AI-Era TA, part 2 of 3
Where most growth-stage TA scorecards sit today
If you’re running TA at a company between 80 and 200 people, your scorecard probably looks like time-to-fill, offer acceptance rate, cost-per-hire, and a gut sense of whether you’re hiring well. Most TA reporting at this stage hasn’t evolved past what worked when the company was 30 people, even though the operating reality has changed several times since then.
That’s understandable given compounding pressure. At 30 people, hiring runs on founder instinct. By 200 people, founder instinct stops scaling, and the absence of structured measurement starts costing real money. The scorecard problem catches up to companies right around the doubling point.
The first post in this series argued that TA becomes the moat in a compressed org. This post is about the version of that scorecard that fits a company hiring 50 to 100 people in the next 12 months on a lean TA team, in an environment where AI does most of the funnel work.
What old metrics still tell you, and what they don’t
Time-to-fill, cost-per-hire, offer acceptance rate, recruiter productivity. These metrics still measure something real. They’re useful when you’re explaining hiring spend to your CFO, or trying to figure out why three engineering reqs have been open for 90 days.
What they don’t tell you, and never have, is whether the people the function is producing deliver on the job. At a company hiring 30 a year, you can run on the old metrics because a few weak hires get absorbed. At a company hiring 100 a year on a lean TA team in a compressed market, you can’t. A 5% bad-hire rate at 30 hires is one or two people you can manage out. A 5% bad-hire rate at 100 hires is five people, which translates to real dilution of talent density at exactly the moment you need to be sharpening it.
The reframe underneath all of this: speed and quality are independent axes, and what links them in practice is signal quality in the funnel. Old metrics measure process. New metrics measure whether the signal is clean enough to let velocity and bar move together. That distinction is the whole game.
Gartner’s 2026 TA trends report found that 45% of recruiting leaders face heightened pressure to improve quality of hire, a sharp jump over prior years. LinkedIn’s Future of Recruiting research found that 89% of TA professionals agree quality of hire will become more important to measure, and only 25% feel confident their organization can measure it well today. That’s a 64-point gap between knowing and doing.
At growth-stage companies, that gap is the opportunity. Most companies above 200 people are starting to fix this. Most below 50 don’t have the volume to need it. The 50-to-200 range is where the lift is highest and the competition is lowest. A growth-stage TA leader who closes that gap walks into the next conversation with a CEO holding evidence the rest of the function couldn’t produce.
The five metrics that earn their place at this stage
A 100-to-200 person company needs a five-metric scorecard. The 7-metric enterprise version is too much for this stage and obscures the signal it’s meant to produce.
1. Quality of Hire (composite at 12 months). The headline metric. At growth-stage scale, the simplest version works: Performance Rating + Retention + Hiring Manager Satisfaction, scored on a 100-point composite. Drop time-to-productivity from the standard formula at this stage because most growth-stage companies don’t yet measure ramp time consistently. Layer it back in as the function matures.
SHRM’s 2025 benchmarking research found that high-performing TA functions hold a Quality of Hire score consistently above 80, and organizations measuring it for the first time typically score between 58 and 67. Your starting score will look ugly. That’s the point. Visibility forces investment.
2. 90-day performance attainment rate. The leading indicator. Quality of Hire is lagging by 12 months, which is too long for a company doubling in a year. The 90-day metric (the percentage of new hires hitting the performance expectations set during the hiring process at the 90-day mark) gives you a directional read in three months instead of twelve. If 90-day attainment drops, your QoH outcome a year out is in trouble before the lagging data confirms it.
3. Hiring manager satisfaction at 6 months. The qualitative complement. Asking the hiring manager whether they’d hire the same person again at 6 months buys you another measurement cycle in the same calendar year compared to the 12-month version. Single question on a simple scale. Easy to instrument in any modern HRIS.
4. Candidate authenticity / verification pass rate. The risk metric. AI-driven candidate fraud has gone from edge case to top-of-mind in 24 months, and it sits inside a broader top-of-funnel composition problem: candidates who look strong enough to advance but don’t convert later, now compounded by AI-driven misrepresentation. Gartner found that 59% of hiring managers suspect candidates of using AI to misrepresent themselves, and 62% of recruiters say candidates are getting better at faking faster than detection systems are catching up. Checkr reported that 23% of companies have already encountered identity fraud among new hires. GoodTime’s 2026 Hiring Insights Report found that fraudulent candidates are now the #1 anticipated TA challenge for 2026, surpassing talent shortage. At a company hiring 100 a year, even a 2% fraud rate is two people who shouldn’t be there.
5. Time-to-fill. Kept, demoted. Useful as a hygiene number to watch for trend. A jump in time-to-fill alongside flat or improving Quality of Hire is fine. It might mean you’re being more selective, which is the right move at this stage. A drop in time-to-fill alongside dropping QoH is the failure mode the rest of the scorecard is built to catch. Read together with the outcome and risk metrics, time-to-fill stops being a tradeoff against quality and starts showing whether velocity and bar are moving together.
Five metrics. Outcomes lead, risk gets its own line, hygiene runs underneath, and operational metrics like recruiter productivity stay internal to the function.
What it takes to build this
The scorecard is the easier half. The harder half is making the data work.
The good news for growth-stage companies is that the modern stack supports this. AI-native ATSs like Ashby ship the QoH frameworks and native analytics that didn’t exist five years ago, with Greenhouse continuing to add capability on the same vector. The adjacent AI tooling layer matters just as much: sourcing platforms like Gem and Juicebox generate pipeline signal that feeds the leading-indicator metrics, and interview intelligence tools like Metaview capture structured data from every conversation that ties back to post-hire performance. Modern HRISs like Rippling pipe the performance data back through. The tooling has caught up to the question. The discipline of using it is what’s still missing.
In practice, building the scorecard at growth stage involves four moves:
- Map the data sources. What lives in the ATS, what lives in the HRIS, what lives in the performance system, and which of those connect to each other today. At a 100-to-200 person company, this is usually a one-week audit rather than a one-quarter project.
- Pick one metric to start. 90-day performance attainment is the easiest first metric because it requires the fewest data sources (just the ATS req data and a hiring-manager survey at the 90-day mark). It also produces actionable signal fastest.
- Stand up the survey and feedback loop. This is where most growth-stage companies stall. Without a clean cadence for hiring managers to fill out the surveys, the data never accumulates. A short, recurring HRIS-triggered survey solves it.
- Build the executive view. Dashboard or one-pager. Doesn’t matter. What matters is that a CEO can read it in 30 seconds and ask the right next question.
The Year 1 investment runs roughly $5K to $30K in tooling depending on what’s already deployed, plus a meaningful slice of TA leadership time in the first 90 days. The investment is small. The discipline to execute against it is the harder ask.
A 30/60/90 view
For a Head of TA walking into a 100-to-200 person company today, the buildout maps cleanly to a quarter:
- Days 1 to 30: audit current state. What’s tracked, what’s not, what data is available, what tooling is already deployed (ATS, HRIS, performance system), where the gaps are between the three. The audit alone usually surfaces three or four insights leadership didn’t have.
- Days 31 to 60: ship the first version of the scorecard. Speed matters more than polish here, because the goal is to get the executive cadence reviewing the right numbers as soon as possible. The mature methodology gets layered in over the next two quarters as the data and the tooling catch up.
- Days 61 to 90: standardize one weak metric (probably candidate authenticity given the fraud pressure), and start the 12-month QoH baseline so meaningful data lands by mid-2027.
By day 90 you have an executive scorecard, a leading indicator running monthly, a forward path on the lagging metric, and the foundation for the harder build that comes next.
The two-line version for the CEO
The right way to surface this scorecard at a growth-stage company is short. Two lines, three numbers:
“Our 90-day performance attainment is X. Our candidate authenticity rate is Y. Our Quality of Hire baseline starts measuring this quarter and will produce its first full read in Q2 next year.”
The rest of the metrics live below the fold. Time-to-fill stays internal to the TA function. Hiring manager satisfaction surfaces inside the QoH composite. The CEO has three numbers to track, and those three numbers map directly to the question they care about most: whether the company is hiring well enough, fast enough, to support the business plan.
That sentence, said to a CEO at a growth-stage AI-native company, distinguishes a TA leader who is operating in 2026 from one still operating on the metrics from 2016.
What this gets you
Most growth-stage companies don’t have this scorecard. They have time-to-fill, a recruiter or two running open reqs, a CEO who can’t tell whether the function is working well, and a CFO who only sees the cost. The TA leader who walks in and ships this scorecard in their first 90 days is doing the work most TA leaders won’t, because it requires building data pipelines, training hiring managers on what to fill out, convincing the CFO that quality-weighted thinking is more useful than raw process metrics, and getting the executive cadence to look at the numbers month over month.
The payoff is that the next compression event, when AI takes another 20% of operational hiring work and the team gets pressure to absorb it, the TA leader with the scorecard is the one with the data the CEO trusts. Every other function gets compressed. The function with the receipts gets a seat at the strategy table.
Post 3 in this series gets concrete on the other side of the equation: the hiring signals that predict performance against this scorecard, building on the work of Adriano Herdman at Move and what AI-mature companies have learned over the past two years.
The scorecard is the lens. The signals are what you look for through it.