How AI Search Is Recommending Web Conferencing: Monthly Trends

Executive Summary

August 2026 shows valid recommendation coverage moving down across five of the ten tracked web conferencing brands, while the remaining five held within normal month-to-month variation. The category leader, Zoom, saw its valid recommendation coverage fall from 69.9% in July to 63.3% in August, a drop of 6.6 points. The gap between Zoom and the second-place brand, Google Meet, widened from 1.7 points to 3.3 points as both brands' rates fell, though at different magnitudes.

RingCentral posted the largest single-month move in the category, falling 15.0 points from 38.5% to 23.5% coverage. This occurred alongside a drop in raw mention presence, from 50.2% to 41.0%. Alongside RingCentral, Google Meet fell 8.2 points (68.2% to 60.0%), Cisco Webex fell 7.7 points (61.5% to 53.8%), and Dialpad Meetings fell 3.9 points (11.0% to 7.1%). No brand posted a rise in valid recommendation coverage this month. The five brands with declines beyond typical month-to-month variation — Zoom, Google Meet, Cisco Webex, RingCentral, and Dialpad Meetings — account for most of the change; the remaining five brands held within normal variation.

The benchmark universe expanded from 418 qualified observations in July to 493 in August, and that growth is an important factor behind this month's rate movements for several top brands, since a brand's absolute recommendation count can rise even as its share of a larger observation pool falls. The recommendation-shaped answer share held roughly flat at 35.3% in August versus 34.9% in July, while the valid recommendation shortlist share dipped from 70.6% to 64.5%.

Each monthly run begins with 800 prompt-surface observations (535 unique questions in July, 577 in August) across the benchmark's defined AI/search surface universe. Of those, all 800 mentioned a tracked brand or competitor in both months; 574 were relevant and 226 were irrelevant in July, versus 632 relevant and 168 irrelevant in August. The public metrics use the 418 observations (July) and 493 observations (August) 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. August's funnel shows a larger qualified set than July (493 versus 418 observations) despite the same 800 raw prompts, driven by a higher share of relevant prompts.

Benchmark-Level Metrics

The following metrics summarize the qualified benchmark set at the aggregate level.

The valid recommendation shortlist share fell from 70.6% in July to 64.5% in August, even as the total qualified observation count rose. This means a larger share of August's observations were not recommendation-shortlist responses, even though the absolute number of valid recommendation shortlists increased from 295 to 318. The category is being discussed in a broader set of answer formats, not just direct recommendations.

AI Recommendation Trend

The top three brands retained their rank order, while coverage rates fell across most of the category

The category-level change came from several brands moving beyond typical month-to-month variation rather than from one outsized mover. Zoom, Google Meet, Cisco Webex, RingCentral, and Dialpad Meetings all saw their valid recommendation coverage rates fall by more than normal variation would explain. RingCentral's 15.0-point drop was the largest single move, but for Zoom, Google Meet, and Cisco Webex, absolute recommendation counts mostly held steady or rose even as their rates fell, because the qualified observation base grew from 418 to 493 between July and August.

What Changed This Month

Zoom

Zoom retained the leadership position with valid recommendation coverage of 63.3% in August, down from 69.9% in July, a decline of 6.6 points. July was the baseline month in this two-month series, so the month-over-month and cumulative movement are the same figure. Zoom's raw mention presence held steady at 98.4% in August versus 98.3% in July, meaning the brand was still surfaced in nearly every qualifying observation (485 of 493, up from 411 of 418).

The rate decline did not come from fewer recommendations in absolute terms. Zoom's valid recommendation count rose from 292 to 312, its top-three placement count rose from 266 to 282, and its rank-one placement count rose from 228 to 237. But because the total qualified observation base grew from 418 to 493, each of those counts represents a smaller share of the total: the top-three rate fell from 63.6% to 57.2%, and the rank-one rate fell from 54.5% to 48.1%.

The distinction here is that Zoom's raw output grew, but the benchmark's overall observation pool grew faster, diluting Zoom's share of the top ranking positions. The net sentiment score also eased from 0.8 to 0.7.

Highest-priority diagnostic: Which prompt types or surfaces are adding to the observation pool without adding proportionally to Zoom's top-three or rank-one placements?

Google Meet

Google Meet's valid recommendation coverage fell from 68.2% in July to 60.0% in August, a decline of 8.2 points beyond typical month-to-month variation. Unlike Zoom, Google Meet's raw mention presence rate also fell, from 91.6% to 87.8%, even though its raw present count rose from 383 to 433.

As with Zoom, Google Meet's absolute recommendation counts increased even as its rates fell: valid recommendations rose from 285 to 296, top-three placements rose from 233 to 253, and rank-one placements rose from 17 to 18. The rate declines reflect a qualified observation base that grew from 418 to 493, faster than Google Meet's own count growth.

The distinction to notice is that Google Meet gained in raw counts across every recommendation tier, but its share of the larger observation pool fell across the board, from presence through rank-one placement.

Highest-priority diagnostic: Which surfaces or prompt categories are contributing to the larger observation pool that Google Meet's own growth is not keeping pace with?

RingCentral

RingCentral posted the sharpest decline in the category, with valid recommendation coverage falling from 38.5% in July to 23.5% in August, a drop of 15.0 points and the largest single-month move among tracked brands. Unlike Zoom, Google Meet, and Cisco Webex, RingCentral's raw mention presence also fell in absolute terms, from 210 to 202 observations (50.2% to 41.0%).

RingCentral's valid recommendation count fell from 161 to 116, a real decline of 45. Its top-three placement count held roughly flat, from 32 to 33, but the top-three rate still fell from 7.7% to 6.7% because the total observation base grew. Its rank-one count rose from 9 to 18, more than doubling from a small base; the rank-one rate rose from 2.1% to 3.6%. The net sentiment score eased from 0.8 to 0.7.

RingCentral is the only one of the five decliners where both presence and valid recommendation counts fell in absolute terms, not only as a share of a larger observation pool. The gap between RingCentral and Whereby narrowed from 24.2 points to 9.3 points, while the gap to leader Zoom widened from 31.4 points to 39.8 points.

Highest-priority diagnostic: Which prompt categories drove the drop in valid recommendations from 161 to 116, and did the decline concentrate on specific surfaces such as ChatGPT or Copilot?

Cisco Webex

Cisco Webex's valid recommendation coverage fell from 61.5% in July to 53.8% in August, a decline of 7.7 points beyond typical month-to-month variation. The brand's raw mention presence held roughly flat, moving from 80.9% to 81.1%, with its present count rising from 338 to 400.

Cisco Webex's valid recommendation count rose slightly, from 257 to 265, but its top-three placement count fell from 47 to 43, a real decline of four placements. Its rank-one count held at one placement in each month. The net sentiment score slipped from 0.8 to 0.7.

Cisco Webex is one of the few brands in the category where the top-three placement count fell in absolute terms, not just as a share of a larger observation pool.

Highest-priority diagnostic: Which competitor captured the four top-three placements that Cisco Webex held in July but not in August?

Dialpad Meetings

Dialpad Meetings' valid recommendation coverage fell from 11.0% in July to 7.1% in August, a decline of 3.9 points beyond typical month-to-month variation. The absolute counts are small: valid recommendations fell from 46 to 35, a decline of 11. The rank-one count moved up slightly, from zero to one placement.

This is a small-count case where percentage movement should be read with care. The brand's raw present count also fell, from 60 to 56 (14.3% to 11.4%), though that move stayed within the range of normal month-to-month variation.

Highest-priority diagnostic: Which specific prompt types account for the 11 lost valid recommendations, and are these concentrated in comparison or shortlist answer formats?

Buyer-Intent Interpretation

In August, all 493 qualified observations fell into the Brand Recommendation cluster. None of the qualified observations were classified as Pricing & Value or Multi-Brand Comparison in either month, meaning the public benchmark cannot yet answer questions about price positioning, value perception, or head-to-head comparison. The commercial questions the benchmark does answer are narrower: which brand an AI system recommends, at what rank, and with what sentiment. The absence of pricing and comparison observations is itself a finding, since it suggests the current prompt universe is weighted toward direct recommendation rather than evaluative or cost-sensitive queries.

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 movements: Dialpad Meetings and Microsoft SharePoint have small absolute observation counts, so their percentage movements carry more uncertainty than brands with larger bases.
• Qualified denominator vs raw collection: All brand-level percentages are calculated within the qualified benchmark set of 493 observations, not the 800 raw collected prompts.
• Counts vs rates: Several brands, including Zoom and Google Meet, increased their absolute recommendation counts this month while their rates fell, because the qualified observation base grew faster than their own counts.
• 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 answer the question of who is being recommended, but they do not reveal which high-intent prompts are being won, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each option, or which external sources shape those answers. For a category where most of the top tier saw coverage rates fall in August, those underlying patterns are the difference between understanding a shift in the observation pool and diagnosing a genuine competitive change.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. For RingCentral, the audit would identify which surfaces and prompt categories drove the 15.0-point decline. For Zoom, it would explain why the brand's top-three and rank-one placements grew in absolute terms while its share of the qualified observation pool fell.

Request an AI visibility audit