How AI Search Is Recommending Customer Service Software: Monthly Trends

Executive Summary

The tracked brand set in the customer service software benchmark changed materially in September 2026. Three full-suite platforms — Zendesk, HubSpot Service Hub, and Zoho Desk — were replaced in the tracked comparison by Zendesk Chat, HubSpot Live Chat, and Zoho Inventory. That is a change in which brands were tracked, not a measured decline in AI recommendation standing for the brands that were removed.

Freshdesk retains the category lead at 56.1% valid recommendation coverage, up from 52.6% in July 2026 and 49.1% in August 2026, and now sits 5.5 points ahead of second-place Zendesk Chat, which entered tracking this month at 50.6% coverage.

The strongest upward mover was Zendesk Chat, which posted 50.6% valid recommendation coverage in its first tracked month, reaching the top rank in 92 of 360 qualified observations (25.6% rank-one rate). Intercom was the strongest riser among brands tracked in both July and September, climbing 12.3 points from 32.4% to 44.7%, a significant move driven by a 19.9-point gain in raw mention presence to 69.4%.

Front was the sharpest decliner among continuously tracked brands, down 6.6 points from 17.7% in July to 11.1% in September, a significant move on the baseline-to-current comparison. Among the seven brands tracked continuously since July, six gained coverage and only Front declined.

Each monthly run begins with a full prompt-surface collection before qualification. In July 2026, the run began with 800 prompt-surface observations (484 unique questions), of which 800 mentioned a tracked brand or competitor; 709 were relevant and 91 were irrelevant, leaving 481 qualified observations. In September 2026, the run began with 800 prompt-surface observations (605 unique questions), of which 800 mentioned a tracked brand or competitor; 563 were relevant and 237 were irrelevant, leaving 360 qualified observations reported below.

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 funnel above shows how the raw collection narrows to the qualified analysis set used for every brand-level percentage below.

Benchmark-Level Metrics

The qualified set narrowed from 481 observations in July 2026 to 360 in September 2026, with August 2026 at 342 as an intermediate month. The tracked brand composition also changed: each month tracks 10 companies, but three of the tracked identities changed — Zendesk, HubSpot Service Hub, and Zoho Desk were replaced by Zendesk Chat, HubSpot Live Chat, and Zoho Inventory — while the other seven brands were tracked continuously. The recommendation-shaped answer share fell from 34.1% to 31.1%, while the valid recommendation shortlist share held nearly flat at 62.5%.

AI Recommendation Trend

Chat-Specific Offerings Reshape the Leadership Table

The zero-percent rows for Zendesk, HubSpot Service Hub, and Zoho Desk reflect their removal from the September tracked set, not a measured drop in AI recommendation standing; their replacements, Zendesk Chat, HubSpot Live Chat, and Zoho Inventory, entered with 50.6%, 25.3%, and 2.2% coverage respectively. Among the seven brands tracked continuously since July, six gained coverage and only Front declined.

What Changed This Month

Freshdesk

Freshdesk's valid recommendation coverage rose 3.5 points from 52.6% in July 2026 to 56.1% in September 2026, a move within normal month-to-month variation but enough to keep the brand in the category lead. The brand was the most recommended in the set with 202 valid recommendations out of 360 qualified observations.

Mention presence rose 5.0 points to 85.3%, while top-three rate slipped 5.2 points from 40.8% to 35.6%. Rank-one rate rose slightly to 6.9% (25 of 360 observations) from 6.2% in July, and net sentiment stayed flat at 0.8.

Freshdesk is the only brand with both leading coverage and near-universal presence; its strength lies in breadth across AI answers.

Highest-priority diagnostic: which prompt themes are driving the coverage gain, and whether the top-three rate decline signals weaker positioning in shortlist responses.

Zendesk Chat

Zendesk Chat entered the tracked set in September 2026 with 50.6% valid recommendation coverage, the second-highest in the category. The brand appeared in 73.6% of qualified observations with 182 valid recommendations.

The brand posted a 35.0% top-three rate and a 25.6% rank-one rate, the highest rank-one share in the category, with 92 of 360 observations placing it first.

The entry of Zendesk Chat alongside Zendesk's removal from the tracked set means the full-suite platform and its chat product are not being compared in the same table; the rank-one strength is notable within the current tracked universe.

Highest-priority diagnostic: which prompt types are driving the rank-one placements, and whether they overlap with the prompts where Zendesk previously appeared.

Intercom

Intercom rose 12.3 points from 32.4% in July 2026 to 44.7% in September 2026, a significant move and the strongest among brands tracked continuously across the quarter. Valid recommendations reached 161 of 360 observations, up from 156 of 481 in July.

Rank-one placement rose 3.3 points from 2.3% to 5.6% (20 of 360 observations), and top-three rate climbed 8.4 points to 21.7% (78 of 360 observations). Net sentiment held at 0.8.

The gain is driven by a 19.9-point presence increase to 69.4%, meaning more AI answers now name Intercom and more of those answers place it in a top-three position.

Highest-priority diagnostic: which AI surfaces or prompt categories drove the presence jump, and whether the top-three gains are concentrated in shortlist-style answers.

Front

Front declined 6.6 points from 17.7% in July 2026 to 11.1% in September 2026, a significant move and the sharpest among continuous brands. Valid recommendations fell from 85 of 481 observations in July to 40 of 360 in September.

The decline bottomed in August at 5.9%, then Front recovered 5.2 points in September, a significant prior-month move; even with that recovery, the brand remains well below its July level. Top-three rate fell 4.1 points from 6.9% to 2.8% (10 of 360 observations), with rank-one rate flat at 0.6% (2 observations).

Front's mention presence was 21.1% in September, close to its July level of 22.7%, meaning the brand appears in answers more often than it is recommended within them.

Highest-priority diagnostic: what AI systems say when Front is mentioned but not recommended, and which prompt themes formerly produced Front recommendations.

Help Scout

Help Scout rose 2.8 points from 35.3% in July 2026 to 38.1% in September 2026, within normal month-to-month variation, with 137 valid recommendations out of 360 observations.

The brand had dropped to 22.8% in August 2026, then recovered 15.3 points in September, a significant prior-month move. Raw mention presence rose 11.7 points to 58.1%, a significant increase.

Help Scout's recovery is broad-based on presence, but top-three rate fell 4.9 points to 9.4%, meaning the brand is more visible yet placed lower when recommended.

Highest-priority diagnostic: what drove the August decline and the September recovery, and whether the top-three rate erosion reflects a positioning shift.

Gorgias

Gorgias rose 4.8 points from 20.2% in July 2026 to 25.0% in September 2026, within normal variation, with 90 valid recommendations out of 360 observations.

Raw mention presence rose 7.2 points to 33.6% (121 mentions), a significant gain, and top-three rate climbed 3.2 points to 6.1% (22 observations). Rank-one rate nearly tripled from 0.2% to 0.6% (2 observations).

The coverage gain was driven by a sharp September rebound: Gorgias had fallen to 14.9% in August, then rose 10.1 points in September, a significant prior-month move.

Highest-priority diagnostic: which prompt themes or surfaces drove the September recovery, and whether the gains extend beyond ecommerce-focused queries.

Salesforce Service Cloud

Salesforce Service Cloud rose 2.1 points from 26.2% in July 2026 to 28.3% in September 2026, within normal month-to-month variation, with 102 valid recommendations out of 360 observations.

The brand dropped to 19.6% in August 2026, then recovered 8.7 points in September, a significant prior-month move. Net sentiment improved from 0.7 to 0.8.

Rank-one rate fell 1.3 points from 2.7% to 1.4% (5 of 360 observations), even as overall coverage recovered, indicating a position-quality challenge.

Highest-priority diagnostic: which prompts recovered in September, and why rank-one placements did not recover at the same pace.

HubSpot Live Chat

HubSpot Live Chat entered the tracked set in September 2026 with 25.3% valid recommendation coverage, with 91 valid recommendations out of 360 observations. The brand appeared in 35.6% of qualified observations.

Top-three rate reached 6.4% (23 of 360 observations) with rank-one at 1.1% (4 observations). Net sentiment was 0.8.

HubSpot Live Chat's entry replaces HubSpot Service Hub in the tracked set, meaning the two HubSpot products are not directly compared in the same table.

Highest-priority diagnostic: which prompt categories drive the brand's recommendations, and how the live chat focus maps to buyer-intent queries.

Zoho Inventory

Zoho Inventory entered tracking in September 2026 with 2.2% valid recommendation coverage, a small but significant move given its tiny base. Valid recommendations reached 8 of 360 observations.

The brand held 2.8% raw mention presence, with one rank-one placement and one top-three placement. Net sentiment was 0.8.

Zoho Inventory is adjacent to the customer service category rather than a direct service platform, so its small coverage base is expected for the prompt universe.

Highest-priority diagnostic: whether the brand appears in customer-service-adjacent prompts or only in inventory-focused queries.

Gladly

Gladly rose 0.4 points from 0.2% in July 2026 to 0.6% in September 2026, within normal month-to-month variation for its very small base. Valid recommendations totaled 2 of 360 observations.

Mention presence rose 0.9 points to 1.9% (7 mentions), and the brand earned its first top-three placement at 0.3% (1 of 360 observations). Net sentiment improved from 0.4 to 0.6.

Gladly's movement is driven by a handful of observations; percentage changes on such a small denominator should be read cautiously.

Highest-priority diagnostic: whether the gains reflect durable presence improvements or small-count variation.

Buyer-Intent Interpretation

In September 2026, all 360 qualified observations fell into the Brand Recommendation cluster. The response mix included recommendation shortlists, comparison analyses, factual answers, and ranked lists, but none of those responses were classified into dedicated pricing or multi-brand comparison clusters.

The public benchmark therefore answers "which brand does AI recommend" but cannot yet answer "how does AI discuss price and value" or "which brand wins a head-to-head comparison." Those commercial questions remain open for company-level analysis.

Brand Opportunity Summary

The benchmark identifies where attention is warranted across the tracked set; a company-level analysis is needed to explain why specific brands gain or lose recommendation coverage.

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: Gladly's coverage of 0.6% is based on 2 valid recommendations out of 360 qualified observations; percentage movement on such a small base should be interpreted with caution.
• Brand-set rotation: September 2026 replaced three tracked brands (Zendesk, HubSpot Service Hub, Zoho Desk) with chat- and inventory-focused offerings (Zendesk Chat, HubSpot Live Chat, Zoho Inventory); coverage of 0.0% for the departed brands reflects the tracking change, not a measured decline in AI recommendation standing.
• Qualified denominator: all percentages are calculated against the qualified set (360 observations in September 2026, versus 481 in July 2026), not the raw prompt collection.
• Directional analysis: 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 leave the important questions open. Which high-intent prompts is a brand actually winning? When a brand loses the recommendation, which competitor takes the slot? What attributes do AI systems associate with each option, and which external sources shape those answers? A category-wide benchmark cannot answer those questions from percentages alone.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It identifies the specific prompts where a brand is vulnerable, the evidence sources that drive AI recommendations, and the surfaces where presence does not convert to placement.

Request an AI visibility audit