How AI Search Is Recommending Chatbots: Monthly Trends

Executive Summary

Tidio is now the clear category leader in AI search recommendations, with valid recommendation coverage of 59.5% in September 2026, narrowly ahead of Intercom at 58.4%. This reverses the prior order: Zendesk had led (or tied for the lead) in both July and August 2026, but Zendesk's coverage dropped to 0.0% in September after a tracking change split it into Zendesk Chat at 43.5%. The gap between Tidio and Intercom is just 1.1 points, making the top of the category a two-brand race.

Intercom was the strongest sustained upward mover, with valid recommendation coverage rising 7.6 points from 50.8% in July 2026 to 58.4% in September 2026, a significant gain across the full series. Tidio also rose significantly, up 15.6 points from 43.9% in July to 59.5% in September. Among brands not affected by the tracking change, the sharpest single-month decline came from Freshdesk, down 3.4 points from 22.2% in August to 18.8% in September, though its coverage remains within the range of normal variation across the full series.

The most consequential development this month is the tracking split of Zendesk and LiveChat into separate brand entities, materially reshaping the category table.

Each monthly benchmark run begins with 800 prompt-surface observations (551 unique questions in September 2026, down from 575 in August and up from 497 in July) across the benchmark's defined AI/search surface universe. All 800 observations mentioned a tracked brand or competitor in each month; 661 were relevant and 139 were irrelevant in September 2026, compared with 732 relevant and 68 irrelevant in August and 719 relevant and 81 irrelevant in July. The public metrics use the 437 observations that survived both qualification stages in September 2026, down from 500 in August and 476 in July. A tracking change affecting how Zendesk and LiveChat are counted took effect in September 2026, and is addressed throughout this report.

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 fell from 476 in July 2026 to 437 in September 2026, with August 2026 at 500 as an intermediate peak. The valid recommendation shortlist share rose from 59.9% to 67.0% over the full series, meaning a larger share of the September set was eligible for recommendation analysis despite the smaller qualified base. The share of recommendation-shaped answers rose from 32.6% in July to 43.7% in September.

Benchmark-Level Metrics

The September 2026 benchmark set contains fewer qualified observations than July but a higher share of recommendation-shaped and valid recommendation responses. This makes September a stronger basis for recommendation analysis per observation, though the smaller base warrants caution when comparing brand-level percentages across the full series.

AI Recommendation Trend

Recommendation Coverage Across the Category

Tidio and Intercom now hold the top two positions at 59.5% and 58.4%, while Zendesk Chat and LiveChat (Text S.A.) occupy the next tier following the tracking change. Intercom and Tidio both moved beyond the threshold for normal month-to-month variation across the full series, while Zendesk and LiveChat fell to zero because their coverage now appears under the new brand entries.

What Changed This Month

Tidio: Significant Rise to Category Leader

Tidio's valid recommendation coverage rose 15.6 points from 43.9% in July 2026 to 59.5% in September 2026, a significant gain across the full series. The month-over-month rise was also significant, up 12.3 points from 47.2% in August 2026.

Raw mention presence rose 23.1 points from 55.9% in July to 79.0% in September. Top-three recommendation rate rose 8.6 points from 22.5% to 31.1%, and rank-one rate rose 2.8 points from 5.9% to 8.7%, with 38 first-position recommendations in September 2026.

The distinction to notice is that Tidio's presence gain outpaced its coverage gain, meaning the brand appeared in a wider set of AI answers even as its recommendation share rose.

Highest-priority diagnostic: Which prompt families drove the simultaneous presence and coverage gains, and whether the rank-one rate can follow the top-three rate upward.

Intercom: Significant Sustained Rise

Intercom's valid recommendation coverage rose 7.6 points from 50.8% in July 2026 to 58.4% in September 2026, a significant gain across the full series. The brand also gained 5.4 points from August 2026, though that month-over-month movement fell within the normal range.

Raw mention presence rose 5.4 points from 78.1% to 83.5%. Top-three rate rose 9.8 points from 29.8% to 39.6%, with 173 top-three placements in September. Rank-one rate rose 2.6 points from 8.6% to 11.2%, with 49 first-position recommendations.

The distinction to notice is that Intercom's top-three gain exceeded its overall coverage gain, indicating stronger positioning within the recommendations it does receive.

Highest-priority diagnostic: Which surfaces drove the top-three rate improvement, and whether Intercom can close the remaining gap to Tidio.

Zendesk and LiveChat: Tracking Change Reshapes the Table

Zendesk's valid recommendation coverage fell from 50.8% in July 2026 and 53.8% in August to 0.0% in September 2026, with Zendesk Chat appearing at 43.5%. LiveChat fell from 22.2% in August to 0.0%, with LiveChat (Text S.A. (formerly LiveChat Software S.A.)) appearing at 19.7%. These movements are classified as significant.

The raw mention presence for Zendesk Chat is 62.9% and for LiveChat (Text S.A.) is 26.8% in September. Zendesk Chat's top-three rate is 23.6% and rank-one rate is 9.4%, while LiveChat (Text S.A.) shows a top-three rate of 14.2% and rank-one rate of 9.2%.

The underlying brands remain highly visible, but the public benchmark now reports them under separate entities, which changes how their coverage compares with Tidio and Intercom.

Highest-priority diagnostic: Whether the new Zendesk Chat and LiveChat (Text S.A.) tracking reflects distinct product offerings or a reclassification of the same brand coverage.

Freshdesk: Continued Decline Within Normal Range

Freshdesk's valid recommendation coverage fell 3.4 points from 22.2% in August 2026 to 18.8% in September 2026, following a 1.5-point decline in the prior month. Across the full series, coverage is down 4.9 points from 23.7% in July 2026, which stays within the threshold for normal variation.

Raw mention presence fell 13.5 points from 43.7% in July to 30.2% in September. Top-three rate fell 7.4 points from 15.6% to 8.2%, with 36 top-three placements in September. Rank-one rate fell 1.4 points from 2.3% to 0.9%, with 4 first-position recommendations.

Freshdesk's presence and top-three placements declined by a wider margin than its overall coverage movement, indicating the brand is being surfaced in fewer AI answers overall.

Highest-priority diagnostic: Which surfaces or prompt types accounted for the lost presence, and which brands appear in the prompts where Freshdesk previously held placement.

Buyer-Intent Interpretation

In September 2026, all 437 qualified observations fell into the Brand Recommendation cluster. There were no qualified observations in the Pricing & Value or Multi-Brand Comparison clusters in July, August, or September 2026, meaning the public benchmark does not yet answer which brands win pricing discussions or head-to-head comparisons. The recommendation signal in this category is driven entirely by direct brand recommendation prompts.

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, including the 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 AI Industry Market Discovery research program.

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

Report-Specific Interpretation Notes

• Movement between months identifies changes worth investigating, not established causes.
• Small observation counts for brands like Landbot, Chatfuel, and Ada mean their coverage percentages rest on a narrow base of valid recommendations (15, 14, and 36 respectively in September 2026).
• The qualified benchmark set (437 observations) is the denominator for all brand-level percentages, not the raw collection universe (800 prompts).
• A tracking change in September 2026 altered how Zendesk and LiveChat are tracked, splitting their coverage into Zendesk Chat and LiveChat (Text S.A.).
• A brand's coverage gain or loss identifies a benchmark movement, not a result attributable to any single cause.

Next Step

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

Beneath the aggregate percentages lie specific questions: which high-intent prompts is a brand winning or losing, which competitor appears when a brand loses placement, what attributes do AI systems associate with each option, and which external sources shape those answers? The September 2026 benchmark identifies where attention is warranted, but it does not by itself explain the mechanism, particularly the tracking change affecting Zendesk and LiveChat.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It answers the why behind the movement, including whether the new brand entities reflect distinct products or reclassified coverage.

Request an AI visibility audit