Why Attribution Models Matter in Recruitment

A staffing agency spending $40,000 per month across job boards, LinkedIn, and employee referrals needs to know which of those channels is generating hires, not just clicks or applications. Without an attribution model, the agency is flying blind: it might be cutting the channel responsible for 60 percent of its conversions because it generates fewer top-of-funnel impressions, and doubling down on a channel that looks productive because it generates traffic but closes nothing.

Attribution modelling in recruitment is less mature than in marketing, but the core problem is identical: a candidate typically encounters an employer through multiple touchpoints before applying. If you only credit the last touchpoint, you underinvest in awareness channels that start the relationship. If you only credit the first touchpoint, you underinvest in channels that close undecided candidates. Choosing the right model shapes where your sourcing budget goes.

How Attribution Models Work

An attribution model is a rule or algorithm that assigns credit for a conversion, in this case a hire or application, across the multiple channels or touchpoints a candidate interacted with before converting. The most common models used in recruitment analytics are:

Last-touch attribution gives 100 percent of the credit to the final channel a candidate used before applying. It is simple to implement and tells you where people convert, but it systematically undervalues awareness-stage channels like employer brand content, display advertising, and social media that candidates encounter early in their consideration process.

First-touch attribution gives 100 percent of the credit to the first channel through which the candidate discovered the role or employer. It highlights what brings new candidates into the funnel but ignores everything that kept them engaged through to conversion.

Linear attribution divides credit equally across all touchpoints a candidate experienced before converting. A candidate who saw a LinkedIn ad, read a Glassdoor review, clicked an Indeed listing, and applied via a referral link would give 25 percent credit to each of those four channels. This is more balanced but treats all touchpoints as equally influential, which may not reflect reality.

Time-decay attribution gives more credit to touchpoints closer to the conversion, on the theory that the channels a candidate engaged with most recently were most influential in the final decision. This tends to favour direct sourcing and apply-path channels over brand and awareness channels.

Position-based or U-shaped attribution gives weighted credit to the first and last touchpoints, with the remainder distributed across middle interactions. A common version gives 40 percent to first touch, 40 percent to last touch, and 20 percent spread across intermediate channels. This acknowledges that both discovery and conversion are important without ignoring the middle of the journey.

A technology staffing agency using last-touch attribution has been cutting its LinkedIn brand content budget for two years because applications credited to LinkedIn Careers are declining. Switching to a linear model reveals that LinkedIn content is appearing as a first or second touchpoint for 38 percent of candidates who ultimately apply through Indeed or directly via the careers site. The agency restores the LinkedIn budget and reduces its Indeed cost-per-click spend, which had been capturing candidates already warmed by LinkedIn rather than sourcing new ones.

Attribution Models vs Source Tracking

Source tracking identifies where a candidate came from. Attribution modelling assigns credit across multiple sources when a candidate experienced more than one. Most ATS platforms offer source tracking but very few offer multi-touch attribution natively, which is why many agencies default to last-touch by default without realising it. Proper attribution requires either a dedicated analytics tool or manual UTM parameter tracking combined with candidate journey mapping.

Attribution Models in Practice

A recruitment operations analyst at a national staffing agency is tasked with optimising a $180,000 annual sourcing budget. Running a position-based attribution analysis across 6 months of hire data reveals that the agency's employee referral program, which was receiving 8 percent of budget, is the first touchpoint for 31 percent of placed candidates. Reallocating $18,000 from underperforming job board spend to referral incentives increases placed candidates attributed to the referral channel by 24 percent in the subsequent two quarters.