AI org benchmark · US equipment rental

In equipment rental, almost nobody is staffing AI, and nobody is talking about it.

Q: What counts as an "AI role," and why combine AI with data science?

Dedicated titles only: core AI/ML roles plus data-science titles, one talent system in practice. The headline number is the combined bench, with the core-AI vs data-science split shown in each exhibit's stacked bars.

Q: Where does the data come from?

Public workforce data resolved to each corporate family, public job postings, the six most recent earnings-call transcripts per company, and company press. All public data, pulled August 2026.

Q: Why is Sunbelt counted at 24,000 employees, and why does EquipmentShare's coverage look odd?

Sunbelt uses its consolidated 24,000 employees (about 21,800 in North America). EquipmentShare's public workforce records exceed its reported headcount, so its absolute count is flagged; its per-10,000 density is the number to read.

Q: Does a bigger bench mean deployed AI, or verified ROI?

No. Headcount measures owned capability, not deployed capability, and every product claim is disclosed by the company or its vendor, not independently verified. No cohort company has disclosed dollar-denominated AI ROI.

The AI Org Diagnostic scores what each company staffs, ships and says across Herc Rentals, United Rentals, Sunbelt Rentals, EquipmentShare, Custom Truck One Source, WillScot, McGrath RentCorp and Alta Equipment Group, measured the same way for every company from public data.

Book a 15-minute call → How it’s measured ↓

8 companies

4 lenses

no survey, no login

See this for your peer set

Want this run on your competitive set?

Pick your company and the peers you’re measured against. We build the same lenses from public data and walk you through them on a read-out call.

Book a 15-minute call →

Executive summary.

Herc Rentals – 2 data-science titles across 9,600 employees, no core-AI title; ProControl described as “utilizing AI and advanced telematics” on the July 2026 call.
United Rentals – 6 data scientists and no core-AI title inside a 2,713-person technology org; AI in pricing came up only when an analyst asked, July 2026.
Sunbelt Rentals – the only company with machine-learning engineering titles (3 of 6), and zero AI mentions across six earnings calls.
EquipmentShare – a 5-person bench including an applied-AI lead; “AI tools we are developing and deploying into the field” on its first public call, August 2026.
WillScot – the only AI executive title in the cohort (Vice President, Artificial Intelligence, Mobility & eCommerce), 10.6 combined roles per 10,000 employees, and the only substantive AI discussion on a call: pricing engine, sales-call coaching, branch video.
Custom Truck, McGrath and Alta – zero AI- or data-science-titled people, and at most one passing AI mention across six calls each.

WillScot and EquipmentShare field the densest benches. The three largest rental companies staff AI at one-fifth the rate.

Pound for pound, WillScot leads: 10.6 combined roles per 10,000 employees. Herc, United and Sunbelt sit at 2.1–2.5. Cohort median 2.1.

Core AI Data science

Company	AI + data-science headcount¹	Employees	Per 10k employees	Share of tech org¹
WillScot	5	4,700	10.6	0.9%
EquipmentShare²	5	7,700	6.5	0.6%
Sunbelt Rentals³	6	24,000	2.5	0.7%
United Rentals	6	28,500	2.1	0.2%
Herc Rentals⁴	2	9,600	2.1	0.2%
Custom Truck One Source	0	2,500	0.0	0.0%
McGrath RentCorp	0	1,306	0.0	0.0%
Alta Equipment Group	0	2,750	0.0	0.0%

¹ Combined core-AI/ML plus data-science titles on a corporate-family basis, counted from public workforce data; observed floors of dedicated titles, not a census, as of August 2026. Per 10k = combined roles ÷ reported employees × 10,000. Share of tech org = combined roles ÷ people in IT and engineering functions in the same data.

² EquipmentShare: public workforce records exceed reported employees; density per 10k is the reliable read, the absolute count less so.

³ Consolidated headcount, formerly Ashtead Group; about 21,800 of the 24,000 are in North America.

⁴ Corporate-family basis: H&E Equipment Services, acquired June 2025, is counted inside Herc.

WillScot is the only company that talks about its own AI. Sunbelt staffs the most machine-learning titles and says nothing.

Across 43 earnings calls, the eight companies made 14 statements about using AI themselves. Talk of the AI build-out driving data-center demand is counted separately and excluded here.

← Scroll for the quiet half of the chart

Own-use mentions = sentences on the six most recent earnings calls per company (one for EquipmentShare) where the company or an analyst discusses the company using AI; sentences about the AI build-out as an end market are excluded. Quadrant lines sit at the cohort medians.

How the AI org is layered.

WillScot is the only company with an executive and a director over its bench. United’s six data scientists report with no AI-titled leader above them; Herc’s two sit inside BI and cyber security.

Exec Director Manager IC bars · share of counted heads by level

Company	Mgr+ / IC¹	Exec+ / Mgr¹
United RentalsIndividual contributors only	0.00	—
Sunbelt RentalsDirector-led	0.20	—
EquipmentShareLead-level, no director	0.25	0.00
WillScotAI executive in place	1.50	2.00
Herc RentalsIndividual contributors only	0.00	—
Custom Truck One SourceNo AI- or data-titled people found	—	—
McGrath RentCorpNo AI- or data-titled people found	—	—
Alta Equipment GroupNo AI- or data-titled people found	—	—

Peer median · 0.20 Mgr+ per IC · 1.00 Exec+ per Mgr

¹ Mgr+ = exec, director, manager & lead heads; Exec+ = exec and director heads. Levels classified from the counted titles; a dash means the ratio has no denominator.

How these numbers are built.

Measure the bench

Combined AI + data-science headcount from public workforce data, resolved to each corporate family before counting: dedicated core-AI/ML and data-science titles only, the same way for all eight companies.

Read the flow and the narrative

Open roles from public job postings screened to dedicated AI and data-science titles, and every AI mention on the six most recent earnings calls, split into the company using AI versus AI as an end market, read against the measured bench.

Count only what we can see

Counts are floors of dedicated AI- and data-titled roles, not a census. Disclosed products, partners and performance claims are recorded as disclosed, never as verified. The same outside-in discipline as our G&A benchmarks.

Sunbelt Rentals is hiring AI or data talent. The other 7 show nothing open.

Open AI- and data-titled roles on public job boards, August 2026. Who’s hiring is read against who’s built.

Core AI Data science

Company		Open AI + data-science roles¹
Sunbelt Rentals		1
United Rentals		0
EquipmentShare		0
WillScot		0
Herc Rentals		0
Custom Truck One Source		0
McGrath RentCorp		0
Alta Equipment Group		0

¹ Public job postings pulled August 2026, matched to the posting company and screened to dedicated AI and data-science titles; postings from other companies that the job board back-fills into a search are excluded.

What this example can, and can’t, tell you.

Free: this example

Eight named companies, measured with one identical method
The combined AI + data-science bench, with the core-AI split and density per 10,000 employees
Open roles and earnings-call narrative read against the measured bench, not taken at face value
Disclosed use cases and partners per company, labeled as disclosed, not verified

The full diagnostic: what we deliberately withhold

The named-profile roster behind every count, with level and location
Your company and your chosen peer set, measured identically
Hiring flow and narrative tracked quarterly as a trend, not a snapshot
Full disclosed-use-case and partner detail per peer
A walkthrough read-out call

Questions & answers

FAQs

What counts as an “AI role,” and why combine AI with data science?+

We count dedicated titles only: core AI/ML roles (machine-learning engineers, MLOps, applied-AI leads, AI executives) plus data-science titles. In practice the two benches are one talent system, staffed and managed together, so the headline number is the combined bench, with the core-AI vs data-science split shown in each exhibit’s stacked bars.

Why are these counts a floor, not a census?+

Only people whose job title says AI or data science are counted. Capability that sits under other titles is invisible to this method: United Rentals has 76 people with “digital” in their title and Herc has a telematics organization, none of which count here. Coverage of the public workforce data also varies: it reaches under a third of McGrath’s and Alta’s employees, so their zeros are weaker evidence than United’s six. The paid diagnostic resolves counts to a named-profile roster.

Where does the data come from?+

Public workforce data for the bench, resolved to each corporate family before counting (H&E Equipment Services is counted inside Herc, which acquired it in June 2025). Public job postings for openings, earnings-call transcripts for narrative (the six most recent calls per company, one for EquipmentShare, public since January 2026), and company or vendor press for disclosed products. All public data, pulled August 2026; no surveys, no interviews.

Why is Sunbelt counted at 24,000 employees, and why does EquipmentShare’s coverage look odd?+

Sunbelt Rentals Holdings (formerly Ashtead Group, NYSE-listed since March 2026) reports 24,000 employees consolidated, of which about 21,800 are in North America; we use the consolidated figure because its AI and data titles are not split by region. EquipmentShare’s public workforce records exceed its reported 7,700 employees, which flags them as an outlier: its per-10,000 density is the number to read, its absolute count is less reliable.

Does a bigger bench mean deployed AI, or verified ROI?+

No, twice over. Headcount measures owned capability, not deployed capability: vendor and contractor staff appear in no company’s numbers, and most of what rental companies call AI today sits inside telematics and pricing platforms built with partners. And every product claim here is disclosed, not independently verified; no company in this cohort has disclosed a dollar-denominated AI return.