How AI Search Is Recommending PR Management Agencies: Monthly Trends

Executive Summary

The category experienced a significant leadership rotation and broad de-concentration between July 2026 and September 2026. Highwire is now the coverage leader at 16.7%, but only 0.8 points ahead of Walker Sands at 15.9%, making this a narrow lead in a tightly packed field. The former leaders declined sharply over the period: FINN Partners fell from 34.8% in July to 14.9% in September, Real Chemistry from 30.8% to 12.3%, and Burson from 20.1% to 8.3%. No single brand now dominates the way FINN Partners did in July.

The strongest upward mover was 5WPR, which entered the tracked set in August at 13.3% and settled at 7.6% in September. Its August entry gain of 13.3 points from 0.0% in July was significant, followed by a significant 5.7-point decline in September. Walker Sands was the quiet riser this month, moving from 12.6% in August back up to 15.9% in September, now the closest challenger to the leader.

Four brands were significant decliners in September: Burson, FINN Partners, Real Chemistry, and Ruder Finn. Ruder Finn at 5.1% has now fallen 11.8 points from its 16.9% July baseline and holds the weakest mid-tier position of the previously established brands. The category is measurably less concentrated than at baseline, with the gap between the leader and the fifth-place brand narrowing substantially.

Each monthly run begins with 639 prompt-surface observations (473 unique questions) in September 2026, up from 478 (339 unique questions) in July and 585 (441 unique questions) in August, across the benchmark's defined AI/search surface universe. Of the September observations, 639 mentioned a tracked brand or competitor; 329 were relevant and 310 were irrelevant. The public metrics use the 276 observations that survive both qualification stages, compared with 279 qualified observations in July and 278 in August.

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.

The qualified observation count remained nearly flat across the series (279 in July, 278 in August, 276 in September), even as the raw collection grew from 478 to 639 prompts. August's 585-prompt collection sat between the two endpoints, with 584 brand mentions and 321 relevant prompts. The higher raw volumes reflect a growing share of irrelevant prompts over time.

Benchmark-Level Metrics

The recommendation-shaped answer share contracted from 38.4% in July to 27.9% in September, and the valid recommendation shortlist share fell from 59.1% to 40.6%. That contraction, which also showed in August at 32.0% and 46.4% respectively, means a smaller proportion of qualified observations produced a usable recommendation shortlist. This helps explain why several brands saw coverage declines across the series even as their presence held or rose.

AI Recommendation Trend

A De-Concentrated Field Replaces a Single Leader

The category has moved from a clear single-brand lead in July to a closely packed cluster in September. Highwire leads at 16.7%, but Walker Sands at 15.9% and FINN Partners at 14.9% sit within 1.8 points. The four significant decliners (Burson, FINN Partners, Real Chemistry, Ruder Finn) account for the de-concentration, while the stable mid-tier brands now hold more relative weight.

The category-level change came primarily from the combination of several significant declines. Four brands moved beyond normal month-to-month variation between July and September, and 5WPR's August entry was itself significant. No single decline explains the full reshuffling; together they compressed the top of the field.

What Changed This Month

Highwire: New Leader in a Compressed Field

Highwire holds 16.7% valid recommendation coverage in September, down 4.8 points from 21.5% in July, a decline that is not significant. Its prior-month movement was negligible, down 0.2 points from 16.9% in August.

Highwire's placement signals strengthened even as coverage eased. Its top-three recommendation rate rose from 2.1% in July to 3.6% in September, and its rank-one rate improved from 0.7% to 1.5%. The brand recorded 46 valid recommendations out of 276 qualified observations in September, down from 47 in August and 60 in July. Net sentiment held at 0.9.

The distinction is that Highwire became the category leader by relative stability: it declined far less than the former leaders, not by gaining share. Its raw mention presence rate fell from 27.6% to 19.9% across the series, yet it now holds the top coverage position.

Highest-priority diagnostic: Which prompt categories and surfaces retain Highwire's presence while former leaders lost theirs, and how narrow is its lead beneath the aggregate rate?

Walker Sands: Closest Challenger After a Strong Month

Walker Sands holds 15.9% valid recommendation coverage in September, up 3.3 points from 12.6% in August, though that gain is not significant by itself. The brand sits 0.8 points behind Highwire, the narrowest leadership gap in the current field.

Walker Sands recorded 44 valid recommendations in September. Its top-three rate was 4.7% and its rank-one rate 1.5%. Raw mention presence came in at 17.8%. Net sentiment was positive at 0.9.

The distinction is that Walker Sands converts a moderate presence into a high coverage rate relative to the broader field.

Highest-priority diagnostic: Which prompts recommend Walker Sands ahead of the former category leaders, and did any specific surface drive the September gain?

5WPR: Significant Rise Then Significant Decline

5WPR entered the tracked set in August with 13.3% valid recommendation coverage, up from 0.0% in July, a significant gain. That level did not hold: September coverage fell to 7.6%, a 5.7-point decline that is also significant.

The brand recorded 21 valid recommendations in September, down from 37 in August. Its top-three rate fell from 6.5% to 3.3%, and its rank-one rate from 2.5% to 1.5%. Raw mention presence eased from 19.1% to 14.5%. Net sentiment remained positive at 0.8.

The distinction is that 5WPR's August entry over-indexed on a concentrated set of prompts, and September's decline reflects a return to a lower but still measurable presence.

Highest-priority diagnostic: Which prompt types drove the August spike, and did those same prompts stop recommending 5WPR in September or simply shift to other agencies?

FINN Partners: From Leader to Third Place

FINN Partners fell 19.9 points from 34.8% in July to 14.9% in September, the largest cumulative decline in the category and significant. The brand recorded 41 valid recommendations in September, down from 97 in July.

Raw mention presence dropped from 45.2% to 22.5% across the series, a 22.7-point decline across the period. Top-three rate fell from 12.2% to 5.4%. Net sentiment held at 0.9. FINN Partners retains third place, but its position is now within a cluster rather than atop the field.

The distinction is that FINN Partners lost presence and coverage together, indicating it appeared in fewer AI answers overall rather than being displaced within answers.

Highest-priority diagnostic: Which surfaces or prompt types account for the 22.7-point presence drop, and did the decline concentrate on specific AI platforms?

Real Chemistry: Significant Decline with Held Rank-One Strength

Real Chemistry fell 18.5 points from 30.8% in July to 12.3% in September, a significant decline. The brand recorded 34 valid recommendations in September, down from 86 in July.

Top-three rate fell from 14.3% to 6.2%, while rank-one rate eased slightly from 5.4% to 4.7%. Real Chemistry still holds the strongest rank-one position in the category at 4.7%, more than three times the level of the next brand. Raw mention presence was 41.7% in September.

The distinction is that Real Chemistry's presence held up better than its coverage: a 41.7% presence rate produced only a 12.3% coverage rate. The brand is still mentioned broadly but recommended in a smaller share of answers.

Highest-priority diagnostic: Which prompts mention Real Chemistry but recommend a different agency, and which competitor captures those recommendations?

Burson: Significant Decline Despite High Presence

Burson fell 11.8 points from 20.1% in July to 8.3% in September, a significant decline. The brand recorded 23 valid recommendations in September, down from 47 in August and 56 in July.

Burson's raw mention presence rate actually rose across the series, from 51.2% in July to 58.7% in September. Top-three rate fell from 16.9% to 5.1%. Net sentiment eased from 0.7 to 0.5. The presence-to-coverage gap is the widest in the category.

The distinction is the opposite of Real Chemistry's situation: Burson holds the highest presence in the category yet converts it into one of the lowest coverage rates among established brands.

Highest-priority diagnostic: Which prompts mention Burson but recommend another agency, and which evidence sources drive those alternative recommendations?

Buyer-Intent Interpretation

In September, all 276 qualified observations fell into the Brand Recommendation class. The comparison and pricing clusters recorded zero observations, as they did in July and August. The public benchmark therefore measures which agency AI systems recommend for general PR management needs, but it cannot yet answer how AI systems compare agencies head-to-head or how they assess pricing and value. Those commercial questions remain outside the current measurement scope.

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. Source presence 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 Methodology
• AI Industry Market Discovery Metrics
• AI Industry Market Discovery Standards

Report-Specific Interpretation Notes

• Small-count movement: brands with fewer than 5 valid recommendations (Allison Worldwide, SparkPR, WE Communications) show movement that is more sensitive to individual observations.
• Qualified denominator vs raw collection: the 276 qualified observations in September are a subset of 639 collected prompts after relevance and eligibility filtering.
• Directional analysis: significant month-over-month movement 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 percentages in this report raise questions that the benchmark alone cannot answer: which high-intent prompts does a brand win outright, which competitor takes the recommendation when the brand loses, what attributes do AI systems associate with each agency, and which external sources shape those answers? For a leader like Highwire holding a narrow 0.8-point edge, the question is which prompts keep it on top; for brands like Burson and Real Chemistry with wide presence-to-coverage gaps, the question is which competitor captures the recommendations when they lose.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. The public benchmark shows the scoreboard; the audit explains the game.

Request an AI visibility audit