micro1 review: what sellers actually get

micro1 wants established companies with 30 or more full-time staff, mature operations and several years of records, mostly in English. Its pages used to cap the range at 200 staff; the current partnership and referral pages set only the 30+ floor.

It is prioritising US companies, then other Western markets, and says demand is strongest for the US, UK and Canada. It sets no minimum file count or data volume.

The common thread in what it asks for is repeat work: the documents, tickets, threads and CRM activity that show how your team handles the same kinds of task again and again. That is exactly what AI agents need to learn a job.

micro1 now publishes three tiers: $100K to $500K for a qualified partnership, $500K to $2M for large datasets or ongoing participation across several teams, and $2M+ for highly unique data. Its estimator page headlines $200K to $2M+. These are micro1's own ranges, not audited figures, and real offers can land below them.

It prices on dataset size, workflow complexity, data quality, the expertise the records show and how unique they are. Business Insider reports that it usually wants a company's whole data corpus, and pays more for larger and more complex sets.

The company behind the budget is substantial. Forbes put its latest raise above $100M at a $4B valuation, and in October 2026 micro1 said it had passed a $1B annual run rate while staying profitable. During the Spirit Airlines bankruptcy it came in late with a $12.5M offer for the airline's records.

The contract sellers sign is public: an order form that points to micro1's Data License Master Terms. The version posted on the day you sign governs your deal, and later changes apply only if both sides sign an amendment. The license is non-exclusive, and it is also irrevocable, perpetual and sublicensable through multiple tiers: micro1 can pass the data on and you cannot take the permission back.

Some limits protect you. micro1 must de-identify personal data before any customer trains on it, may not resell your data as a standalone product, may not use it to compete with you or reverse-engineer your pricing, and may not name you as the source without consent. Raw records are deleted within 120 days of the final delivery; any model trained on them is kept.

Read the warranties closely. You promise the data has not already been licensed to someone else on the same or similar terms, that it holds no health records, payment card numbers or data on children, and that you have every consent needed. Each side's liability is capped at three times the fee, and disputes go to arbitration in California.

Introducers get paid too. micro1's referral page lists $50,000 for enterprise deals up to $700,000 and 7.5% above that, and $75,000 for closed-company deals up to $1M and 7.5% above that. The bonus is paid by wire only after the company's data is transferred and approved. Referrers cannot use micro1 email addresses or pose as its staff.

We are in this program. The fee comes from micro1, never out of your license price.

Paid social is how most owners first hear of micro1. A few quick questions, then a Facebook lead form, with no micro1 web page in between. In September and October 2026 Meta pulled some of these ads for breaking its ad standards. It did not say which claims were the problem.

So before you share anything, confirm the email comes from a micro1.ai, micro1.io or micro1.co address, the domains micro1 lists as official.

micro1 calls itself a data lab: it sells expert human data, training environments and agent testing to AI labs and enterprises. Company records feed that business, because micro1 turns de-identified records into simulated workplaces where AI agents practise real tasks. It started as a recruiting company, and its AI recruiter, Zara, now sources and screens the experts who do its project work. We found no separate hiring product for employers on its current site.

1. 1

   Profile and data manifest

   Your contact creates a company profile and lists the data you hold. micro1 then runs a short AI interview to judge fit.

2. 2

   Discovery and pricing

   micro1 identifies which datasets it can license and makes an offer based on their nature, scale and value.

3. 3

   Contract

   You sign an order form under its Data License Master Terms, including its promise to de-identify before passing data on.

4. 4

   Delivery

   You transfer the agreed data through its APIs or another approved method. Under its terms, micro1 bears the export and delivery costs.

5. 5

   Review

   micro1 checks the data for completeness and quality, with up to 30 days to approve or reject it.

6. 6

   Invoice and payment

   Once the data is approved you send an invoice, and micro1 says it pays within 30 days.

The best fit is an English-language company of 30 or more staff that has run the same kinds of work for years and kept the email, tickets and documents to show it. Software, finance, support-heavy firms, professional services and logistics all suit.

If that is you, get micro1's written offer and at least one other quote. Its license is non-exclusive, but its warranty about earlier licenses means you should agree the order of deals up front. The perpetual terms are the part to negotiate.

We run the same checks on every buyer we rank. They take an afternoon and they protect the one thing you cannot take back: the copy of your records.

• Find the legal entity. A registered company name, an address and at least one named person. A Facebook form alone is not enough.
• Get the offer and the contract in writing before any export. A sample agreement tells you more than a sales call.
• Ask who strips personal data, how, and when. You should approve the method and see a sample before anything leaves your systems.
• Ask where the data goes. Which AI companies receive it, whether you are told, and whether it can be resold.
• Ask how everyone is paid. By you, by the buyer, or out of your deal. Referral fees are normal; hidden ones are not.
• Get a second quote. Terms from at least one other buyer is the fastest way to know if an offer is fair.