How AI Search Is Recommending Customer Service Software: Monthly Trends

Executive Summary

The customer service software benchmark reached its most active month of the four-month series in October 2026, and leadership changed hands. Freshdesk moved into the category lead at 67.9% valid recommendation coverage, up 11.8 points from 56.1% in September 2026 and up 15.3 points from 52.6% in July 2026, with 222 valid recommendations out of 327 qualified observations. That places it well clear of the next brand, Intercom at 54.7%.

Intercom was the strongest riser on the baseline-to-current comparison, up 22.3 points from 32.4% in July 2026 to 54.7% in October 2026, a significant 22.3-point gain, and up 10.0 points from September 2026, also significant. Its top-three rate rose 13.0 points to 26.3% and its rank-one rate rose 4.7 points to 7.0%, both significant.

The sharpest movement in the table is structural rather than competitive. Zendesk, HubSpot Service Hub, and Zoho Desk carry 0.0% current coverage because they were replaced in the tracked set in September 2026 by Zendesk Chat, HubSpot Live Chat, and Zoho Inventory. That reflects a change in which brands are tracked, not a measured decline in AI recommendation standing. Zendesk Chat and HubSpot Live Chat each fell back in October from their September entry levels, down 24.9 points and 15.5 points respectively.

Against the prior month, the category was active in both directions. Seven brands gained coverage on the baseline-to-current comparison, three were significant decliners tied to the tracked-set change, and three were stable.

Each monthly run begins with 800 prompt-surface observations. In July 2026 that collection contained 484 unique questions; 800 prompts mentioned a tracked brand or competitor, 709 were relevant and 91 were irrelevant, leaving 481 qualified observations. In October 2026 the collection contained 617 unique questions; 800 prompts mentioned a tracked brand or competitor, 522 were relevant and 278 were irrelevant, leaving 327 qualified observations. The public metrics use the 327 observations that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Two critical or high-severity factual inconsistencies were detected this month, spanning three AI platforms.

Freshdesk

One high-severity inconsistency concerns free plan availability, with a confidence of 0.9. When asked "Is Freshdesk no longer free?", AI platforms provided conflicting information about whether a free offering still exists. Gemini stated that Freshdesk's permanent free plan is gone and that only a 6-month Free Program remains, citing pages including a Freshdesk versus Freshservice comparison and a pricing explainer. Copilot stated that Freshdesk is still free, with a free tier for up to 2 agents, often capped at 6 months, citing a free-plan overview and the company's own Free Program support page. The two claims cannot both be true. Flagged sources supporting the two sides include a third-party review noting a free plan for up to 2 agents, the company's Free Program support article describing a 6-month free window with up to two agents, and a pricing explainer stating that the permanent multi-agent free plan is gone.

Help Scout

One high-severity inconsistency concerns the starting monthly price per agent, with a confidence of 0.85. When asked "What is the best email ticketing system?", responses differed on the entry-level offering. ChatGPT stated that a free plan supports up to 5 users and 100 contacts per month, citing a competitor pricing page and the company's own site. Copilot stated a starting price of $25 per agent per month, citing a helpdesk comparison and two email ticketing roundups. One flagged source on the pricing side references email-first support at $25 for 5 agents.

Benchmark Context

Questions This Section Answers

• How did the qualified observation base change between July 2026 and October 2026?
• What response-composition shifts occurred across the two qualification stages?

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

The following figures summarize the category-level shift in response composition across the two qualification stages.

The qualified set continued to narrow, with August 2026 at 342 and September 2026 at 360 as intermediate months. The more consequential shift is in response composition: recommendation-shaped answers rose from 34.1% of the set in July 2026 to 59.9% in October 2026, and valid recommendation shortlists rose from 62.8% to 78.0%. The tracked set still contains 10 companies, but three of those identities changed in September 2026.

AI Recommendation Trend

Questions This Section Answers

• Which brands gained the most recommendation coverage, and who now leads the category?
• Why do Zendesk, HubSpot Service Hub, and Zoho Desk show 0.0% current coverage?

A New Leader Emerges as Intercom Closes the Gap

Several brands moved beyond normal month-to-month variation, and the category-level change came from a combination of those movements rather than a single brand. The zero-percent current readings for Zendesk, HubSpot Service Hub, and Zoho Desk reflect their removal from the tracked set rather than a measured decline in recommendation standing.

What Changed This Month

Questions This Section Answers

• What drove Freshdesk's move into the category lead, and how broad is its coverage strength?
• Why did Zendesk Chat and HubSpot Live Chat fall back in October 2026 after entering tracking in September?
• How did Front's October recovery differ from its shortlist-placement trend?

Freshdesk

Freshdesk moved into the category lead, with valid recommendation coverage up 15.3 points from 52.6% in July 2026 to 67.9% in October 2026, a significant baseline move. The month-over-month move was also significant: up 11.8 points, the brand's largest single-month gain in the series. The gain is a two-month upward streak.

Placement improved alongside coverage. Top-three rate rose 3.2 points from 40.8% in July 2026 to 44.0% in October 2026, and rank-one rate rose 2.1 points from 6.2% to 8.3%. Raw mention presence rose 6.0 points to 86.2%, a significant move. The brand recorded 144 top-three placements and 27 rank-one placements out of 327 qualified observations.

The distinction worth noting is that Freshdesk now leads on both presence and coverage, but its rank-one rate sits well below its top-three rate. Coverage strength is broad rather than concentrated at the top position.

Highest-priority diagnostic: which prompt themes and surfaces account for the 11.8-point September-to-October gain, and whether the top-three to rank-one gap reflects answer structure rather than positioning.

Intercom

Intercom was the strongest riser in the series, with valid recommendation coverage up 22.3 points from 32.4% in July 2026 to 54.7% in October 2026, a significant move, and up 10.0 points from 44.7% in September 2026, also significant. The brand has now risen in two consecutive months.

Placement gains were equally broad. Top-three rate rose 13.0 points from 13.3% in July 2026 to 26.3% in October 2026, a significant move, and rank-one rate rose 4.7 points from 2.3% to 7.0%, also significant. Raw mention presence rose 22.4 points to 71.9%, a significant move. The brand recorded 86 top-three placements and 23 rank-one placements out of 327 qualified observations.

The distinction here is that Intercom is gaining on both coverage and placement quality at once, which separates it from brands gaining coverage while losing shortlist position. The brand still trails Freshdesk by 13.2 points.

Highest-priority diagnostic: which surfaces drove the presence increase from 49.5% in July 2026 to 71.9% in October 2026, and whether the top-three gains concentrate in shortlist-style answers.

Help Scout

Help Scout rose 6.9 points from 35.3% in July 2026 to 42.2% in October 2026, a significant baseline move, and 4.1 points from 38.1% in September 2026, which was not significant on its own. The pattern across the series is volatile: the brand fell to 22.8% in August 2026, recovered to 38.1% in September 2026, and rose again in October 2026. The 15.3-point August-to-September swing remains the largest single-month move in the brand's series.

Presence rose 9.6 points from 46.4% in July 2026 to 56.0% in October 2026, a significant move, with 183 mentions across the qualified set. Top-three rate moved down 1.5 points to 12.8%, and rank-one rate rose 0.5 points to 1.5%. The brand recorded 42 top-three placements and 5 rank-one placements out of 327 qualified observations.

The distinction is visible versus recommended: Help Scout is named in more answers but is not converting that into shortlist position at the same rate.

Highest-priority diagnostic: which prompts produce a Help Scout mention without a shortlist placement, and what drove the August-to-September swing.

Salesforce Service Cloud

Salesforce Service Cloud rose 11.4 points from 26.2% in July 2026 to 37.6% in October 2026, a significant baseline move, and 9.3 points from 28.3% in September 2026, also significant, its largest single-month move in the series. The brand had dipped to 19.6% in August 2026 before two consecutive monthly gains.

Top-three rate rose 3.9 points to 14.1% and rank-one rate rose 0.7 points to 3.4%, neither significant on the baseline comparison. Raw mention presence rose 4.2 points to 49.5%, also not significant. The brand recorded 46 top-three placements and 11 rank-one placements out of 327 qualified observations.

The distinction is that the coverage recovery is outpacing the placement recovery, so the gain is concentrated in valid recommendations rather than in higher positions.

Highest-priority diagnostic: which prompts returned in the September and October gains, and why top-three placement did not recover at the same pace.

Zendesk Chat and HubSpot Live Chat

Zendesk Chat and HubSpot Live Chat both entered tracking in September 2026 and both fell back in October 2026. Zendesk Chat was down 24.9 points, the sharpest month-over-month decline among currently tracked brands. HubSpot Live Chat was down 15.5 points from 25.3% to 9.8%, also significant. On the baseline-to-current comparison both still read as significant risers, up 25.7 points and 9.8 points respectively, because their July 2026 baseline predates their entry into the tracked set.

Zendesk Chat's current position is stronger than its coverage alone suggests. Top-three rate stands at 21.4% and rank-one rate at 12.2%, the highest rank-one share among currently tracked brands, with 70 top-three placements and 40 rank-one placements out of 327 qualified observations. Raw mention presence is 34.2%. HubSpot Live Chat recorded 11 top-three placements and 1 rank-one placement out of 327 observations, with presence at 14.1%.

The distinction is that Zendesk Chat retains disproportionate rank-one strength relative to its reduced coverage, while HubSpot Live Chat shows a much thinner placement base. Both brands are on a one-month downward streak after a single month of tracking data.

Highest-priority diagnostic: whether the September entry levels reflected a first-month tracking effect, and which surfaces drove the October pullback in each brand.

Gorgias, Front, Gladly, and Zoho Inventory

Gorgias rose 7.3 points from 20.2% in July 2026 to 27.5% in October 2026, a significant baseline move, with 90 valid recommendations out of 327 observations. Raw mention presence rose 6.6 points to 33.0%, significant, while top-three rate rose 2.3 points to 5.2% and rank-one rate rose 0.4 points to 0.6%, neither significant. The brand has risen in two consecutive months.

Front was stable on the baseline comparison, down 0.6 points from 17.7% in July 2026 to 17.1% in October 2026, but rose 6.0 points from 11.1% in September 2026, a significant month-over-month move. Presence rose 2.4 points to 25.1%, not significant. Top-three rate fell 4.1 points to 2.8%, a significant decline, and rank-one rate fell 0.3 points to 0.3%. The brand recorded 9 top-three placements and 1 rank-one placement out of 327 observations. The distinction is that Front's October recovery was driven by valid recommendations rather than by stronger shortlist placement.

Gladly was stable, up 0.4 points from 0.2% in July 2026 to 0.6% in October 2026 and flat from September 2026, with 2 valid recommendations out of 327 qualified observations. Presence was 2.5% with 8 mentions, and the brand held 1 top-three placement and no rank-one placements. Percentage movement on this base should be read with the absolute counts in view.

Zoho Inventory was flat at 0.0% coverage across the baseline-to-current comparison, down 2.2 points from 2.2% in September 2026, a significant month-over-month move. The brand recorded 1 present observation out of 327, no valid recommendations, and no top-three or rank-one placements.

Highest-priority diagnostic: for Gorgias, whether the two-month rise extends beyond ecommerce-oriented prompts. For Front, which prompts now produce valid recommendations without shortlist placement. For Gladly and Zoho Inventory, whether the small observation base supports any durable inference at all.

Zendesk, HubSpot Service Hub, and Zoho Desk

These three brands carry 0.0% current coverage because they were removed from the tracked set in September 2026, not because of a measured change in AI recommendation standing. Their removal is a tracking change, and their baseline figures from July 2026 remain in the table for continuity.

Highest-priority diagnostic: how the full-suite platforms would compare against the chat-focused offerings now tracked in their place, since the current table does not place them side by side.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters were captured in the October 2026 qualified observations?
• What commercial questions can the public benchmark not yet answer?

In October 2026, all 327 qualified observations fell into the Brand Recommendation cluster. The response mix included recommendation shortlists, ranked lists, factual answers, comparison analyses, and pricing analyses, but none of those responses were classified into dedicated pricing or multi-brand comparison clusters.

The public benchmark therefore answers which brand AI recommends, and in what position, but cannot yet answer how AI systems discuss price and value or which brand wins a head-to-head comparison. Those commercial questions remain open. The inconsistency alerts above show that pricing and availability questions are reaching AI systems from outside the qualified cluster, which is one reason the benchmark keeps pricing and comparison as distinct classes rather than folding them into recommendation coverage.

Brand Opportunity Summary

Questions This Section Answers

• Which brands warrant the most attention based on October 2026 coverage and current signals?
• What is the highest-priority diagnostic question for each brand in the tracked set?

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 Visibility Market Discovery research program.

• AI Visibility Industry Market Index
• AI Visibility Industry Market Methodology
• AI Visibility Industry Market Metrics
• AI Visibility Industry Market Standards

Report-Specific Interpretation Notes

• Small-count movement: Gladly's 0.6% coverage rests on 2 valid recommendations out of 327 qualified observations, and Zoho Inventory's 0.0% rests on 1 present observation with no recommendations. Movement on those bases should be read alongside the absolute counts.
• Tracked-set rotation: Zendesk, HubSpot Service Hub, and Zoho Desk carry 0.0% current coverage because they were removed from the tracked set in September 2026, not because of a measured decline in AI recommendation standing. Their replacements, Zendesk Chat, HubSpot Live Chat, and Zoho Inventory, have shorter series histories.
• Qualified denominator: all percentages are calculated against the qualified set, 327 observations in October 2026 versus 481 in July 2026, not the raw prompt collection of 800.
• 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? On pricing and availability questions in particular, this month's inconsistency alerts show that AI systems are drawing on conflicting external sources, and a category-wide benchmark cannot resolve which one is shaping the recommendation.

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