How AI Search Is Recommending Recruiting Agencies: Monthly Trends

Executive Summary

The August 2026 benchmark records zero valid recommendation coverage for every tracked recruiting agency in the qualified observation set. Robert Half, the July 2026 category leader at 46.8% valid recommendation coverage, recorded 0.0% in August 2026 — a decline of 46.8 points, the largest movement of any tracked brand. All ten tracked brands moved from a July 2026 range running from 19.4% (Kelly Services) to 46.8% (Robert Half) down to 0.0% coverage in August 2026.

This is not a case of one or two brands gaining or losing ground relative to competitors. Every one of the ten tracked brands registered a significant decline in valid recommendation coverage this month, and none recorded a significant increase. The ranking order that existed in July 2026 no longer applies in the current data: no brand holds a lead over another in August 2026. The recommendation-shaped answer share fell from 32.1% of qualified observations in July 2026 to 0.0% in August 2026, and the valid recommendation shortlist share fell from 56.0% to 0.0% over the same period. The qualified observation set expanded from 521 to 579 records, yet the larger set produced no recommendation signals for any tracked brand.

The data describes a difference in the measured output distribution between the two months. It does not establish why AI systems returned factual answers instead of recommendation-shaped responses in August 2026.

This report tracks prompt-surface observations across the AI/search surface universe for recruiting agencies. The August 2026 collection began with 791 prompt-surface observations (551 unique questions); all 791 mentioned a tracked agency or competitor, of which 676 were relevant and 115 were irrelevant, leaving 579 qualified observations that feed the public benchmark metrics below. The July 2026 baseline began from 792 prompt-surface observations (514 unique questions), yielding 521 qualified observations after the same relevance filtering.

Key Findings

Benchmark Context

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set.

Benchmark-Level Metrics

Between the baseline and current months, the response-type mix collapsed into a single category. In July 2026, the benchmark recorded 184 factual answers, 126 recommendation shortlists, 119 ranked lists, 43 comparison analyses, 41 mixed responses, and 7 pricing analyses. In August 2026, all 579 observations were classified as factual answers.

AI Recommendation Trend

Every Tracked Brand Recorded Zero Valid Recommendation Coverage in August 2026

No brand outperformed any other in August 2026; every tracked brand's valid recommendation coverage fell to zero within the same measurement period, and the movement was uniform in both direction and near-uniform in magnitude across all ten brands.

What Changed This Month

Category-wide: All brands

The defining story of August 2026 is that no brand was recommended. Every tracked agency saw its valid recommendation coverage fall to 0.0% from July 2026 levels that ranged from 19.4% (Kelly Services) to 46.8% (Robert Half). The qualified set was larger in August 2026 (579 observations vs 521), yet it contained zero valid recommendations for any brand.

Raw presence also fell to zero. In July 2026, raw mention presence rates ranged from 27.6% (Kelly Services) to 70.6% (Robert Half). In August 2026, every brand recorded 0.0% presence: no tracked agency was mentioned, named, or referenced in any qualified observation.

The August 2026 observations were classified entirely as factual answers. This means the AI systems returned general information rather than naming or recommending specific agencies. When the response type is uniformly factual, the benchmark cannot attribute any recommendation credit, and the absence applies equally across all ten tracked brands.

Highest-priority diagnostic: Which prompts shifted from recommendation-shaped answers to factual answers, and is the change a surface-level behavior, a prompt-set difference, or a broader pattern in how AI systems responded to the August 2026 question set?

Leader: Robert Half

Robert Half entered August 2026 as the category leader with 46.8% valid recommendation coverage, ahead of Adecco (35.9%) and Randstad (38.2%). Its August 2026 result of 0.0% removes that leadership position from the current data. This is a movement of 46.8 points, the largest single movement in the category.

The supporting metrics for Robert Half all moved to zero. Its raw mention presence rate fell from 70.6% to 0.0%, its top-3 rate fell from 30.9% to 0.0%, and its rank-one rate fell from 22.1% to 0.0%. In July 2026, Robert Half held the rank-one position in 115 observations; in August 2026, there were no rank-one positions to hold.

The distinction is between visibility and recommendation. In July 2026, Robert Half had both: it was present in most observations and was the most recommended agency. In August 2026, it had neither. The category leader going into the month is indistinguishable from every other tracked brand in the current month's data.

Highest-priority diagnostic: Which July 2026 prompts produced Robert Half recommendations, and did the August 2026 prompt set include those same questions, or did the mix change?

Notable gap: Kelly Services vs Robert Half

The widest baseline gap in the category was between Kelly Services and Robert Half, at 27.4 points in July 2026. That gap closed to 0.0 in August 2026, as both brands fell to zero coverage. The gap between Heidrick & Struggles and Robert Half followed the same pattern, closing from 27.0 points to 0.0.

These gaps existed in the baseline because Robert Half held the highest coverage in the category and the other brands trailed by substantial margins. The current month shows neither a leader nor trailing positions: there is no competitive hierarchy to observe in August 2026, only a uniform absence of recommendation signals.

Highest-priority diagnostic: Did the August 2026 prompt set favor factual, information-seeking queries over the recommendation and comparison queries that produced the July 2026 coverage gaps?

Buyer-Intent Interpretation

In July 2026, 56.0% of the 521 qualified observations (292) contained a valid recommendation shortlist, and all ten tracked brands appeared in the qualified set. The August 2026 qualified set, while larger at 579 observations, produced no valid recommendation shortlists and no brand mentions, leaving no basis for buyer-intent classification in the current period.

Because the current month's qualified observations produced no brand mentions, the public benchmark cannot answer questions about price, value, or head-to-head comparison for August 2026. The baseline month showed only minimal coverage of those classes, with 7 pricing analysis responses and 43 comparison analysis responses out of 521. The current month provides no basis for evaluating commercial questions beyond basic factual information.

Brand Opportunity Summary

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations (query, surface, recommendation outcome, rank, sentiment, and citations where exposed). Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Presence in a response is not automatically treated as proof of causation.

About This Benchmark

This report is part of the LLM Authority Index AI Market Discovery research program.

• AI Industry Market Discovery
• AI Industry Market Discovery Methodology
• AI Industry Market Discovery Metrics
• AI Industry Market Discovery Standards

Report-Specific Interpretation Notes

• Because all ten tracked brands recorded the same zero-coverage result, this pattern is treated as a difference in the measured output distribution rather than as a competitive outcome for any single brand. Matched-prompt evidence would be needed to determine why the recommendation-shaped answer share fell to zero.
• The qualified denominator grew from 521 to 579 observations, but the current month produced no brand mentions, making the denominator irrelevant to brand-level comparison this month.
• Movement between months identifies changes worth investigating; it does not by itself establish the cause of those changes.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

The aggregate percentage for August 2026 tells a simple story: no tracked brand received recommendations. But the questions beneath that percentage remain open. Which high-intent prompts produced coverage in July 2026 and did not reappear in August 2026? Which competitor takes the recommendation when a brand loses coverage? What attributes do AI systems associate with each agency, and which external sources shape those answers?

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. Rather than waiting for the next monthly signal, an audit can identify which specific queries and surfaces drive recommendation outcomes and where the August 2026 signals went.

Request an AI visibility audit