Handshake AI vs micro1 vs Mercor: what each publishes about buying company data

The three side by side

Every cell below is the company’s own wording, taken from its row in the buyer index, and the table changes when the index does. Each name links to the full record with the exact quotes and the dates we read them.

Three things stand out. None of the three publishes its payout model, so you do not know from the outside whether you get one payment or several. None names the AI labs it sells to. And none publishes completed company data deals for these programs, so the ranges are what each company says it may pay, not what anyone has been paid.

Who each one is looking for

Handshake AI asks for “20+ full time employees & 3+ years in operation”. Its list of example sources is the broadest of the three and reaches well beyond chat and email: AutoCAD, Dropbox, Figma, Gmail, Google Drive, Outlook, Monday.com, NetSuite, Notion, Paychex, QuickBooks, ServiceTitan and Slack, “+1K more”. A company whose history sits in finance, payroll, design files or field-service records has the clearest published match here.

micro1 runs two tracks. Its Enterprise Data Partnership is for “operationally mature companies with 30+ employees, established documentation, and high-quality operational data such as SOPs, knowledge bases, CRM data, project histories, and internal processes”. It says it prioritizes US companies, “followed by other Western markets”, and that current demand is strongest in the United States, the United Kingdom and Canada. Its Defunct Company Data Partnership is for companies that are no longer operating, with “50+ peak employees” and substantial Microsoft Teams or Slack data, in a list of ten countries and in English.

Mercor publishes no minimum size and no geography. Its data page lists the tools you can connect (Slack, Microsoft Teams, Gmail, Outlook, Confluence, Notion, Salesforce, GitHub and Google Drive among “100+ supported tools”) and target industries from legal and healthcare to logistics. With no published minimum, it is the only one of the three a company under 20 people can approach without falling short of a stated rule.

What they say about money

micro1’s data page splits its range into three tiers: $100k+ for qualified partnerships, $500k+ for large-scale datasets or ongoing participation across several teams, and $1M+ for highly unique proprietary data. Handshake AI’s announcement of its program puts the typical range at “$100K to $4M depending on the data”.

Mercor is the only one that tells you, before you talk to anyone, when the money arrives and that nothing comes off it. Handshake AI ties payment to delivery milestones agreed in the contract, so the schedule depends on what you negotiate. micro1 publishes nothing about when companies are paid.

How the data leaves your systems

The route matters because platform terms treat an app pulling data through an API differently from your own administrator exporting it.

• Mercor works by connection: you link your accounts and “our platform extracts and de-identifies data into an asset”.
• Handshake AI says it offers “remote, in-person, and physical transfer options” and helps choose one based on where the data is stored and what you agreed to share.
• micro1 publishes nothing about ingestion.

Slack’s API terms bar the provider of an app offered outside its own organization from using API data to “train a large language model”, and Google’s Workspace developer policy prohibits using user data to train AI models beyond that user’s personalized model. We do not judge whether any buyer’s pipeline complies. Ask each one how its route fits these terms before you authorize a connection, and read the ingestion section of the offer checklist.

Who ends up with the data

Handshake AI says “We partner with all of the leading frontier AI labs” and that licensed data is used to train and evaluate private AI models, without naming the labs. Mercor says the data your business generates “is exactly what frontier labs need” and that it licenses, packages and pays for it. micro1 does not say. If it matters to you which lab receives your data, or whether it can be resold, put that question in writing before you sign.

Who gets paid for introducing you

All three pay people who refer companies. That is legal and common, and you should know it when someone suggests one of them.

• Handshake AI: $10K per referred company, paid when the deal closes.
• micro1: up to $50,000 per referred company on the Enterprise track, $75,000 listed in the data behind its defunct-company posting, and 7.5% of the license or purchase price where no amount is specified. Its referral terms let micro1 withhold or reduce any payout “in its sole and absolute discretion”.
• Mercor: 5% of the deal, up to $100,000, paid after the company’s first delivery, with terms shown in a referral portal after sign-in.

If an adviser, broker or friend points you to one of these three, ask whether they are paid for the introduction. The same rule applies to us: a buyer that pays us is marked on its row, and the disclosure page explains how we make money.

Which one fits your company

This is not a ranking. It matches situations to what each company publishes.

These three are not the whole market. The index lists 51 companies that buy, broker or list company data, and the buyers by company size guide shows which ones publish a minimum your company meets.

Before you talk to any of them

1. Ask whether the license is exclusive and for how long. An exclusive license to one buyer can rule out a second sale.
2. Get the payment trigger in writing: signing, delivery, acceptance or a milestone.
3. Ask who removes personal data, before or after it leaves your systems, and what happens to the data when the license ends.
4. Decide who runs the export: your administrator, or a connector you authorize.
5. Ask who, if anyone, is paid for the introduction.

If you already hold an offer from one of them, paste it for a free review. We check it against these terms and against what other buyers publish, and it never goes to a buyer.