dataconto.com

# Dataconto

> Dataconto shows owners of company records (email, chat, tickets, docs, databases) and owners of private code which AI buyers would license what they have, checked against each buyer's published terms. The fit check is free and shows the result before asking for contact details. Owners pay nothing. Some buyers pay Dataconto a referral fee when an owner we sent closes a deal, and that fee plays no part in fit check results. Dataconto doesn't hold, broker or price the data.

Last checked: 2026-10-21. Every buyer page shows the date we last read that buyer's terms and links to its sources.

## Who it's for
- Owners of companies that are operating, winding down or closed, with business records: Slack and other chat, email, support tickets, docs, CRM and databases.
- Owners of private code repositories, with or without git history.

## Buyers we list (terms checked 2026-10-20)
- [micro1](https://dataconto.com/buyers/micro1): data partnership for companies with 30+ staff licensing SOPs, knowledge bases, CRM and support records, in published tiers of $100K to $2M+. A second program takes companies that have closed.
- [Scale AI](https://dataconto.com/buyers/scale): Data Partnerships license de-identified records from a company's business systems. 40+ employees preferred; Scale's illustrative value is $10K to $1M+.
- [Mode](https://dataconto.com/buyers/mode): buys filtered email, chat, CRM and billing records from companies in the US, Canada, UK and Australia. Mode says they could be worth $100K to $5M.
- [Grepped](https://dataconto.com/buyers/grepped): licenses de-identified CRM, support, contract, invoice and project records to AI labs. Grepped says a company's data could be worth $20K to $5M, with a private estimate in about a minute.
- [Maryah](https://dataconto.com/buyers/maryah): licenses Slack, email, CRM, tickets and call transcripts, with no published size or country limit. Paid by revenue share or one payment.
- [Pangea](https://dataconto.com/buyers/pangea): licenses private source code and git history to AI labs at $100 per million tokens of code and $40 per million tokens of history. Non-exclusive, so the owner keeps the code.
- [All buyers](https://dataconto.com/buyers): what each takes, who qualifies and what it publishes about pay.

## Tools
- [Company fit check](https://dataconto.com/company): 8 questions for an operating company, 10 for one that is closing or closed, checked against four buyers' published terms.
- [Code fit check](https://dataconto.com/code): 6 questions, checked against Pangea's published terms.
- [Company worth estimator](https://dataconto.com/company/worth-estimator): runs micro1's, Telegraph's and Grepped's own published estimators side by side from one set of answers (team size, years in business, headquarters, optional revenue), and lists the other buyers that fit. No Dataconto figure.
- [Worth calculator](https://dataconto.com/code/worth-calculator): prices private repositories at Pangea's published rates. Connect GitHub, GitLab or Bitbucket, or pick a folder. The count runs in the browser, and we never see the code.
- [Offer check](https://dataconto.com/company/offer-check): an owner who got an offer for the company's records sees it next to that buyer's own estimator or published range, the three published estimators for the same company, and who else takes records like theirs. Runs in the browser; nothing typed is sent.
- [Records summary](https://dataconto.com/company/records-summary): the company fit check's questions on one form, turned into a short text in the order AI companies' intake forms ask (status, headquarters, headcount, industry, years of records, systems, language, access, who can sign, sensitive records), plus which listed companies fit and why, with each program's check date. Copy it or download it as .txt or .md. Runs in the browser; nothing is sent. The fit check's result screen offers the same summary, for code too.
- [Rights check](https://dataconto.com/company/rights-check): five questions (company status, who can sign, customer contracts, personal data, staff and contractor IP) plus the tools the company uses give a verdict (ready to license, ready with some work, not yet) and the steps to settle before signing, each linked to its source, with each tool's own terms on ownership, export and AI training. Runs in the browser; nothing is sent.
- [Deletion clock](https://dataconto.com/company/deletion-clock): the last day to export from each of 20 business tools once the account ends, by plan, from each vendor's own terms, with a calendar file. Runs in the browser.

## Short answers (checked 2026-10-26)
Each answer links the page that holds its sources and dates.
- Q: Who buys a company's Slack, email and support records? A: AI companies and data buyers that license business records. Dataconto lists micro1, Mode, Scale, Turing, Telegraph, Grepped, Miro Advisory, License My Data and Maryah, each with its published criteria. https://dataconto.com/buyers
- Q: Can a closed company still sell its data? A: Yes, if the records still exist and someone can sign. micro1's Defunct Company Data Partnership, Turing and License My Data take records from closed companies. https://dataconto.com/guides/sell-company-data-after-shutdown
- Q: How much is a company's data worth to AI buyers? A: micro1, Telegraph and Grepped publish estimators. For a 60-person US company with 8 years of records they give $286K to $599K, $290K to $605K and $31.5K to $45K. None is an offer. https://dataconto.com/company/worth-estimator
- Q: Has anyone published what it paid a company? A: Telegraph lists three deals: $350,000 to a 60-person accounting firm, $450,000 to a 70-person engineering firm and $650,000 to a 100-person freight broker. https://dataconto.com/buyers/telegraph
- Q: Is my offer for company data fair? A: Compare it with the buyer's own estimator or published range and with other buyers' estimates for the same company. https://dataconto.com/company/offer-check
- Q: How much does a private code repository sell for? A: Pangea's published list rates are $100 per million tokens of source code and $40 per million tokens of git history, non-exclusive. https://dataconto.com/code/worth-calculator
- Q: How long does Slack keep data after you cancel? A: On paid plans, 30 days to export after the subscription ends. On the free plan, export before the account ends. https://dataconto.com/company/deletion-clock
- Q: How long does Google Workspace keep data after cancellation? A: Its data processing addendum allows up to a 30-day recovery period, then deletion within 180 days, with no export after the account ends. https://dataconto.com/open-data/tools/google-workspace
- Q: Does selling company data cost the owner anything at Dataconto? A: No. Owners pay nothing; some buyers pay Dataconto a referral fee when a deal closes. https://dataconto.com/disclosure
- Q: Does Dataconto contact owners first? A: No. Owners find Dataconto through search and run the fit check themselves. https://dataconto.com/company
- Q: Can my company license its records to AI companies? A: If someone with authority signs, customer contracts allow it, personal data is handled and the company owns its staff's work. Buyers such as Turing and Maryah ask you to confirm the right to license. https://dataconto.com/company/rights-check
- Q: Who can sign a data license for a company in administration? A: Usually the administrator, trustee or assignee, not the founders. https://dataconto.com/guides/company-in-administration-can-its-data-be-sold
- Q: Do business tools train AI on customer data? A: It differs by tool and plan; each vendor's own words are in the tool-terms dataset. https://dataconto.com/open-data/tools

## How the results are built
- Each fit check rule matches one criterion a buyer publishes. A buyer appears in a result only if its published terms fit the answers.
- Figures on the site are the buyer's own, quoted with a source and the date we checked them. We don't estimate what a company's records are worth.
- The worth calculator uses pre.dev's published counting script, which prices repos for Pangea, and Pangea's list rates.

## Guides
- [Sell company data after shutdown](https://dataconto.com/guides/sell-company-data-after-shutdown): which buyers take closed companies, and what to export before you cancel.
- [Can a company in administration or bankruptcy sell its data?](https://dataconto.com/guides/company-in-administration-can-its-data-be-sold): who has authority to sign, and why customer personal data stays out.
- [Acqui-hired? What happens to the old company's records](https://dataconto.com/guides/acquihired-what-happens-to-the-records): who owns the Slack, tickets, docs and code after an acqui-hire.
- [Sell source code to AI companies](https://dataconto.com/guides/sell-source-code): what Pangea pays per token, what rules a repo out, how a sale runs.
- [Selling your SaaS on Flippa? Sell the code separately](https://dataconto.com/guides/selling-your-saas-on-flippa-sell-the-code-separately): a code license can run next to a marketplace sale.

## Records by type
- [Records by type](https://dataconto.com/data): which listed AI companies take each kind of record, how to export it and when it gets deleted.
- [Slack and Teams chat](https://dataconto.com/data/slack-and-teams): export by Slack plan, free-plan deletion, Teams Export APIs.
- [Business email](https://dataconto.com/data/email): Google Data Export and Vault, Microsoft Purview, deadlines after cancellation.
- [Support tickets](https://dataconto.com/data/support-tickets): Zendesk, Intercom and Freshdesk exports with the full ticket text.
- [Jira and Linear tickets](https://dataconto.com/data/jira-and-linear): Jira site backups, Linear CSV and API exports.
- [CRM and sales records](https://dataconto.com/data/crm): Salesforce Data Export Service and HubSpot exports.
- [Call recordings and transcripts](https://dataconto.com/data/call-recordings): Zoom and Gong exports, and all-party consent.
- [Documents, wikis and SOPs](https://dataconto.com/data/documents-and-sops): Notion, Confluence and Google Drive exports.

## Industries
Each page quotes the listed AI companies that name the industry on their own pages, word for word, with the date read (2026-10-27), plus size floors and the records that matter.
- [Industries](https://dataconto.com/industries): which listed company names which industry, in one grid. micro1 and Telegraph also take any industry.
- [Accounting and finance](https://dataconto.com/industries/accounting-and-finance): named by Telegraph, micro1, Mode, Scale, Turing and Miro Advisory. Telegraph says it paid a 60-person accounting firm $350K. Mode buys from accounting and finance firms with 10 or more employees.
- [Law firms and legal](https://dataconto.com/industries/legal): named by Mode, micro1 and Scale. Mode buys from legal firms with 6 or more employees. Telegraph says privileged information needs extra review and may be excluded.
- [Healthcare](https://dataconto.com/industries/healthcare): named by Mode, Scale, Turing, Telegraph and Miro Advisory. Under HIPAA (45 CFR 164.508(a)(4)) selling protected health information needs each patient's authorization; de-identified information under 45 CFR 164.514 falls outside that rule. Mode buys from medical firms with 10 or more employees.
- [Software and SaaS](https://dataconto.com/industries/software-and-saas): named by Turing, micro1, Scale, Telegraph and Miro Advisory. Private code can be licensed on its own (Pangea: $100 per million tokens of source code, $40 per million tokens of git history).
- [Logistics and manufacturing](https://dataconto.com/industries/logistics-and-manufacturing): named by Telegraph, Scale, micro1 and Miro Advisory. Telegraph says it paid a 100-person freight broker $650K and a 70-person engineering firm $450K.
- [Agencies and professional services](https://dataconto.com/industries/professional-services-and-agencies): named by Scale and Telegraph; Pangea licenses an agency's own code but keeps client work out.
- [E-commerce and retail](https://dataconto.com/industries/ecommerce-and-retail): named by Scale, Telegraph, Turing and Miro Advisory. Telegraph lists Shopify among the systems it takes records from.

## Open data: who pays owners for company records and code
- [Buyer database](https://dataconto.com/open-data): 34 buyer programs (12 take private code), 148 dated facts quoted from the buyers' own pages: payout ranges, per-token rates, platform fees, payment timing, eligibility. Press-only figures are flagged. History is kept when a term changes. No introducer or partner terms. Updated 2026-10-21. CC BY 4.0.
- Downloads: https://dataconto.com/data/buyers.json (programs with current facts), /data/buyers.csv, /data/facts.jsonl (every fact, history included), /data/facts.csv, monthly snapshots under /data/buyers/.
- API, keyless: https://dataconto.com/api/buyers.json?side=code|company&kind=pays_owner|marketplace|invite_only&min_payout=100000 and https://dataconto.com/api/facts.json?program=<id>&field=payout_range. Spec: https://dataconto.com/openapi.json
- MCP server, keyless and read-only: https://dataconto.com/mcp (Streamable HTTP; registry name com.dataconto/buyers). Tools: list_buyers, get_buyer, get_market_facts, get_tool_terms, plus fit_check (which listed buyers' published criteria an owner's answers meet), estimate (micro1's, Telegraph's and Grepped's own estimators side by side, and where an offer sits against them), code_value (code priced at Pangea's published per-token rates with pre.dev's counter), deletion_clock (last day to export from each tool after the account ends) and rights_check (whether a company's records are ready to license, and the steps to settle first). Every result links to its Dataconto page. Docs: https://dataconto.com/open-data#mcp
- [SEC filings on AI data licensing](https://dataconto.com/open-data/sec-filings): 19 US public companies whose own SEC filings say they license data or content to AI companies (Reddit, Getty Images, Tempus AI, Wiley, GoPro and others), each with a verbatim quote, any amount the filing states, and a link to the filing on sec.gov. Downloads: /data/disclosures.json, /data/disclosures.csv. Updated 2026-10-22. CC BY 4.0.
- [What your tools' terms say about your data](https://dataconto.com/open-data/tools): 20 business tools (Slack, Microsoft Teams, Microsoft 365, Google Workspace, Zendesk, Intercom, Freshdesk, HubSpot, Salesforce, Notion, Atlassian, Linear, Asana, Zoom, Gong, Discord, Dropbox, GitHub, Sentry, Vercel), six questions each: who owns the customer's data, whether the vendor trains AI on it (generative and other models, by plan), aggregated use, what can be exported on which plan, how much history the plan keeps, and what happens after cancellation. 376 facts, each a verbatim quote from the vendor's own terms, AI page, DPA or help article with a link and the date read. "The terms don't say" only after every contract was read; readings of the text are marked inferred. One page per tool at /open-data/tools/<tool> with an anchor per fact. Updated 2026-10-23. CC BY 4.0.
- Downloads: https://dataconto.com/data/tools.json (tools, headline answers by plan, every fact), /data/tools.csv, /data/tool-facts.jsonl, /data/tool-facts.csv, monthly snapshots under /data/tools/, changed answers at /data/tool-changes.atom.
- API, keyless: https://dataconto.com/api/tools.json?tool=slack | ?record_type=slack-and-teams | ?field=ai_training&answer=opt_out | ?plan=Free. MCP tool: get_tool_terms (same filters).
- Suggested citation: Dataconto (2026). Dataconto open buyer database: who pays owners for company records and code. Version 2026-10-21. https://dataconto.com/open-data
- Attribution: "Data from Dataconto (https://dataconto.com/open-data), CC BY 4.0."

## About
- [How we get paid](https://dataconto.com/disclosure): buyers pay the referral fee, owners pay nothing.
- [About Dataconto](https://dataconto.com/about): who runs it and why.
- [Privacy](https://dataconto.com/privacy) · [Terms](https://dataconto.com/terms)
- Sister site: [githubworth](https://githubworth.com) prices the code in a GitHub profile at Pangea's rates.