We interviewed 100 chief marketing officers at B2B companies about how they use AI. 90% use it every day. 88% use it for competitive analysis and 73% for vendor or market research. And when they start looking for a new vendor, more of them now open an AI chat (22%) than a search engine (16%).
Most of these leaders use AI more than they did a year ago, and research has become one of the main jobs they hand to it. Below is what that looks like in practice, which models they pick for which jobs, and where AI for research still runs out.
What is AI for research?
AI for research means using large language models and other AI tools to gather, summarize and interpret information. Most of it today is secondary research: reading competitor sites, analyst reports, call transcripts and dashboards faster than a person could. A smaller group of AI research tools run primary research, such as AI-moderated interviews that collect new answers from real people. AI market research applies both to the questions marketers care about most: competitors, categories and customers.
How CMOs use AI for research today
Content creation still leads, with summarizing close behind. The next three tasks are research. Competitive analysis (88%), data interpretation (81%) and vendor or market research (73%) all sit within 20 points of writing copy. Work that used to go to an analyst or an agency now goes to a model first.
The use is getting deeper, too. 85% of CMOs use AI more than they did 6 to 12 months ago, and 81% have added new use cases in that time. 47% have moved past one-off prompts into structured workflows and custom agents, such as a competitive report that writes itself every week.
"CMOs were testing prompts a year ago. Now they're building whole systems, optimizing LLM visibility and hiring for AI fluency," said Alfred Wahlforss, CEO of Listen Labs.
Which AI is best for market research? CMOs use more than one
Most CMOs run a small stack of models and pick one by task.
Model | What CMOs prefer it for | Share of CMOs |
|---|---|---|
Claude | Long-form writing and strategy | 44% |
ChatGPT | Content and copy | 29% |
Gemini | Work inside Google Workspace and analytics | 22% |
Copilot | Work inside Microsoft 365 | 12% |
So the answer to "what's the best AI for market research" depends on which part of the research you mean. CMOs reach for the model with the strongest output when the work is thinking-heavy, and for the assistant built into their office suite when the work has to live there.
"Different models are earning different roles inside the organization, and that specialization is starting to influence how teams create content, evaluate vendors, and make purchasing decisions," said James Cadwallader, co-founder and CEO of Profound.
AI is now the second stop in B2B vendor research
We asked CMOs where they start when they evaluate a new tool or service. Each picked one starting point (99 respondents; shares are rounded).
Starting point | Share of CMOs |
|---|---|
Peer recommendations | 49% |
An AI chat assistant | 22% |
A search engine | 16% |
Industry publications and analyst reports | 7% |
Other | 5% |
Peer recommendations lead by a wide margin. An AI chat comes second, ahead of search engines, and analyst reports trail both.
The reasons for switching are practical. 48% cite speed and 42% cite synthesis: an AI answer pulls ten pages into one paragraph, while a results page hands you the ten pages. Most haven't dropped search. 65% use the two together, and 27% keep search around specifically to check sources. The common pattern is a first take from the AI, then search to verify the claims and follow the links.
For B2B companies, that means more than one in five CMOs will first hear about your category, your competitors and your product from an AI chat, before they reach your website or your sales team.
Where AI research tools stop
The AI tools for research that CMOs use most all work on information that already exists. A model can read competitor sites, reports and transcripts faster than any analyst, compare them and summarize what they say.
What a model can't do is know things nobody has written down. It can't tell you how buyers will react to a product they haven't seen, why a segment churned last quarter, or what a customer means when they say a price feels high. Those answers live with your customers, so someone has to ask them.
The CMOs we interviewed seem to work this way already. Two-thirds use search alongside AI, more than a quarter keep search around to check sources, and peer recommendations still outrank every other starting point.
How to use AI for research: a practical workflow
If you're starting to use AI for research, or want to get more out of it, this is the pattern the CMOs in our study follow.
Start with the decision. Write down what you'll do differently depending on the answer. It keeps the model's summary pointed at something useful.
Let AI do the first pass. Competitor positioning, pricing pages, analyst reports and past call notes are where 88% of CMOs already use it for competitive analysis.
Check anything you'll repeat. Verify claims and follow the source links before a finding goes into a deck. That's why 27% keep search in the loop.
Automate what recurs. A weekly competitive summary or a monthly report is where 47% of CMOs have moved from prompts to workflows and custom agents.
Ask people for the why. When the question is about what customers think or will do, run primary research. AI-powered market research, such as AI-moderated interviews, makes that fast enough to run alongside everything else.
AI is becoming a requirement
Asked how AI will change marketing leadership over the next one to two years, 82% of CMOs pointed to productivity and 47% to creative work. Two smaller groups described bigger changes. 32% expect smaller or restructured marketing teams, and 28% say AI fluency is becoming a requirement for new hires.
Productivity was the most common answer. Smaller teams and AI fluency requirements came up less often, but they're the two changes that affect headcount and hiring.
What to do with this
Run a stack of models. Pick the model by task, the way most CMOs already do.
Verify before you repeat. 65% of CMOs use AI and search together. Treat the AI's answer as a first draft.
Find out what AI says about you. With 22% of CMOs starting vendor research in an AI chat, check how the major assistants describe your company and category every month.
Hire for AI fluency now. 28% of your peers already screen for it.
Methodology
Listen Labs ran AI-moderated interviews with 100 chief marketing officers at B2B companies, all employed full time and responsible for marketing strategy and budget. 48% work in tech or software, and the rest in financial services, healthcare, professional services and other industries. Listen Labs and Profound published the findings on July 15, 2026. Percentages are the share of respondents who answered each question, on bases of 96 to 100. Open-ended answers were coded into themes, and a respondent can fall into more than one theme, so percentages don't add to 100. No individual respondents are quoted.