How AI Search Is Recommending Live Chat Software: Monthly Trends

Executive Summary

Tidio remains the category leader in Live Chat Software, holding 68.9% valid recommendation coverage in September 2026 against 67.9% in July 2026, a 1.0-point gain the benchmark classifies as stable. tawk.to holds second position at 61.3%, down from 63.0%, keeping the gap to the leader at 7.6 points.

The one brand the benchmark flags as a significant riser this period is LiveChat (Text S.A., formerly LiveChat Software S.A.), whose valid recommendation coverage rose from 11.2% in July 2026 to 23.3% in September 2026, a 12.1-point gain. That movement came after an August spike to 40.2%, followed by a 16.9-point pullback in September. Despite the retreat, LiveChat remains well above its July baseline. The gap between LiveChat and Zendesk Chat narrowed from 34.4 points in July 2026 to 19.3 points in September 2026.

Every other tracked brand moved within the range the benchmark classifies as normal month-to-month variation. Zendesk Chat posted the largest decline from baseline, down 3.0 points from 45.6% to 42.6%, while Intercom made the strongest steady upward move, rising 1.0 point to 46.3%. Intercom also recorded the category's notable placement gains, with its top-3 rate rising 10.3 points and rank-one rate rising 5.5 points from baseline.

Each monthly run begins with 800 prompt-surface observations (607 unique questions in July 2026, 626 in August 2026, 580 in September 2026) across the benchmark's defined AI/search surface universe. All 800 mentioned a tracked brand or competitor in each month; 446 were relevant and 354 were irrelevant in July 2026, 478 were relevant and 322 were irrelevant in August 2026, and 520 were relevant and 280 were irrelevant in September 2026. The public metrics use the 349 qualified observations in July 2026, 371 in August 2026, and 408 in September 2026 that survive both qualification stages.

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.

These stages define the denominator used for every brand-level percentage that follows; the qualified benchmark observations row is the base for all coverage, presence, and placement rates in this report. August 2026 fell between these two months with 371 qualified observations.

Benchmark-Level Metrics

The qualified observation count grew steadily across the three months, from 349 to 371 to 408, while the recommendation-shaped answer share followed a different path: 53.6% in July, down to 46.6% in August, then recovering to 47.8% in September. The valid recommendation shortlist share rose to 77.9%, meaning a larger proportion of the qualified set contained usable recommendation signals than at baseline.

AI Recommendation Trend

Leadership holds while the mid-field recalibrates after a spike

The top of the category held firm in September 2026, with Tidio and tawk.to retaining their positions. The more notable story is the mid-field: LiveChat's August surge to 40.2% did not hold, settling back to 23.3% in September, which reshaped the competitive spacing around the third-through-sixth positions.

The only movement the benchmark's series-based check classifies as a significant riser this period is LiveChat's 12.1-point coverage gain from baseline; every other brand's movement falls within its normal range. The category-level change in September comes from the combination of several smaller movements plus LiveChat's retreat from its August spike rather than from any single unstable shift.

What Changed This Month

LiveChat (Text S.A., formerly LiveChat Software S.A.): Significant Riser With a Two-Month Story

LiveChat's valid recommendation coverage rose from 11.2% in July 2026 to 23.3% in September 2026, a 12.1-point gain the benchmark's series-based check classifies as significant. However, the path was not linear: the brand spiked to 40.2% in August 2026 before pulling back 16.9 points in September. The September level still stands well above the July baseline, which is why the brand is classified as a significant riser in the cumulative series despite the month-over-month decline.

The supporting metrics moved with the baseline-to-current story. Raw mention presence rose from 18.1% in July to 34.3% in September, a significant 16.2-point gain. The recommended top-3 rate climbed from 7.7% to 20.3%, up 12.6 points and significant, while the rank-one rate rose from 4.3% to 12.5%, up 8.2 points and significant. In absolute terms, valid recommendations grew from 39 in July 2026 to 95 in September 2026, with rank-one placements rising from 15 to 51.

LiveChat's presence rate nearly doubled from 18.1% in July to 34.3% in September, while its top-3 and rank-one rates both stayed far above baseline even after the August peak receded. The benchmark records the size and direction of this change; it cannot establish what caused the August spike or the September pullback.

Highest-priority diagnostic: Which prompt categories drove the September retention above baseline, and did the August spike concentrate on particular surfaces that then normalized?

Intercom: Placement Gains While Coverage Holds

Intercom's valid recommendation coverage moved from 45.3% in July 2026 to 46.3% in September 2026, up 1.0 point — within the benchmark's stable range for the brand.

The more notable signal is in placement quality. Intercom's top-3 rate rose from 22.1% to 32.4%, a significant 10.3-point gain, with top-3 placements increasing from 77 to 132. The rank-one rate rose from 6.3% to 11.8%, a significant 5.5-point gain, with rank-one placements rising from 22 to 48. Raw mention presence moved in the opposite direction, from 64.8% in July 2026 to 62.0% in September 2026.

Intercom is winning top-3 and rank-one positions in a larger share of the answers where it appears, even as its raw presence edges down — a placement gain that outpaces its visibility trend.

Highest-priority diagnostic: Which prompt types are yielding the additional top-3 and rank-one placements, and which brands are being displaced in those positions?

Zendesk Chat: Steady Declines From Baseline

Zendesk Chat's valid recommendation coverage fell from 45.6% in July 2026 to 42.6% in September 2026, down 3.0 points. That is a two-month decline, with the brand moving from 45.6% to 45.0% in August and then to 42.6% in September. The brand remains 4th in the category, but the gap to 3rd-place Intercom widened.

Raw mention presence fell from 61.6% in July 2026 to 56.1% in September 2026, down 5.5 points. The top-3 rate moved in the opposite direction, from 17.5% to 19.9%, while the rank-one rate held nearly flat at 3.2% versus 3.1%. In absolute terms, valid recommendations rose from 159 in July 2026 to 174 in September 2026 as the qualified observation count grew.

Zendesk Chat's coverage decline comes alongside a falling presence rate, meaning the brand appears in fewer answers overall, even though its placement within those answers improved slightly.

Highest-priority diagnostic: Which prompts no longer surface Zendesk Chat, and which brands are taking its place in those answers?

Olark: Continued Marginal Presence

Olark's valid recommendation coverage fell from 6.0% in July 2026 to 3.2% in September 2026, a 2.8-point decline that falls within the benchmark's stable range but continues a pattern of marginal presence for the brand.

Raw mention presence fell from 8.6% to 4.7%, a significant 3.9-point decline. The top-3 rate fell from 0.6% to 0.2%, with top-3 placements falling from 2 to 1. In absolute terms, valid recommendations fell from 21 in July 2026 to 13 in September 2026, and the brand's total mentions fell from 30 to 19 across the surface universe.

Olark remains visible in a narrow set of prompts, but within that set it is placed in recommendation lists and shortlists less often than at baseline.

Highest-priority diagnostic: Which prompts still mention Olark, and what attributes appear alongside the brand in those remaining contexts?

Buyer-Intent Interpretation

In September 2026, all 408 qualified observations fell into the discovery and consideration cluster. The benchmark run captured no qualified pricing or value observations, and no multi-brand comparison observations. This means the public metrics describe which brands AI systems recommend during initial discovery, but they cannot yet describe how those systems handle price sensitivity or direct competitive comparison.

The practical implication is that the benchmark measures first-position visibility and recommendation presence, not the full buyer journey. A brand could perform strongly in discovery prompts while faring differently in later-stage comparison questions. The current data cannot distinguish between those scenarios, which is exactly the kind of question a company-level analysis can answer.

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 AI Industry Market Discovery research program from CiteWorks Studio.

• Methodology
• Metrics
• Standards

Report-Specific Interpretation Notes

• Small-count movements: brands with low absolute counts (LivePerson at 2 valid recommendations in September 2026, Olark at 13) can show large percentage swings from small changes in raw numbers. Read their movements with appropriate caution.
• Qualified denominator: all percentage figures use the 408 qualified observations in September 2026, not the 800 prompts collected. The funnel from collection to qualification is disclosed in Benchmark Context.
• Directional analysis: month-over-month movement identifies changes worth investigating, but it does not by itself establish the cause of those changes. LiveChat's August spike and September pullback both require investigation.
• LiveChat is the only brand the benchmark's series-based check classifies as a significant riser this period; its presence, top-3, and rank-one metrics all moved significantly above baseline even as the brand retreated from its August peak.

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 describe which brands AI systems recommend, but they do not explain why those recommendations form. Beneath LiveChat's 12.1-point baseline gain are questions about which high-intent prompts it now wins, why its August spike did not hold, which brand takes the recommendation when it loses, what attributes AI systems associate with each option, and which external sources shape those answers. The same questions apply to every brand in the category, whether stable or moving.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. Rather than reacting to aggregate movements, the audit identifies the specific levers that influence where and how AI systems recommend a brand.

Request an AI visibility audit