>>> PRESS_RELEASE
MFOUR ADDS ANOTHER 1M+ CHATGPT CONVERSATIONS LAST MONTH, EXTENDING FULL CONSUMER JOURNEY VISIBILITY|||MFOUR ADDS ANOTHER 1M+ CHATGPT CONVERSATIONS LAST MONTH, EXTENDING FULL CONSUMER JOURNEY VISIBILITY|||
ChatGPT Conversation Data

See what your consumers ask AI — and what AI recommends.

Real opt-in ChatGPT conversations, joined to verified first-party demographics. Query it with DANI, get the full report, or license the raw dataset.

100% opt-inAccount-level · every deviceJoined on PANELIST_IDPII tokenized
ChatGPT conversation
Best cordless vacuum under $300?A few strong picks: Dyson V11, Shark Stratos
panelist_id · 6680048
Identity recordKnown consumer
Verified
GenderFemaleAge25–34HHI$90kTierPlusDMAToledo
first-party · Fair Trade Data® · PII tokenized
Three ways in · one connected dataset

Access your way.

Real MFour panelists’ opt-in ChatGPT conversations, paired with direct survey. Query it yourself, get the full report, or license the raw dataset — whichever fits how you work.

Self-serve

Query it yourself

Ask DANI in plain English and get answers from the live dataset in seconds.

What do shoppers ask AI before choosing us?
Reports

Get the full report

Boardroom-ready analysis on the questions that matter, delivered to you.

License

License the dataset

Take the raw first-party conversations into your own stack, models, and BI.

panelist_id  6680048
prompt  "best cordless…"
brands[]  Dyson, Shark
Inside the data

The who and the what behind every AI conversation.

Connect what your consumers ask AI to who they are. The demographic truth and the full conversation — the inputs brands need to inform product, category, and messaging strategy.

Who’s asking

Demographics

Anonymous chat data is just noise. Every conversation joins to first-party demographics on PANELIST_ID — so you know exactly who is asking.

Membership tier
FreePlusProTeamEnt
Age, gender & income on PANELIST_ID
18-2425-3435-4445-5455+
Age · gender · incomeGeographyPlan tierUsage depth
The full record

Actual chat data

Every conversation preserved as a complete thread — reconstruct any session exactly as it happened, both sides intact. Captured at the account level, so you see every conversation from every device — desktop or mobile, any OS.

Best running shoes for flat feet?For overpronation, look at stability shoes — the Brooks Adrenaline GTS…
model gpt-6osession 3 · 2citations 2image 1
Full threadsEvery device & OSModel & sessionCitationsMultimodal
Fair Trade Data®

Scraped data can vanish overnight. Consented data can’t.

Most AI conversation data on the market is scraped — harvested without permission, in a legal gray zone the platforms are actively shutting down. The day the source gets blocked or challenged, the data — and everything you built on it — goes dark. MFour’s data is collected the opposite way.

The industry default
Scraping
Harvested without the consumer’s permission
Breaches the platform’s terms of service
No verified identity — who said it is a guess
One policy change or cease-and-desist from disappearing
When the source pulls access, your data — and every model and report built on it — goes dark with it.
The MFour way
Fair Trade Data®
Panelists opt in and are paid fairly for what they share
Full disclosure and consent — CCPA-compliant, PII stripped
Permissioned at the source — it can’t be pulled out from under you
Connected to verified, first-party demographics
Consented, compliant, and defensible — data you can build on and still have next year.
Ask how it’s collectedBefore you license AI data, ask how it was collected. If no one can tell you, it’s living on borrowed time.
Conversations + survey

Conversations show what they ask AI. Survey tells you why it matters.

Chat data is the behavior — the questions, the brands, the moment. But it can’t tell you what consumers did with the answer, how much they trusted it, or where it sat in the journey. Pair the conversation dataset with direct consumer survey on the same panelist and you close the loop: from what they asked to why it matters, how it influenced the decision, and the role AI is playing.

The behavior
In the conversation
Best running shoes for flat feet? ChatGPT recommended Brooks, Hoka & ASICS.
We see exactly what was asked, what AI recommended, and who asked it — but not what happened next.
The meaning
We ask them directly
When ChatGPT recommended those shoes, how much did it shape your final choice?
It decided my purchase
Strong influence
No real influence
The complete pictureWhat they asked AI + what they did with the answer = the full role AI is playing in the decision. One consumer, both halves of the story.
What the survey answers

Six questions chat data can’t answer on its own.

A research framework purpose-built to read the meaning behind every AI conversation — mapped to the moments that move your category.

GoogleRedditReviewsChatGPT

What is AI replacing?

Google, Reddit, reviews, support or social — and at which step.

AI
Search
Reviews
People

How much do they trust it?

AI recommendations vs. search, reviews, and people they know.

Where does it move the funnel?

Awareness, consideration, purchase intent, and the final decision.

priceis it worth it?fitalternativeswhat’s best

What’s driving the questions?

The unmet needs, frustrations, and motivations behind the prompt.

Who acts on AI most?

Which demographics convert — and how it shifts by category.

How is perception shifting?

Brand & product perception as AI usage climbs, wave over wave.

Templated framework

Run it on repeat — not once.

The conversation-plus-survey study ships as a standardized template. Field it once for a baseline, then re-field on a cadence you set. AI adoption is moving every quarter — a one-off read is stale before the deck is done. A recurring program turns it into a trend line you own.

Benchmarksa standardized baseline you measure every brand and category against
Trends over timethe same questions, wave over wave, so shifts are real and comparable
A recurring programquarterly waves, not a one-off, so insight compounds instead of expiring
AI influence index · trailing 4 waves
Q1
Q2
Q3
Q4
Share of category decisions AI influenced+19 pts
Anatomy of a record

Every message, structured and privacy-safe.

Delivered row-per-message with a consistent schema. Personal identifiers are stripped and replaced with typed tokens before delivery.

chatgpt_messages · one row
PII tokenized
PANELIST_ID6680048
MEMBERSHIP_INFOPlus
MESSAGE_ROLEuser
MODELgpt-6o
CONVERSATION_TITLECordless vacuum picks
MESSAGE_TEXTDoes <ORG> ship to <LOCATION>?
SESSION_SORT · SORT3 · 2
CREATED_AT_UTC2026-02-04T21:51Z
Field legend
PANELIST_IDJoin key to demographics & every other MFour data stream.
MEMBERSHIP_INFOPlan tier — Free, Plus, Pro, Team or Enterprise.
MESSAGE_ROLEuser, assistant or system — reconstruct both sides.
MODELWhich model produced the response.
<ORG> <LOCATION>Typed tokens replace names, places & PII in text.
New · AI ad capture

We capture the ads ChatGPT shows your consumers.

As OpenAI rolls out advertising inside ChatGPT, MFour’s opt-in panel sees exactly what real consumers are served — the advertiser, the creative, the prompt that triggered it, and who saw it. The first behavioral view of AI advertising.

Advertiser & creativethe brand, copy, and product promoted in every sponsored placement
Placement & triggerthe prompt and response context the ad appeared in
Who saw itdemographics & plan tier of every consumer served, on PANELIST_ID
Share of voiceyour AI-ad presence vs. competitors, tracked over time
ChatGPT
Ad captured
Best running shoes for flat feet?
For overpronation and flat arches, look for max-support stability shoes:
Sponsored
B
Brooks Running
Adrenaline GTS 23
Built for overpronators — GuideRails support keeps flat feet aligned mile after mile.
Shop the GTS 23
Data licensing

License the data vs querying the data.

Want the raw data instead of querying it through DANI? Take the full dataset into your own stack — delivered to your warehouse on a cadence you control, connected to OmniTraffic on PANELIST_ID, first-party and consent-backed.

PANELIST_IDThe same join key as every OmniTraffic stream — connect ChatGPT behavior to location, app, web and purchase for one validated consumer.
Delivery
Daily
go-forward + full history
Format
Parquet
Snowflake · BigQuery · S3 · API
Coverage
Full threads
prompts, responses & metadata
Capture
Account-level
desktop · mobile · any OS
Privacy
PII stripped
tokenized · consented · CCPA
Join key
PANELIST_ID
connects to all streams
Analyze with DANI

Ask the ChatGPT dataset anything.

Skip the warehouse. DANI runs analysis across every conversation in plain English — brand share-of-voice, prompt-intent clusters, membership cohorts and trends.

Brand share-of-voicePrompt-intent clusteringMembership cohortsTrend & forecastingAI citation tracking
Q
“Which cordless-vacuum brands do Plus & Pro users ask about most — and how has it shifted this quarter?”
Dyson leads at 28% of mentions, up 6 pts QoQ. Tineco is the fastest-growing among Plus users.
Dyson
Shark
Tineco
Bissell
Other

Get the ChatGPT data.

License the connected dataset for your own stack, or analyze it inside Studio with DANI. First-party, opt-in, and joined to one consumer.