Pave Data Puts Sales at 18.4% of Staff at AI Hypergrowth Companies

AI hypergrowth companies run smaller sales teams, and sales still makes up close to a fifth of their staff.

Jason Lemkin published Pave data at SaaStr comparing headcount by function at AI hypergrowth companies with the rest of tech. Sales is 18.4% of employees at the AI companies and 20.7% elsewhere in tech. Outside tech, the figure is 10.4%.

Pave said in August that "sales headcount proportions are broadly similar across both groups" and that customer support is underweighted at AI companies. That comparison set top AI firms against all of tech. In the November figures, support and the catch-all "other" group run lean at AI companies, while sales holds its proportion of the company. The smaller go-to-market teams in ICONIQ's survey likely come from smaller companies whose reps each carry more.

Support is where the headcount thins

Pave's figures, as SaaStr published them, are medians for each function, so neither column adds to 100%:

• Sales: 18.4% at AI companies, 20.7% across the rest of tech
• Support: 1.8% against 4.4%
• Engineering: 47% against 33.9%
• The "other" group: 20.8% against 28.8%
• HR: 5.6% against 3.2%
• Legal: 1.1% against 0.5%

Support at the AI companies runs at well under half the rest of tech's level. The sales gap is about two points. Lemkin reads the support figure as the first place agents remove headcount.

The back office isn't lean across the board. HR and legal take a bigger slice at the AI companies. Companies growing 100% or more raised median headcount 133% in the first half of 2026, per SaaStr, and hiring at that pace likely keeps HR staffed up. Pave's August post, which averages each company's mix across 162 AI firms with 50 or more employees, also put engineering at nearly half of staff.

Smaller companies explain the smaller teams

ICONIQ's survey of about 150 GTM executives, run by the growth investor in January 2026 and cited by SaaStr, counts a median of 20 GTM employees at AI companies with $10 million to $25 million in ARR. Other companies at that size have 35. That 43% gap narrows in each larger revenue band, down to 21%.

"They have fewer reps because they have fewer employees overall," Lemkin wrote. The narrowing gap fits that reading as the companies grow.

The survey also points to more output per rep. Companies that have embedded AI more fully report 67% of reps at quota, against 59% at the rest. Sales-sourced pipeline is 62% of the total in the same survey.

The mix inside sales is still moving

Pave's 2025 analysis found SDR-titled roles fell from 1.98% of its dataset in January 2023 to 1.45% in August 2025. Over the same period, the share of companies employing a GTM engineer rose from 0.02% to 0.51%. A steady sales total can hold a changing set of jobs.

Pave says its AI cohort skews earlier-stage than the all-tech group, and stage alone shapes headcount mix. SaaStr's post also places forward-deployed engineers, who do solutions and implementation work, inside engineering's 47%. Some customer-facing staffing likely sits outside the sales column.

The earlier-stage skew could account for part of both gaps. Pave's two methods still agree on the main result: sales near parity with the rest of tech, and support underweighted.

Sales stays within two points of the rest of tech

The two-point gap in sales is small next to the 13-point gap in engineering. The bigger change is inside sales, where SDR roles have shrunk and GTM engineers have appeared.

They have fewer reps because they have fewer employees overall.

The question the smaller sales teams raise is whether company size explains them. ICONIQ's GTM headcount gap, 43% at $10 million to $25 million in ARR, narrows to 21% as revenue bands grow.

By Business of GTM