Recruit's AI Agent Moat: Indeed Outgrows the Lapping Worry, But Guidance Stays in Neutral
HR Tech's $11.9B outlook and 46.5% margin are raised again—yet Recruit leaves consolidated guidance untouched and says the enterprise inflection is still gradual.
6098.T · Earnings Call · 2026-11-11
The beat that didn't travel
Recruit's FY2026 Q2 was, on the surface, another AI monetization victory lap. HR Technology revenue rose 27.5% year-on-year in U.S. dollar terms and 36.6% in the U.S., a new quarterly record. CEO Hisayuki Idekoba admitted he had worried that lapping last year's AI product launches would drag growth. “Q2 delivered strong results, up 27.5% year-on-year in U.S. dollar terms and up 36.6% year-on-year in the U.S.” — Hisayuki Idekoba, HR Technology Business Executive · 2026-11-11 Instead, he raised full-year HR Tech revenue to $11.9 billion, up 23.2%, from $11.4 billion and 18.7% in August, and guided segment EBITDA+S margin to roughly 46.5%. Then the odd part: consolidated guidance did not move. CFO Junichi Arai said “we are leaving the full year guidance unchanged from what was upwardly revised in August” — Junichi Arai, CFO · 2026-11-11. He explained that at the assumed JPY156 to the dollar, the HR Tech upgrade is worth only about 4% of EBITDA+S and about 5% of base EPS versus August's consolidated guide. That is a small flow-through—and a telling one. Management is not letting a segment beat become a company-wide promise, especially with FX and seasonal Q3/Q4 hiring uncertainty. The stock reaction likely hinges less on the $0.5 billion segment raise than on why it stops at the segment line.Muse, agents, and the two-sided moat
The call's most charged exchange was about Meta's Muse, a personal AI agent that can search jobs. An analyst noted Recruit's shares fell after the announcement and asked whether agent-led job search changes the monetization premise. Idekoba's answer was the strategic centerpiece: job discovery is commoditizing; the bottleneck is the flood of applications.He argued Indeed sits on both sides of the marketplace, so it can move from pay-per-click toward performance billing—cost per interview or cost per qualified application—and use AI screening to price actual hiring outcomes. This is the new AI agent narrative, and it is genuinely fresh in the company's keyword trail: a quarter ago the language was AI automation and automation tool; now the frame is agent-versus-agent, with Recruit claiming the two-sided data moat. The market's fear is that a single-sided agent disintermediates the job board. Recruit's rebuttal is that a single-sided agent cannot solve the employer's constraints or the candidate's latent signals—and that the application flood makes matching more valuable, not less.So it's not just about Muse, but AI agent looking for jobs, AI agents can find jobs. Google for jobs does that too and it's becoming quite commoditized. So looking for jobs, finding jobs can be done. But the key — the problem here is the application for jobs are receiving many applications by AI agents.