Note: This is part 2 in our story of how Johan and Alice used AI, “Moneyball!” and data analytics to rebuild their most costly sourcing channel. We asked Claude AI to validate the figures to determine if they were reasonable. They are. Here’s its full analysis.


After uncovering the staggering $200,000 cost of each failed job posting hire, Alice O’Hara and her analyst Johan Evans dug deeper into their division’s recruiting channels. Ethan shared this table highlighting the average profit generated per hire in year one from job boards in comparison to the company’s other sourcing channels.

Financial Impact per Hire is a Better Metric than Cost per Hire

Alice thought the indirect costs of job postings seemed high but Johan said he checked and 60% of their total talent acquisition budget was being consumed by job board hiring, yet this channel accounted for only 30% of their actual hires. “When we factor in the technology costs, recruiter time, hiring manager hours, and most importantly, the downstream costs of turnover, job boards are our most expensive channel by far.” He pointed to the analysis showing that internal moves and boomerang hires, in contrast, delivered better results at a fraction of the cost.

Let’s Redesign Our Job Board Talent Strategy

While concerning, Alice wasn’t ready to abandon job postings entirely. “We don’t need to eliminate job boards,” she told Johan. “We need to reinvent how we use them. We need to attract the right people in, not weed out those who are just randomly applying.”

First, they needed to stop the flood of unqualified applications before they even applied, that were driving up overhead costs. Second, and perhaps most crucial, they needed to address the lack of job understanding and role clarity issue that kept emerging in the exit interviews.

Alice had recently attended a DEI conference where a leading labor attorney discussed performance-based job descriptions that could solve both problems.

She excitedly told, Johan, “Here’s the idea. Rather than listing required skills and experience, these descriptions focused on what successful candidates would actually accomplish in the role.”

Job Branding is the Key Differentiator

“Think about it,” Alice explained. “We’re seeing high turnover because people don’t fully understand the role before they start the job. What if we described jobs in terms of expected outcomes rather than required credentials? As part of this let’s tie the role to an important company project. That’s called job branding and studies have shown that it’s critical for companies that don’t have a big employer brand.”

“Here’s a sample of what this posting for a product manager role would look like. This one was created by our AI agent in less than a minute. Then as part of our application process we could then ask candidates to submit an accomplishment most comparable to the major performance objective. This would be a better way to screen candidates. Just as important it would attract stronger candidates and exclude those who weren’t as qualified.”

This Job is Mine! – It’s Describing Me

Johan learned from the “Moneyball for HR!” course he just took on LinkedIn Learning that a chi-square test would be perfect for comparing this type of posting to their tradition approach using an A/B test.” He said, “It’s designed for comparing outcomes between two groups. We could track key metrics like candidate quality, interview success rates, offer acceptance, Q12 scores, and first-year retention.”

“We have a unique opportunity here,” Alice told the executive team to get approval for this approach. “By rethinking job postings from the ground up – from how we write them to how we evaluate candidates – we can transform our most expensive hiring channel into one that actually delivers on its promise. Success would be measured not just in reduced turnover, but in improved engagement scores and faster time to productivity.”

“The numbers don’t lie,” Alice concluded. “We’re spending millions on a broken process. But now we have a data-driven path forward. The question isn’t whether we can afford to make these changes – it’s whether we can afford not to.”