Hiring Faster Without Hiring Better: The Speed-Volume Trade-Off
When a Fortune 500 insurance carrier more than doubled its hiring, its production rate fell from 41.5% to 21.1%. On its own that looks like a failure. Connecting the systems told a fuller story. The carrier ended up with a comparable number of producers, each ramping faster, 62 days versus 109, and generating more per month, about $812 versus $395. The trade was a lower share of hires producing, in exchange for faster and more productive ones. You cannot see that trade-off until screening data is joined to outcomes.
Source: "Decision Traces," Saad Bin Shafiq, NODES, 2026, comparing the 2022 and 2025 cohorts. Read it on arXiv.
The surface story
The production rate, the share of hires who reach the production milestone, fell from 41.5% in 2022 to 21.1% in 2025. Over the same period, hiring volume more than doubled, from 1,486 to 3,623 agents, and the dominant job board grew from 44% to 67% of hires. Looked at alone, a falling production rate during a hiring surge reads as a quality problem.
The fuller story
Connecting the systems changed the picture. Despite the lower rate, the higher volume produced a comparable number of producers, 680 in 2025 against 616 in 2022. Each producer ramped faster, a median of 62 days to the production milestone versus 109, and generated more per month of tenure, about $812 versus a pooled $395 in the earlier cohorts. Using the speed constant, the faster ramp alone accounts for roughly $1.12M in additional annual production across 680 producers. See the speed-to-production findings.
Why it stays hidden
A carrier looking only at the production rate would see decline and miss the compensating gain in speed and per-producer output. The trade-off is only visible when screening inputs are connected to production outcomes across systems. Without that connection, you manage to the wrong number. See how the connection works.
What this does not say
The paper is candid about the limits, and so is this page. The causes of the rate decline cannot be fully separated: the carrier went deeper into the talent pool, the channel mix changed, market conditions may have shifted, and the scoring system may itself have selected for speed over probability of producing. The per-producer comparison carries offsetting biases, since the earlier cohort's production data favors long-tenured survivors and tenure normalization mechanically favors the newer cohort. This is a quasi-experimental observation with a historical control, not a clean causal result. The takeaway is the trade-off itself, and the fact that you need connected data to see it.
Frequently asked questions
Does hiring at higher volume lower quality? Not necessarily in the way a single metric suggests. In this study the production rate fell as volume rose, but producers ramped faster and generated more per month, so the picture depended on which number you looked at.
What is the speed-volume trade-off? It is the pattern where a process produces a lower share of producers but faster and more productive ones, so total production can hold up even as the production rate falls.
Why did the production rate fall? Several causes that cannot be fully separated: a deeper talent pool, a changed source-channel mix, possible market shifts, and a scoring system that may have favored speed.
How do you measure this correctly? By connecting screening data to production outcomes, so you can see ramp speed and per-producer output alongside the headline production rate.
Related reading
See the real trade-offs hiding in your own hiring data. Book a 30-minute walkthrough.