Hays (HAYPY), SelectQuote (SLQT): AI Absorbs the Matching Work

Hays cut permanent placement volumes 14% as average fees rose 2%; SelectQuote ties over $30m of annual savings to AI and keeps licensed agents on the call.

On 20 and 25 August 2026, UK specialist recruiter Hays plc (HAYPY) and US Medicare distributor SelectQuote, Inc. (SLQT) each used their earnings calls to describe the same reallocation: AI now does the information-gathering and matching work, and human capacity is being concentrated on the part of the transaction where someone still carries the consequence [1][2].


What commission-per-transaction intermediaries actually sell

Both companies earn headcount multiplied by transactions. Hays places professionals and charges the employer a percentage of the hired candidate's salary. SelectQuote sells Medicare Advantage — the US senior health plans administered by commercial insurers — through licensed agents who compare plans with the consumer by phone and earn a commission from the insurer on each completed enrollment.

The fee the buyer paid once covered two separate things. One was information and matching: who is looking, what they cost, which plan covers a given drug. The other was someone willing to answer for the consequences if the choice turns out wrong. AI now supplies the first at close to zero marginal cost. The second does not transfer, because a model does not bear the cost of a bad decision.

Hays priced the second half using exit costs. Measured as a percentage of salary, the cost of removing a bad hire at director level can run more than twice as high as at associate level, and the company describes its own fee as a modest insurance premium against that risk [1]. Intermediaries are therefore redrawing the line in two directions at once — selecting revenue where the cost of failure is high, and detaching the cost of everything around the transaction from transaction count.


Volumes, average fees and cost savings moving in opposite directions

Hays' unit economics already reflect that selection. Permanent placement volumes fell 14% year over year while the average fee per placement rose 2%, which management attributes to a continued focus on higher-value roles. Like-for-like net fees fell 8% to GBP 906m, operating expenses came down by GBP 70m, consultant headcount fell 12%, and pre-exceptional operating profit rose 3% to GBP 48.6m [1]. The new CEO's Momentum strategy narrows the business to 16 countries and six global specialisms and sets a target of more than 50% higher net-fee productivity per consultant [1].

SelectQuote is reworking the cost side. The company has deployed AI-enabled enrollment support tools across the business to match agent capacity to demand, streamlined agent workflows with sales-assist technology, and is expanding AI-powered quality assurance to review and coach agents, leaving the licensed agent with the call itself [2]. Fiscal 2026 revenue was $1.62bn, up 6%, with adjusted EBITDA of $109m. Fiscal 2027 guidance is $1.35bn to $1.45bn of revenue, down 14% at the midpoint, with adjusted EBITDA of $90m to $115m — roughly 60 basis points of consolidated margin expansion — and management attributes more than $30m of annualized expense improvement to reduced labor intensity from AI plus back-office streamlining [2].

A peer's own numbers limit how far this reads across. On its 27 July call, Kforce (KFRC) said revenue had grown for three consecutive quarters and returned to pre-pandemic and therefore pre-AI norms, with growth in both consulting and traditional staff augmentation [3].


The control point moves from holding information to signing for the outcome

An intermediary's leverage is shifting. Holding the list and the pricing used to be enough to charge for; what can be charged for now is answering for the outcome of the choice. On the revenue side that means screening business by how costly failure is, with low-consequence transactions either conceded or repriced down. On the cost side, matching, screening and quality assurance are being separated from headcount, so expense no longer scales linearly with transaction count and the operating metric becomes fee or margin per remaining person rather than the number of people.

The fee that gets conceded does not disappear. It moves toward companies selling the substitute tool, and toward companies paid to verify information once it can be generated in bulk. What is confirmed so far is a shift of steps and pricing inside these intermediaries: SelectQuote attributes its own revenue decline to Medicare Advantage carrier behavior and the Inflation Reduction Act, not to AI [2]. Two things can be checked directly from here — whether Hays moves toward its target of more than 50% higher net-fee productivity per consultant, and whether SelectQuote delivers roughly 60 basis points of margin expansion with revenue down 14% at the midpoint [1][2].


Which companies this change may affect:

  • Recruit Holdings (6098.T): Its Indeed unit sells Smart Sourcing and Smart Screening, which perform exactly the matching and screening work being conceded. Management describes the target market as the roughly $200bn hiring-expenditure pool that includes placement services, and raised full-year HR Technology guidance to $11.4bn, up 18.7% year over year, on an assumption that US job postings fall about 4% [4].
  • First Advantage (FA): It is paid per background check on hires actually made. Roughly half of its base growth came from customers' enterprise-wide labor reshaping programs and the labor churn they create, and the company cites "bad AI" bringing fraud into the recruiting process as a reason customers add more verification components [5].
  • KE Holdings (BEKE): Chinese residential brokerage runs on the same commission-per-transaction structure. The company split the agent role and created an AI-assisted client manager who is deliberately not paid per transaction to handle the pre-viewing work, and existing-home contribution margin rose 6.1 percentage points year over year, which it attributes to lower fixed labor cost [6].

Sources

[1] Drillr · Hays plc (HAYPY) · 2026-08-20 · FY2026 full-year results call

"Thirdly, a focus on higher-value roles where the cost of failure is higher and the potential impact from AI on these roles is lower."

[2] Drillr · SelectQuote, Inc. (SLQT) · 2026-08-25 · FY2026 fourth-quarter earnings call

[3] Drillr · Kforce (KFRC) · 2026-07-27 · Q2 2026 earnings call

[4] Drillr · Recruit Holdings (6098.T) · 2026-08-07 · FY2026 first-quarter earnings call

[5] Drillr · First Advantage (FA) · 2026-08-06 · Q2 2026 earnings call

[6] Drillr · KE Holdings (BEKE) · 2026-08-21 · Q2 2026 earnings call

This is only meant to surface industry changes and companies you may have overlooked - it is not a stock recommendation.

Want deeper analysis?

Ask drillr anything about 6098.T, BEKE, FA, HAYPY, KFRC, SLQT — powered by SEC filings, earnings calls, and real-time data.

Try drillr.ai for free

Drillr can make mistakes. Information only — not investment advice. Learn more