data is pulled
Decide what counts as a source: the channel that first brought the candidate, the channel that produced the application, or the channel that influenced the decision. Most systems record the application source, which is not always the first touch. Write the definition and keep it stable across the review period. A change in definition can look like a change in channel performance.
Record the source at the application stage, not after the hire. A recruiter who fills in the source at the end may choose the channel that looks best. If the system allows a candidate to name a referral and a job board, record both and choose one as primary. The secondary source is still useful for context. Include the agency, referral, job board, careers page, event, and direct application categories. Add an unknown category rather than forcing every record into a channel. A large unknown share is a data-quality finding, not a reason to guess.
Measure quality, not only volume
For each source, report the number of applicants, the number interviewed, the number hired, and the number still employed after ninety days. Add the share that passed the probation review if the business has one. A source that produces many applicants but few hires is expensive even when the cost per click is low. Compare the sources on the same role family and location. A channel may work for warehouse roles and fail for software roles. A single overall rate hides that difference. Segment the review before drawing a conclusion. Ask the hiring managers which hires from each source needed extra support in the first month. That question adds a quality signal that the applicant tracking system does not capture. Record the answer as a short note, not a score.
Include cost and time in the comparison
Report the cost per hire by source, including the agency cost, the job board spend, the event cost, and the recruiter time. A referral may look free until the bonus and the coordinator time are counted. A clear cost comparison helps the client choose where to invest. Report the time to fill by source as well. A source that produces a hire in three weeks may be worth more than a cheaper source that takes eight. Do not compare the averages without the role family and the seniority level. A senior role will always take longer. Keep the cost data in the same period as the hire data. A hire that started in one quarter and was paid in the next can distort the comparison. Note the period and the currency.
Check fairness across candidate groups
Review the source mix by the demographic groups the business tracks, where the data is available and lawful. A channel that produces diverse applicants but few hires may have a screening problem rather than a sourcing problem. Do not publish small-cell results that could identify a candidate. Aggregate the data before sharing. Ask whether the application process creates an unnecessary barrier for a particular group. A long form, a specific device requirement, or an interview time that excludes a shift worker can reduce the pool. The source review is a chance to fix the path, not only to change the advertising. Keep the legal and privacy review with the HR owner. The recruiting team should not collect demographic data that the business has not approved. Use the categories the business already reports and no more.
Give the client a short, honest summary
The summary should show the top sources by hires, the cost per hire, the ninety-day retention, and one recommendation. Put the caveats beside the numbers: the sample size, the period, and the unknown share. A client is more likely to act on a small, honest finding than on a large, confident one that cannot be explained. Avoid ranking a source on a single metric. A source can be strong on volume and weak on retention. Show the trade-off and let the client choose the priority. The recruiting team should recommend, not decide, the channel strategy. Record the decision and the date of the next review. If the client changes the budget, the next review should show the effect. A source-of-hire review that is never repeated cannot tell the business whether the change worked.
Close the loop with the candidates
Ask a sample of hires and rejected candidates how they found the role and whether the process was clear. The answer may reveal a source the system never captured, such as a community group or a former employee. Add the category if it recurs. Do not contact candidates who asked not to be contacted. Keep the feedback short and use the approved channel. The purpose is to improve the process, not to build a marketing list. Share the themes with the recruiting team and the client. A source review that stays in a spreadsheet changes nothing. The team should see which channel to invest in and which candidate experience to fix.
Keep the review repeatable by saving the query and the definitions with the report. A new analyst should be able to rerun the same view next quarter without rebuilding it. Note the date the report was produced and the person who checked the numbers. A repeatable query is more valuable than a one-time slide.