Key takeaways
- Intercom shows no storage figure, so you count conversations: a teammate with Can access Reports reads New conversations in a Conversations report (one year at a time, two years back), and the Conversations search API's total_count covers older history.
- An admin starts the export at Settings > Data > Imports & exports > Export data, sending JSON or JSONL to an Amazon S3 or Google Cloud Storage bucket, and confirms it from a verification email.
- The historical export, the reports and the dataset CSV all reach back two years; the dataset CSV holds figures only, with no message text, and deleting a user or lead also deletes their conversations.
- You don't need to export anything to apply: have your conversation count, years of history and teammate seats ready, and the data vendor anonymizes the records before any AI lab sees them.
01
What is Intercom data, and why do AI labs and data vendors buy it?
The short answer
Intercom data is the record of customer conversations a company has in Intercom, its customer messaging and support software: each conversation holds the customer's messages, the replies from the company's AI agent and teammates, private notes and the tags and ratings added along the way. AI labs (companies that build and train AI models) and data vendors (companies that anonymize business records and package them for those labs) buy a copy because a conversation shows a real customer problem being handled, including the moment a bot hands the problem to a person.
What an Intercom workspace records
Intercom puts every channel a customer might use, such as the chat window on a website, email and social messages, into one inbox. Intercom calls the whole exchange a conversation, and a conversation is made of parts, one for each message, note or reply. The company's staff who answer customers are called teammates.
- Conversation
- One thread with one customer, holding every message, reply and note in order.
- Fin AI Agent
- Intercom's own AI agent, which answers customers first and can pass the conversation to a teammate.
- Contact
- A person Intercom knows about: a user (a customer with an account) or a lead (a visitor who has not signed up).
- Attribute, tag and rating
- The details attached to a conversation, such as the kind of customer, the topic label a teammate added and the satisfaction score the customer gave.
Why an Intercom conversation teaches an AI model something
An AI model learns by studying a very large number of examples and finding the patterns in them, and those examples are called training data. A conversation is a complete example of customer service, because it shows the question, the answer and whether the customer was satisfied at the end.
The hand-off from a bot to a human is the part labs find most useful. It shows exactly where a scripted or automated answer fell short and what the person did next, which is what an AI agent, software that carries out a job step by step, needs to learn to do better. The full mechanism is in AI training data: why AI companies want your company's data.
How Intercom fits into a dataset
Intercom is usually one of several apps in a company's dataset, next to email, customer records and documents. Companies are paid $100K to $2M or more per dataset, according to micro1, the data vendor we typically refer companies to. The price covers the whole dataset, not each app on its own, and what your data is worth explains what moves it.
02
Where does Intercom show how much data you have?
The short answer
Intercom shows no storage figure, so the measure is a count of conversations. A teammate with the Can access Reports permission opens a Conversations report, which looks back up to two years and shows one year at a time. For anything older, a developer reads the total from the Conversations search API.
How to count conversations in Reports
Intercom's Reports section includes a ready-made Conversations template, described in its conversations reporting guide. The steps below read the number of new conversations from it.
- Confirm that your teammate role includes the Can access Reports permission, and that it also includes Can create, edit and internally share reports if you need to build the report yourself.
- In Reports, start a new report from the Conversations template.
- Read the New conversations number for the dates you chose.
- A single report covers up to one year, so run it once for each year and add the results together.
How to count older conversations with the API
Reporting data reaches back two years, although Intercom says the conversations themselves are kept separately and are not lost. For the older history, a developer calls the search conversations endpoint, POST /conversations/search, with a filter on the created_at date. The reply carries a total_count value, which is the number of conversations that match.
The developer needs an access token from a developer app created in the workspace.
Which plan and permission you need
| Method | Plan | Permission |
|---|---|---|
| Conversations report | Essential, Advanced or Expert | Can access Reports |
| Custom report | Advanced or Expert | Can create, edit and internally share reports |
| Search conversations API | Any workspace with a developer app | An access token from that app |
03
What should you know about your Intercom data before you apply?
The short answer
Have three facts ready: how many conversations your workspace holds, how many years it covers and how many teammate seats you pay for. They go into the data manifest, the written summary of which apps hold your records, how many years each covers and roughly how much. Applying shares no data, and you don't need to run an export first.
The conversation count
The conversation is the right unit, because one conversation already holds every message, note and reply in its thread and counting messages would inflate the number. Give the total from your reports or the API and say which one it is.
The years of history
Use the year of your oldest conversation. Reports reach back only two years, so a developer may need to run the search conversations call sorted by created date to confirm older history, and the date of your first Intercom invoice is a good backstop.
The number of seats
A seat is a teammate license you pay for, because Intercom prices by seat. The users, leads and companies in your contact list are customer records and do not count.
What you don't need to do first
You don't need to export anything to apply, and you don't need to clean or scrub conversations. The data vendor anonymizes the records before anyone else sees them, and nothing moves until a lawyer-reviewed license is signed. The application, the intro call and each later step are laid out in how to sell your company's data.
04
How do you export your full Intercom history?
The short answer
An Intercom admin goes to Settings > Data > Imports & exports > Export data, starts a New data export to an Amazon S3 or Google Cloud Storage bucket and turns on the historical export. That is the most complete built-in route, because it keeps message text and can include attachments, but the historical export covers two years of conversations. Older history comes through the Conversations API.
Which export to use
Intercom offers several exports, and only some of them contain what was actually said. Intercom's guide to exporting conversations data lists them.
- Cloud storage export
- Writes the full conversations, with message text and optional attachments, to a bucket you own. This is the one to use.
- Dataset CSV
- A spreadsheet of figures about each conversation, such as IDs, timestamps and metrics. It has no message text.
- Inbox transcript
- One conversation at a time as text or PDF, with internal notes left out. Intercom also allows 200 to 300 conversations per bulk download from the inbox.
- Conversations API
- Lets a developer pull full conversations in bulk, including those older than two years.
The steps for the cloud storage export
A bucket is a storage folder in the cloud that you control. You need one set up before you start, and Intercom's cloud storage export guide links the instructions for creating it.
- Sign in as an Intercom admin.
- Open Settings > Data > Imports & exports > Export data and choose New data export.
- Choose JSON or JSONL as the output format, which are both text formats that hold each conversation in full, and switch on Include attachments.
- Under Destination, choose an Amazon S3 or Google Cloud Storage bucket and enter its details, then select Test to confirm Intercom can write to it.
- Under Schedule, turn on Export historical data for the last two years and choose Hourly or Daily for the ongoing files that follow.
- Click Start data export. Intercom emails a verification message to the admin, and the export begins once that admin clicks Confirm.
What the files look like
Each file holds up to 200 conversations, and the filenames begin with the start and end time of the batch. Attachments go into folders named for the conversation and the message they belong to. Intercom can only write to your bucket and cannot read or delete anything in it, and one bucket cannot receive exports from more than one workspace.
What the other exports leave out
The dataset CSV, found under Reports > Dataset export, needs the Can access Chart drill-in and Can export CSV permissions. It holds conversation IDs, timestamps, attributes and metrics and no message text, covers up to two years, and arrives by email when it has more than 10,000 rows. An inbox transcript leaves out internal notes.
For the people behind the conversations, open your contacts, select the users, and choose More > Export users. Intercom emails a CSV link that expires after one hour, and its guide to exporting users, leads or company data describes the options.
Export before you delete anyone
Deleting a user or a lead also deletes their conversations. Archiving is gentler, because Intercom suspends the conversations and restores them if the person returns, but it is wise to export first before any cleanup of the contact list.
Sources: Intercom Help, Conversations reporting, Export conversations data to cloud storage, Export your conversations data and How to export users, leads or company data, and the Search conversations API reference. Steps checked against these pages in November 2026.
05
What stays private when Intercom data is licensed?
The short answer
The company chooses which inboxes and teams are in scope, and the data vendor replaces names and other personal details with consistent stand-ins before an AI lab sees anything. The company keeps ownership of its Intercom data, and the license covers defined records for a defined use.
You choose what is in scope
An inbox that handles internal or sensitive matters can stay out of the dataset entirely. A company can license the conversations of its customer support team and keep the rest. The data vendor also deletes records too sensitive to clean safely instead of anonymizing them.
Names are replaced before a lab sees them
Anonymization means replacing every detail that identifies a person or a company, such as a name or an email address, with a stand-in that is the same every time it appears. The conversation still reads in order, so a lab can learn from it, but it no longer says who the people are. The diagram below shows one conversation before and after.
The method, including what is deleted outright and how you review samples, is in how company data is anonymized for AI training.
Ownership and payment
Licensing gives the buyer limited rights to a copy, while the company keeps ownership of the originals. The terms that matter, such as exclusivity, permitted use and deletion, are covered in data licensing. micro1 pays in full within 30 days of accepting the data, usually as a single payment.
Questions
Frequently asked questions
Does the Intercom CSV export include the text of our messages?
No. The dataset CSV under Reports holds only figures about each conversation, such as IDs, timestamps, attributes and metrics. To get what was actually said, use the cloud storage export, or have a developer pull conversations through the Conversations API.
How far back can we export Intercom conversations?
Two years, for the reports, the dataset CSV and the historical cloud storage export alike. For older conversations, a developer uses the Conversations API.
Can we export Intercom attachments?
Yes, through the cloud storage export. Turn on the attachments option when you set it up, and Intercom writes the files into an attachments folder organized by conversation and message.
Who can set up an Intercom data export?
An Intercom admin. After you click Start data export, Intercom emails a verification message, and the export begins only once the admin confirms it.
Do we need our own cloud storage to export from Intercom?
For the cloud storage export, yes: Intercom writes to an Amazon S3 or Google Cloud Storage bucket that you own. If you have none, the Conversations API and the inbox downloads are the other routes.
Do we need to export Intercom before we apply?
No. The application takes about two minutes and shares no data. You only need your approximate conversation count, your years of history and your number of teammate seats for the intro call. The full export is needed later, when the records are transferred after you sign the license.
What is Intercom data worth to AI labs and data vendors?
Intercom history is priced as part of a whole dataset rather than on its own. Companies are paid $100K to $2M or more per dataset, according to micro1, and the price depends on the years of history and how completely the conversations show each problem being solved. What your data is worth explains the factors.
Bottom line
Know what is in Intercom, then see what it is worth
You don't need to export anything to apply. Apply in two minutes, describe your records on the intro call, and micro1, the data vendor we typically refer companies to, prices the whole dataset and sends an offer 1–2 days later.
Not legal, tax or financial advice. Figures come from micro1's published pages and public reporting as of November 2026.