The HR function has been trying to earn a seat at the revenue table for decades. AI is the most compelling business case the function has ever had — but most HR leaders are presenting it wrong, with the wrong metrics, to the wrong audience. Here's how to build a business case that gets funded.
---
Start with what the data actually shows, because the numbers are strong when you present them correctly.
McKinsey's 2025 analysis puts average ROI on AI investment at 5.8 times within 14 months — and that's across all enterprise functions, not just HR. For recruiting specifically, the efficiency gains are documented and substantial. HireVue clients report 60 to 89% reductions in time-to-hire. Paradox's deployment at GM saved $2 million annually in recruiter time and process overhead. Chipotle cut hiring time by 75%. These aren't projections — they're results from named clients with repeatable use cases.
The broader productivity data is also significant. AI-augmented roles show 37% average productivity gains. Companies deploying AI-driven process automation save an average of $4.6 million annually at the enterprise level. The return on generative AI specifically runs $3.70 for every $1 invested.
That's the headline case. But there are two important nuances that make the difference between a credible business case and a CFO's eye-roll.
The first nuance: not all AI projects produce positive ROI, and the timeline matters. Forrester's research shows that 44% of AI projects that make it to production achieve positive ROI within 12 months. That's a majority — but it's not 100%, and the implication is that the design of your implementation significantly affects whether you land in the 44% or the 56%. The projects that succeed have clear use case definitions, measurable success metrics established before deployment, and a change management plan that drives actual adoption. The projects that don't have those things fail not because the technology doesn't work but because the humans didn't use it effectively or the baseline metrics were never established.
The second nuance: the CFO isn't your only audience, and different audiences need different numbers.
For the CFO, the argument is cost per hire and time-to-hire reduction. The average cost-per-hire in the US runs $4,700 (SHRM benchmark), and that number scales up dramatically for specialized roles. If AI screening reduces the screening time component of that cost by 75%, or if it compresses time-to-hire by 60%, those are dollar figures you can calculate against your current volume. Run your organization's hiring numbers. If you make 1,000 hires per year and AI reduces time-to-hire by 30 days, and you have roles where delayed hiring creates lost revenue or productivity, you can model that value explicitly.
For the CHRO, the argument is quality and compliance. The business case isn't just efficiency — it's better hiring decisions, more consistent evaluation, and a reduced compliance exposure. AI-led interview candidates succeeded in subsequent human interviews at 53% compared to 28% for resume-screened candidates, per WEF research. That quality improvement is a legitimate business case component that the CFO argument often misses.
For the CEO, the argument is talent as competitive advantage. Unilever's AI-assisted hiring program improved 16% in retention of new hires. IBM's predictive attrition AI achieved 95% accuracy in identifying employees likely to leave, contributing to $300 million in saved turnover costs. SAP saw attrition rates fall 20% using predictive algorithms. When you can demonstrate that AI hiring tools improve not just cost-per-hire but 90-day retention and first-year performance, you've moved the conversation from cost reduction to revenue generation.
Building the actual business case requires five elements. One: baseline data. Know your current time-to-hire, cost-per-hire, offer acceptance rate, 90-day retention, and first-year performance ratings. Without a baseline, you can't demonstrate improvement. Two: use case specificity. Don't pitch "AI for HR broadly." Pitch "AI screening for high-volume hourly roles where we make 500 hires per quarter" — a specific use case with specific volume and specific measurable outcomes. Three: vendor-specific outcome data. Every serious AI HR vendor has case studies. Use the ones closest to your company profile — similar industry, similar volume, similar role types. Four: implementation cost realism. Include integration costs, training time, change management overhead, and compliance program development. The total cost of ownership is higher than the software license. Five: risk-adjusted returns. The 44% within-12-months positive ROI finding means you should build your case on conservative assumptions and use best-case scenarios as upside. CFOs who see a realistic range are more confident than those who see only an optimistic single number.
One more thing on timing. The organizations seeing the best AI hiring ROI right now are the ones that started building capability 18 to 24 months ago. The window to be a first-mover has passed for most companies. But the window to avoid being materially behind your competitors is still open — and it closes faster than most HR leaders expect.
---
Quick Hits
Time-to-Hire Reductions of 60-89%
HireVue's data showing 60-89% time-to-hire reductions represents the high end of the range, but it's achievable for high-volume use cases with strong implementation. Even a 30-40% reduction translates to meaningful business impact in industries where slow hiring creates vacancy costs — retail, healthcare, logistics. Translate the time reduction into days, then model what a vacant role costs per day in your context.
Unilever's 16% Retention Improvement
Unilever's AI hiring program produced a 16% improvement in new hire retention — a metric that rarely shows up in AI ROI calculations but is often the largest single lever for reducing total talent costs. Turnover is expensive: SHRM estimates total replacement cost at 50-200% of annual salary depending on role seniority. If AI improves first-year retention by even 10 percentage points, the financial case usually dwarfs the screening efficiency savings.
Cost Per Hire Benchmarks
SHRM's benchmark puts average cost-per-hire at $4,700 for hourly roles, with significant variation by industry and role type. Specialized technical roles run 5-10 times that. When building an AI ROI case, start with your actual current cost-per-hire data, not industry averages — your CFO will ask, and your credibility rests on the specificity of your baseline.
---
The Operator's Take
HR has a long history of underselling its business impact, and the AI opportunity is too large to repeat that mistake. The executives who are getting AI budgets approved aren't presenting AI as a cost center efficiency play — they're presenting it as a capability that directly affects revenue, retention, and competitive positioning. The data supports that framing. A 5.8x ROI within 14 months is a venture-grade return in an operating budget. A 16% retention improvement for new hires moves the needle on the talent stability that enables every other business objective. Build your case at that level and you'll get funded. Build it as a recruiter time-saving project and you'll compete with every other process automation request in the queue.
---
Building a business case is step one. Building the operational playbook that actually delivers the ROI is step two — and where most AI HR investments fall apart. A clear adoption framework, vendor selection process, and implementation roadmap is what separates the 44% that achieve positive ROI within 12 months from the 56% that don't.
Get it here → AI Adoption Playbook for HR Teams ($39)
---
The Screening Room — Editions 19–24