A company's email archive is worth money to AI buyers, but it is almost always priced as part of a wider record with chat, files and tickets, not per message. Reported figures that included email run from "hundreds of thousands of dollars" for a software company's 13 years of records (Fortune, September 2026) to a $10 million auction bid for Spirit Airlines' archive, a sale not yet approved (Axios, August 2026). For a small or mid-size company, published buyer ranges for a full record start near $100,000 (micro1, October 2026).
On this page
- Why do AI labs want business email?
- How much is an email archive worth?
- What makes some email worth more than other email?
- How do I check how much email my company has?
- How do I export company email safely?
- What should I leave out of an email export?
- How is email de-identified?
- What should I do before I talk to buyers?
- Questions
- Sources
Why do AI labs want business email?
Email is where a lot of real work is written down: quotes sent and accepted, supplier back-and-forth, approvals, complaints and how they were resolved. AI companies building agents for office work need long examples of that, from start to finish.
micro1's revenue chief told Business Insider (October 2026) that if a company will not include its Slack, CRM or email, the rest of its records are worth less. Buyers that list email on their program pages include Polyshares, G2i, AfterQuery, Troveo, Grepped and SellMyData. micro1 asks for internal communications. G2i lists Outlook, Gmail and Google Calendar by name (G2i, October 2026).
Data Licensors, the company that publishes this page, also licenses company records, including email. Our buyer page sits next to the others.
How much is an email archive worth?
There is no public price per email. These reported deals and offers included email. Each is labeled by the kind of figure it is, and all are in our deal ledger.
| Company | Records named | Amount | Kind of figure | Source |
|---|---|---|---|---|
| Transcription software company, 13 years | Slack, Jira, email, Google Drive | "Hundreds of thousands of dollars" | Press report of a sale | Fortune, September 2026 |
| Founder-led company | Company data (the post says micro1 asks for Slack, Notion and email) | About $600,000 offer from micro1, declined | Founder post | X, September 2026 |
| Logistics and distribution business | Google Workspace (email, bills of lading), Slack, QuickBooks, warehouse systems listed as possible sources | $350,000 | Broker claim, not independently confirmed | Red Deer Data, September 2026 |
| Spirit Airlines, bankrupt, about 17,000 former staff | About 100M emails across 80,000 accounts, 500M Teams messages, files and code; customer records excluded | $10,000,000 winning auction bid; micro1 later offered $12.5M | Auction bid, not final: hearing set for November 6, 2026 | Axios, August 2026; The Next Web, August 2026; Fortune, September 2026; TravelPulse, October 2026 |
The airline is not a guide for a 40-person company. The useful comparison is the full-record range buyers publish. micro1 lists tiers of $100K to $500K, $500K to $2M and $2M+ (micro1, October 2026). See how much company data is worth for every buyer's range.
What makes some email worth more than other email?
| Worth more | Worth less |
|---|---|
| Long threads with customers and suppliers that end in a decision | Newsletters, receipts and marketing blasts |
| Quotes, change requests and approvals with attachments | Automated alerts and system notifications |
| Years of continuous history from the same team | A few months, or one person's mailbox |
| Mail that links to the same jobs in your chat, CRM or tickets | Mail with no context anywhere else |
| Your industry's own vocabulary (trades, freight, engineering) | Generic scheduling and calendar traffic |
A buyer can usually tell from a small sample which kind you have.
How do I check how much email my company has?
Google Workspace (Gmail). In the Admin console go to Reporting, User Reports, Accounts. With the Reports privilege you can see, per user, Total emails, Emails sent, Emails received and Gmail storage used (Google Workspace Admin Help). Add the storage column across users for a rough size, and note the oldest account.
Microsoft 365 (Outlook). In the Microsoft 365 admin center, open Reports, Usage, Exchange. The Mailbox usage report shows an Item count for each mailbox, which is the total stored, and storage used. The Email activity report shows send and receive counts over the last 7, 30, 90 or 180 days (Microsoft Learn). Archive mailboxes are not in the storage chart, so check them too.
Write down three numbers: total messages, number of mailboxes that were really used, and the year the oldest mail starts.
How do I export company email safely?
| Tool | Who runs it | What you get | Points to know |
|---|---|---|---|
| Google Data Export tool | Super admin with 2-Step Verification, account at least 30 days old | Gmail, Drive, Chat and more for the whole organization | All editions. Ready no earlier than 48 hours, typically about 72 hours, up to 14 days. A Google-provided bucket is deleted 60 days after the export starts |
| Google Vault | User with Vault Manage Matters, Manage Searches and Manage Exports privileges | Search by account and date, export Gmail as PST or mbox | Included in Business Plus, Enterprise Standard and Plus, Frontline and Education editions, not in Business Starter or Standard; ask Google about an add-on. Export files are kept 15 days |
| Microsoft Purview eDiscovery | eDiscovery user with an E3 or E5 license | Mailbox items as PST or single messages, plus SharePoint, OneDrive and Teams | Export packages expire after 14 days. Exports running over 7 days are canceled, so split by date |
| Outlook PST export | Each mailbox user, in Outlook for Windows | One mailbox to a .pst file | Classic Outlook exports mail, calendar and contacts; new Outlook (Settings, Files, Export) exports mail only. Good for a few mailboxes |
Sources: Google Workspace Admin Help, Google Vault Help, Microsoft Learn and Microsoft Support, checked October 2026.
Practical steps:
- Agree the scope first: which mailboxes, which years, which labels or folders.
- Export in date batches. Purview cancels any export that runs over 7 days, and smaller batches are easier to check.
- Keep the files on company storage you control, encrypted, and delete working copies when the delivery is done.
- Send only through the encrypted transfer the buyer gives you. At Data Licensors your team runs the export with our step-by-step guidance; we never ask for passwords.
What should I leave out of an email export?
- Mail with your lawyers. Sharing it could affect legal privilege. Leave it out or ask counsel.
- HR, payroll, benefits and medical mail. Health records are on our list of things we never accept.
- Personal mail of employees. A lawyer quoted by MIT Technology Review (October 2026) pointed to the private details that sit in email and calendars: doctor appointments, school pickups, date-night plans.
- Mailboxes of people who handle sensitive client matters, if contracts limit use of client information.
- Payment details, bank credentials and government ID numbers. We never accept them.
- Newsletters, spam and notifications. They add size, not value.
How is email de-identified?
Email has more identifying detail than most records, because it hides in headers, signatures and quoted replies. De-identification covers:
- From, To and CC addresses, which become role placeholders such as [Sales 2] or [Customer contact].
- Names in greetings, sign-offs and signatures, plus phone numbers, titles tied to one person and street addresses.
- Company and customer names in subject lines and bodies.
- Account, invoice and order numbers that point to a person or a company.
- Attachments, which are processed the same way or left out.
Quoted reply chains repeat the same text many times, so duplicates are removed before review. Thread order and dates stay, because the sequence is the useful part. You review a de-identified sample before you sign. See how de-identification works and, if you use Microsoft 365 for chat and files too, our Microsoft 365 guide.
What should I do before I talk to buyers?
Count your mail, decide the mailboxes and years you would include, and list what stays out. Then get a quick estimate, and ask more than one buyer to price the same scope in writing. Old mail that nobody reads is a common kind of dark data; it is worth checking before anyone deletes it.
