100 Questions, Seventeenth Category: Video Conferencing and Webinar Software, Where Zoom Is the Default and Teams and Meet Are Picked for Their Suites

Zoom is the default

Zoom is named in 76 of 100 answers, recommended in 73 and picked as the answer's own choice in 63. It is picked in 18 of the 20 direct advice questions, and it is the only pick for online teaching, a law firm and a church. It is also picked in 48 of the 75 answers to questions that do not mention Zoom. Only 3 of the 63 lines that pick Zoom mention Microsoft 365 or Google Workspace. Webex is fourth: named in 33 answers, recommended in 30, picked in 21, and 16 of the 21 picking lines name Cisco, security, compliance, regulation, government, enterprise needs or encryption.

Teams and Meet are picked for their suites

Microsoft Teams is named in 55 answers, recommended in 53 and picked in 41; Google Meet is named in 49, recommended in 48 and picked in 35. The picks come with a condition. On 31 of the 41 lines that pick Teams, the answer names Microsoft 365, Office or Outlook: "Already standardized on Microsoft 365 → Teams". On 24 of the 35 lines that pick Meet, it names Google Workspace, Gmail, Google Calendar or Classroom. The same condition shows up where the model describes them without choosing them. Meet and Teams have the largest gaps between recommended and picked of any brand named 15 or more times, 13 and 12 answers. In all 12 of those Teams answers it is described as the fit for Microsoft 365 users, and in 11 of the 13 Meet answers as the fit for Google Workspace, Gmail or Classroom users. Counting both, the suite condition is attached to Teams in 43 of the 53 answers that recommend it and to Meet in 35 of 48.

Webinars and verticals get a specialist

Twenty of the questions are about webinars. Zoom is picked in 16 of them, and a webinar specialist is picked in 15. The model rarely swaps Zoom out; it adds a tool built for the job: Demio is picked in 11 of the 20, Livestorm in 9, ON24 and WebinarJam in 5 each. Across all 100 answers, the specialists are picked every time they are named: Demio 14 of 14, Livestorm 12 of 12, ON24 6 of 6, BigMarker 6 of 6. The six telehealth and therapist questions follow the same pattern: Doxy.me is picked in all six, SimplePractice in five, and all six answers mention Zoom's healthcare offering. Open source and self-hosted questions go to Jitsi, BigBlueButton and Nextcloud Talk. Retired products stay retired: Skype appears only as the subject of the Skype alternatives question, whose answer notes it was retired in May 2025, and BlueJeans, Hopin and Amazon Chime are never named.

Decisive, and far from Google

Ninety-two of the 100 answers pick at least one brand. Seven of the eight that do not still assign brands to stated cases; the eighth, for an enterprise with 5,000 employees, is one sentence naming three contenders. Twenty-six of 30 head-to-heads pick every brand named. Two pick one side: Zoom vs Webex picks Webex for Cisco-heavy enterprises, and Google Meet vs Jitsi picks Jitsi for open source. Of the 271 picks, 227 went to brands whose own website did not rank in Google's top ten for the question, 84 percent. Google's top ten is 87 percent third-party pages.

Vendor pages, and list sites that do not rank

Across 100 answers there were 280 citation events to 92 domains. Vendors' own sites took 170 of them, 61 percent, and 52 answers cited nothing but vendor pages. Zoom's domains are cited in 49 answers, 32 of them to questions that do not mention Zoom, and Zoom's help center alone in 25. microsoft.com is cited in 35 answers and google.com in 20, including their support and documentation pages. The most-cited independent sites are list and comparison pages: dupple.com in 8 answers, stackfyi.com in 6, toolradar.com in 5. None of the three was in Google's top ten for any question that cited it. Reddit and Wikipedia have zero citations.

Forty-three of the 280 citation events involved a domain that also sat in Google's top ten for that query, 15 percent, level with website builders for the lowest overlap in the series so far. For third-party sites alone it was 6 of 110.

Say which buyer you beat the bundle for. The model's first move for Teams and Meet is to ask what the buyer already pays for. A standalone vendor needs a plain, citable statement of the case where it wins anyway: external clients, large webinars, compliance, no downloads.

Be the specialist the model adds. On webinar questions the model kept Zoom and picked a specialist in 15 of 20 answers. Demio was picked in all 14 answers that named it. Name your use case (lead generation, course sales, virtual events) the way a buyer would ask for it.

Your help center is part of your marketing. Zoom's support site was cited in 25 answers, and Microsoft's and Google's documentation pages show up throughout. Pricing, participant limits and feature pages are what the model quotes; keep them current and specific.

List sites that do not rank still get read. Three comparison and list sites were cited in 18 answers between them without ranking in Google's top ten for any of those questions. If a page like that covers your category, it is part of your AI search footprint.

One run per query means answer variance is unmeasured; Benchmark 01 runs each query three times across three surfaces. The teardowns are collected on different days, so differences between categories mix the category with the date. The brand dictionary covers the general video conferencing and webinar software market and deliberately excludes vertical tools, so their appearances are described in prose and not counted. The readers are AI models, not people: they follow a written protocol and agree with each other closely, but a shared blind spot would not show up as disagreement. The line between picked and recommended is a judgment, documented in the protocol with examples. Sentiment is not coded. No vendor-tool cross-check was read for this category. Google's control counts a brand as ranking only when its own domain is in the top ten; a listicle that features the brand does not count. One run per question, one day, United States, English, logged out. Answers vary between runs, so these are frequencies for this sample. The readers are AI models, not people, and the line between picked and recommended is a judgment; the three codes the readers disagreed on were all of that kind. The suite counts check the answer line the readers cite for each pick, so a condition stated only in a neighboring sentence is not counted. google.com and microsoft.com are counted as the vendors' own sites, so their share of first-party citations includes documentation pages that serve many products.

Every number above can be recomputed from these files. CC BY 4.0: use them, cite the page.

• queries.csv: the 100 queries with intent labels.
• brands.csv: the 45-brand dictionary with aliases and canonical domains.
• mentions.csv: 399 coded brand mentions with position, type, a 0/1 picked column and whether the brand's domain was in Google's top ten.
• citations.csv: 280 citation events with domain class and Google overlap.
• observations.csv: one row per query with brand, recommendation, citation and control counts.
• dictionary-changes.md: the one dictionary change made after collection (the short name Teams)
• brands-original.csv: the dictionary as first drafted, before that change
• coder-notes.md: coding method, reader agreement, the judgment calls that matter in this category, and dictionary notes
• codebook.md: the rulebook, with dated amendments through v2.0.
• reader-protocol.md: the written protocol both readers coded against (codebook v2.0).
• picked_stats.json: the picked-level figures.

The earlier teardowns: Teardown 01, project management, Teardown 02, crm, Teardown 03, email marketing, Teardown 04, help desk, Teardown 05, accounting, Teardown 06, payment processing, Teardown 07, payroll, Teardown 08, hr software, Teardown 09, applicant tracking, Teardown 10, password managers, Teardown 11, endpoint security, Teardown 12, business intelligence, Teardown 13, data warehouse and etl, Teardown 14, ecommerce platforms, Teardown 15, website builders, Teardown 16, scheduling.

In video conferencing the model's picks follow the software suite a buyer already uses, and its reading list is vendor help pages plus list sites that do not rank. Finding out what the model is reading for your category is the first thing a Category Salience Brief does, with a query set you approve first.