Sell Agency DataGet a quote

AI data buyers compared for agencies

Four buyers publish terms an agency can check itself against: micro1, Scale AI, Nyne and Labelbox Prism. Nyne has the lowest size floor, Scale AI de-identifies inside your systems, and Labelbox Prism licenses are exclusive by default.

Every cell rechecked against the buyers' sites, 7 November 2026.

Sell Agency Data works on its own account: none of the four companies below has endorsed it, and it has no affiliation with any of them. Each cell paraphrases that buyer's page as of 7 November 2026. A blank on their side shows here as "Not published" rather than a guess.

Published terms, side by side
Termmicro1Scale AINyneLabelbox Prism
Published rangeThree advertised tiers per company: $100K–$500K, $500K–$3M, and over $3M for rare operating dataAn illustrative $10K–$1M+ per partnership, higher for more frequent deliveries; explicitly not a promise$100K or more described as common, varying with the kind and volume of recordsRoughly $50K at the small end and $1M+ for enterprise-scale data, paid as an upfront fee with revenue share on top
Who qualifiesAt least 30 full-time staff and English-language records; buys mostly from American, British and Canadian firms; no set number of years, though a longer record helpsPrefers 40 or more employees but judges each case; wants mature data systems, real workflows and one internal ownerUS firms with 20 or more full-time W-2 staff and at least three years trading; rights still checkedNo general size or age rule (for support logs it leans toward big consumer-facing firms)
Records named that agencies holdSOPs and process write-ups, operations and planning records, knowledge bases, CRM and sales records, bug trackersJira and Confluence records, files with their metadata, email, chat, calendars, CRM and HR systemsAsana and Jira project records, CRM, bookkeeping data such as invoices and estimates, calendarsNothing agency-specific; the nearest are cross-system interaction records and code edit traces
What they will not takeRecords mostly in other languages get lower priorityIdentities of staff or customers; credentials and raw system access; anything the contract does not coverWhatever categories you decide to hold backVoice or likeness without explicit, time-stamped consent
De-identificationDe-identified by micro1, and you may inspect sample outputs; location of that work unstatedStrips identities on your side before export, with your sign-off on every packageDe-identified ahead of release; the page does not say by whomLabelbox does intake, cleanup and PII removal; where is not stated
Who owns the data afterwardYou keep ownership of the underlying recordsNo explicit statement; data described as remaining under your governanceLeft to the license agreementThe seller stays in control of both data and rights
Payment timingPays the whole amount within a month of accepting the dataNot publishedNo timing given; scope, rights and acceptance terms are set before work startsUpfront fee on signing; Labelbox says payment can follow the first call within three weeks
ExclusivityPage describes no exclusive grantNot publishedNot publishedExclusive unless negotiated otherwise, shaped to scope and sensitivity
Referral program for introducersYes. $50,000 on deals to $700,000; 7.5% of value on larger onesYes. Up to $50,000, only when Scale signsNot publishedNot published

Disclosure: micro1 and Scale AI run referral programs, and any of these four could pay us after a deal we introduced is signed. Column order is just the order we opened the pages. We have not tested or audited these buyers; the comparison rests on what they publish.

How do the four buyers differ for an agency?

micro1

The broadest list of process records, and the one that reads most like an agency's shared drive: SOPs, process documentation, product planning, operations records, internal knowledge bases. Its tiers run from $100K to $3M+, and the agency stays the owner of whatever it licenses. The size rule is 30+ full-time employees, with data mainly in English.

Scale AI

Strongest on privacy mechanics. Identities are stripped on the agency's own systems before export, you approve every package, and Scale refuses staff or customer identities, credentials and raw system access. Its $10K to $1M+ figure is labelled illustrative. It prefers 40+ employees and a clear internal owner for the project.

Nyne

The lowest printed bar of the four: twenty full-time staff on a US payroll, in business for three years or longer. It names Asana and Jira project records and accounting data, describes six-figure partnerships as common, and says it will try to top a qualifying written offer someone else made for the same data. Ownership is left to the contract.

Labelbox Prism

No size rule and the quickest money on paper (Labelbox puts first call to payment at less than three weeks), but licenses are exclusive by default and no agency record type is named. It pays an upfront fee plus a revenue share on later use, from around $50K for smaller targeted datasets.

Who else is in the market?

Mercor buys company data too. Its pages would not load for us on 7 Nov 2026, so we print nothing about its terms.

Agencies closing down have a separate route. SimpleClosure AssetHub is not a buyer: a shutdown service that can also place a closing company's assets with buyers: workspace records from tools such as Slack, Notion and Figma, code, IP and domains, each scrubbed of PII before licensing, it says. It publishes neither prices nor its own fee.

Questions about the buyers

Which AI data buyer is best for a small agency?
On printed rules, Nyne sets the lowest bar (a US payroll of 20 full-timers, trading three years or more) and Labelbox Prism sets none. micro1 asks for 30+ full-time staff, and Scale AI prefers 40+. Fit depends on your records more than your headcount.
Do any of these buyers take an agency outside the US?
Not Nyne, whose rule is built around US W-2 payrolls. micro1 buys mainly from American, British and Canadian firms but will look at others. Scale AI and Labelbox Prism print no country rule.
Which buyer keeps the data inside our systems?
Of the four, only Scale AI places its anonymizing step on the seller's side, ahead of export. The rest describe de-identification without a location.
Are these ranges what agencies get paid?
No. They are what each buyer advertises, and Scale AI calls its own range illustrative. No buyer publishes what an agency was actually paid. Treat every range as a ceiling to test, not a price.
Why is Mercor missing from the comparison?
Mercor's pages would not load for us on 7 Nov 2026. Terms appear here only after we have seen them on the company's own site.
Pages this relies on

Get a quote

Describe the agency and where its history is kept, in a few lines. Your answers get compared with what each buyer says it requires, and the email back names any that match. No buyer hears of you until you agree.

Keep records, files, samples and client names out of this form. A description of what the agency holds is plenty.

Step 1 of 3

Where is the company based?

Use the main office. Several buyers limit where they buy from, so location matters.

Get a quote