Zoom's AI Strategy: How the Video Leader Turns Conversations Into Customer Intelligence in 2026

TL;DR#

Zoom's AI strategy in 2026 is a deliberate reinvention: the company that defined video meetings rebranded from Zoom Video Communications to Zoom Communications in late 2024 and now calls itself an "AI-first work platform," anchored by AI Companion 3.0 and a growing fleet of agentic AI features. Zoom posted $4.67 billion in fiscal 2025 revenue and guided fiscal 2026 to roughly $4.85 billion, with its customer experience business crossing $100 million in annual recurring revenue on the strength of AI Contact Center and Zoom Virtual Agent. CEO Eric Yuan, asked what he is investing in, answered "AI, AI, and AI," and has demonstrated a "digital twin" avatar that can speak on his behalf. Yet here is the strategic gap: Zoom literally hosts the richest source of customer truth that exists — an estimated 3.5 trillion minutes of recorded human conversation a year — while its own product-feedback loop still leans on post-call surveys and NPS, a channel whose email response rates have fallen from 20–25% in 2019 to 10–15% in 2025. This post maps Zoom's real AI moves, shows where survey-based listening underuses the conversational goldmine Zoom is sitting on, and explains how conversational AI interviews extend the same "conversation-to-action" logic Zoom built its product on. The takeaway for product and CX leaders: if conversations beat forms inside the product, they beat forms in customer research too.

What is Zoom's AI strategy?#

Zoom's AI strategy is its pivot from a single-purpose video-meetings app into an AI-first work platform, built around AI Companion as a federated AI assistant and a layer of agentic AI that turns meeting conversations into completed work across Zoom Workplace, Zoom Phone, and Zoom's customer-experience products. The strategy reframes Zoom's core asset from "the place meetings happen" to "the system that understands what was said and acts on it," which is why the company dropped "Video" from its name in November 2024 and now positions itself, in Eric Yuan's words, as an AI-first platform for human connection.

That reframing matters for anyone studying how a category leader modernizes its customer relationship. Zoom is betting that value has migrated from hosting the conversation to extracting intelligence from it — the same thesis behind why AI-first customer research cannot start with a web form: the signal lives in what people actually say, not in the fields a form forces them into.

Zoom by the numbers: the scale behind the strategy#

Zoom operates at a scale that makes its AI bet consequential rather than cosmetic. The platform's reach is the reason its conversational data is so valuable — and the reason its survey-based feedback gap is so costly.

• $4.67 billion in fiscal 2025 total revenue, up 3.1% year over year, with fiscal 2026 guided to roughly $4.85–4.86 billion (Zoom Q4 FY2025 results).
• ~470,000 paying business customers, including 192,400 enterprise customers and 4,363 customers contributing more than $100,000 each per year as of late 2025.
• 300+ million daily meeting participants and an estimated 3.5 trillion meeting minutes processed per year.
• $100 million+ in annual recurring revenue from Zoom's customer experience business, which grew ARR in the high double digits in its most recent reported quarter.
• 9 of the top 10 customer-experience deals in a recent quarter included paid AI such as Zoom Virtual Agent or AI Expert Assist (CX Today).

Those last two numbers are the tell. Enterprise buyers are no longer paying Zoom only to host calls — they are paying for AI that acts on the contents of those calls. The customer-experience segment has become one of Zoom's fastest-growing lines precisely because it sells outcomes (resolved issues) rather than minutes.

Where Zoom uses AI today#

Zoom's AI footprint in 2026 spans three surfaces: workplace productivity, telephony, and customer experience — all unified under AI Companion 3.0.

AI Companion 3.0 and agentic workflows#

AI Companion 3.0 is Zoom's agentic assistant that turns scattered meeting conversations into insights, tracked progress, and finished documents without users uploading transcripts or crafting prompts. Announced at Zoomtopia in September 2025 and made generally available on December 15, 2025, it uses what Zoom calls a "federated" AI approach — combining Zoom's own large and small language models with third-party LLMs to balance quality against cost (Zoom newsroom). Its agentic retrieval can pull from meeting summaries, transcripts, and notes inside Zoom Workplace as well as connected apps like Google Drive and Microsoft OneDrive. Zoom also sells AI Companion as a $10-per-month standalone, signaling that it views AI as a revenue line, not a free add-on.

Zoom Virtual Agent and AI Contact Center#

Zoom's customer-experience AI centers on Zoom Virtual Agent, an agentic resolution engine for customer service that the company upgraded with end-to-end resolution capabilities and integrated with Zoom Phone in August 2025 to handle inbound calls around the clock. In March 2026, Zoom expanded its enterprise agentic AI platform with workflow orchestration spanning Workplace, Phone, and CX (Zoom newsroom). Notably, Zoom reframed contact-center success metrics away from handle time and deflection toward first-contact resolution and repeat-contact rates — a quiet but important shift toward measuring whether the customer's actual problem got solved.

The digital twin and the "AI, AI, and AI" mandate#

Eric Yuan has made AI the company's defining priority, telling TechCrunch in October 2025 that AI is the subject of multi-hour strategy meetings and predicting AI will enable three-to-four-day workweeks by 2030. He has gone further than rhetoric: Yuan used an AI "digital twin" avatar to speak during an investor earnings call (TechCrunch). The throughline across all three surfaces is consistent: capture the conversation, understand it, and complete the work.

The gap: Zoom hosts the conversation but still researches with surveys#

Zoom's product philosophy is "conversation to action," but its own customer-feedback loop still runs on the static-survey playbook it has otherwise transcended. This is the central irony of Zoom's AI strategy, and it is instructive for every product and CX team.

Consider what Zoom is sitting on. Every day, hundreds of millions of people describe — out loud, in their own words — what is working, what is frustrating, what they wish a meeting tool could do, and why they almost churned to a competitor. Zoom captures the audio, transcribes it, and now summarizes it with AI Companion. That is the single largest reservoir of unprompted, contextual customer voice in the productivity software market.

Yet the canonical way Zoom and companies like it learn "would you recommend us?" or "what should we build next?" is still a post-call NPS prompt or a quarterly survey. The problem is that this channel is collapsing. Email NPS response rates have fallen from 20–25% in 2019 to 10–15% in 2025, and CX leaders increasingly describe a "low survey response rate crisis" that leaves growing blind spots in why customers stay or leave (Clootrack). A score of 7 tells Zoom a customer is lukewarm; it does not tell Zoom that the customer is lukewarm because Workplace's AI summaries miss action items in technical reviews. That "why" is exactly what surveys flatten and conversations preserve — the core argument in why conversations win over surveys for real customer research.

A summarized meeting transcript is not customer research, either. AI Companion is brilliant at compressing what happened in a meeting, but a sales call or an internal standup is not a structured inquiry into a product decision. Nobody on a Zoom meeting is being asked, in a neutral and probing way, "walk me through the last time the AI summary let you down, and what you did instead." Observed conversation and elicited conversation are different instruments — and product teams need the second one to make roadmap calls, as the 2026 guide to product-market-fit research lays out.

Why conversational AI research is the natural extension of Zoom's own bet#

Conversational AI interviews extend Zoom's "conversation-to-action" logic from internal meetings to deliberate customer research, replacing the survey layer with the same dynamic, follow-up-driven dialogue that makes AI Companion useful. If Zoom believes the conversation is the unit of value inside the product, the most consistent thing it could do for product feedback is stop researching with forms.

This is the wedge for an AI-first customer-interview platform. Perspective AI runs hundreds or thousands of AI interviewer conversations at once, each one asking open questions, listening to the answer, and probing the vague or surprising parts the way a skilled researcher would — at survey scale and survey speed. Instead of a Zoom product manager guessing what a "6 on the AI summary feature" means, an AI interview asks the follow-up: "You mentioned the summaries miss action items — can you describe the last meeting where that happened?" That is the difference between a data point and an insight.

The same approach maps cleanly onto Zoom's surfaces:

• Product Discovery. Before shipping the next AI Companion capability, run conversational interviews with power users about how they actually use summaries today — the kind of continuous discovery product teams need and that modern customer-research stacks are built around.
• Customer Experience and churn. When an enterprise account's usage dips, a concierge agent can open a real conversation rather than fire a survey, surfacing the renewal risk while there is still time to act — the proactive posture described in the complete guide to AI-powered customer experience.
• Voice of customer at scale. A standing program of conversational interviews becomes a living voice-of-customer program, giving Zoom continuous qualitative signal instead of a quarterly survey spike — the build path covered in how to build a voice-of-customer program from scratch.

Zoom is far from alone in this. The same conversational-research opportunity shows up across the enterprise-software leaders modernizing their feedback loops — from how Slack listens to its messaging users and Salesforce's conversational discovery push to ServiceNow's enterprise workflow research and how Docusign is replacing forms with conversations. And it parallels the broader market shift toward conversational customer feedback documented in the 2026 state of customer research and what is replacing the survey layer. For teams evaluating which platforms enable this, the ranked list of AI customer-insight platforms for enterprises is a useful starting map.

What Zoom's playbook teaches every product and CX team#

Zoom's AI strategy offers a transferable lesson: the company learned that hosting a conversation is less valuable than understanding and acting on it, and the same principle applies to how any team gathers customer feedback. You do not need 3.5 trillion meeting minutes to apply it.

The pattern is straightforward. First, treat customer feedback as a conversation, not a form — let people answer in their own words and follow up when they say something interesting. Second, make it continuous rather than quarterly, so signal arrives while it can still change a decision. Third, scale the depth, not just the volume, using AI to run the probing follow-ups a human researcher would. Zoom built a $4.67 billion business by betting that conversation beats the static interface; the same bet is available to any team that replaces its survey layer with conversational research. It is the through-line connecting Zoom's enterprise AI moves to the everyday work of CX teams trying to understand why customers really stay or leave.

Frequently Asked Questions#

What is Zoom's AI strategy in 2026?#

Zoom's AI strategy in 2026 is to become an AI-first work platform rather than a video-meetings app, anchored by AI Companion 3.0 and agentic AI that turns conversations into completed work across Workplace, Phone, and customer experience. Zoom signaled the shift by renaming itself from Zoom Video Communications to Zoom Communications in November 2024, and CEO Eric Yuan has called AI the company's top investment priority.

What is Zoom AI Companion?#

Zoom AI Companion is Zoom's built-in AI assistant that summarizes meetings, drafts content, and — in its 3.0 version — runs agentic workflows that retrieve information and complete tasks across Zoom and connected apps. It uses a federated AI approach combining Zoom's own models with third-party LLMs, became generally available on December 15, 2025, and is also sold as a $10-per-month standalone product.

How does Zoom use AI for customer experience?#

Zoom uses AI for customer experience primarily through Zoom Virtual Agent, an agentic resolution engine in its Contact Center that handles inbound service interactions and was integrated with Zoom Phone in August 2025. Zoom's customer-experience business crossed $100 million in annual recurring revenue, and AI was attached to 9 of its top 10 CX deals in a recent quarter, reflecting strong enterprise demand for AI-powered support.

Why are surveys and NPS losing favor for customer research?#

Surveys and NPS are losing favor because response rates are falling and scores capture sentiment without the reasoning behind it. Email NPS response rates dropped from 20–25% in 2019 to roughly 10–15% in 2025, leaving growing blind spots in why customers stay or churn. A number tells a team that a customer is unhappy; only a conversation reveals the specific cause and what to do about it.

How does conversational AI research differ from a meeting transcript?#

Conversational AI research differs from a meeting transcript because it is a deliberate, neutral inquiry into a specific question, with follow-ups, rather than a recording of whatever happened to be discussed. A meeting summary captures what was said in context; a conversational interview elicits the reasoning behind a product or experience decision and probes the vague parts, producing research-grade insight at survey scale.

Conclusion#

Zoom's AI strategy is one of the clearest examples in enterprise software of a category leader betting that conversation, not the static interface, is where value lives — a $4.67 billion bet expressed through AI Companion 3.0, an agentic Virtual Agent, and a CEO who answers "AI, AI, and AI" when asked about investment priorities. The unfinished part of that bet is internal: the company that hosts 3.5 trillion minutes of customer conversation a year still leans on post-call surveys and NPS to learn what to build next, even as those response rates collapse. The lesson generalizes well beyond Zoom. If conversations beat forms inside the product, they beat forms in customer research too. Teams that want to close that gap can replace the survey layer with conversational AI interviews that ask, listen, and follow up at scale — start a conversational research study and see what your customers will tell you when they are not forced into a dropdown.