Who buys support ticket data for AI training?

Four buyers mention support tickets or support records by name on their sites: micro1, Scale AI, Labelbox Prism and Nyne. Their published ranges run from $10K to $3M+ per company, and their entry rules start at 20, 30 or 40 employees. None publishes a per-ticket price.

Independence note: this site is not affiliated with, owned by or endorsed by micro1, Scale AI, Labelbox or Nyne.

Last verified 7 November 2026

How the published terms compare

Support data terms as each buyer states them (A to Z; checked 7 November 2026)
Buyer	Support data named	Published range	Entry rule	Payment timing
Labelbox Prism	Asks for longitudinal customer support data, and for support logs leans toward consumer brands with heavy ticket volume	Roughly $50K at the small, targeted end; $1M+ for big enterprise deals; a fee paid when the license is signed, then a cut of downstream revenue	Prints no headcount floor; for support logs, busy consumer-facing companies are favored	Says deals can move from a first call to money wired in fewer than 21 days
micro1	Names support records in a long list of operational data that also covers bug tracking, CRM, SOPs and knowledge bases	Headline of $100k-3M+ per company, split into three tiers	Wants established firms with 30 or more full-time staff and records mostly in English; sees most demand from US, UK and Canadian companies	Process section promises full payment within 30 days of acceptance; the FAQ words it as 30 days from invoicing
Nyne	Names support tickets outright, next to accounting, CRM, dispatch and service history records	Typical partnership put at $100K+, moving with the kind and quantity of data on offer	Needs a minimum of 20 US-based, full-time W-2 employees and three-plus years in operation	Not published
Scale AI	Lists support records, and separately chat and messaging histories, among CRM, email, HR and Jira or Confluence data	Puts partnership value at $10K–$1M+, explicitly as an illustration; regular deliveries push it up, and Scale guarantees none of it	Prefers to begin at 40 or more employees, though each company is judged on its own	Not published

Some of these buyers run referral programs, so a license that follows our introduction can earn us a fee. The A to Z order and every word in the rows are the same either way.

micro1

Company data partnerships. Names support records in a long list of operational data that also covers bug tracking, CRM, SOPs and knowledge bases.

micro1 published tiers
Range	Described as
$100K to $500K	The entry tier, for companies micro1 qualifies as enterprise data partners
$500K to $3M	Big operational datasets, or data supplied on an ongoing basis by several teams
$3M+	Operational data that is proprietary and rare

Entry rule: Wants established firms with 30 or more full-time staff and records mostly in English; sees most demand from US, UK and Canadian companies.

De-identification: micro1 runs the de-identification itself, describing processes for transforming sensitive information. Sellers can look over representative samples, and some datasets are rewritten synthetically. Where the work happens is not stated.

Ownership: The underlying data stays the company's property, per the page.

Exclusivity: Nothing on the page gives micro1 exclusive rights.

Runs its own online estimator, which labels every result as indicative and neither a quote nor a binding offer.

From micro1's own page, opened 7 November 2026.

Scale AI

Data Partnership. Lists support records, and separately chat and messaging histories, among CRM, email, HR and Jira or Confluence data.

Published range: Puts partnership value at $10K–$1M+, explicitly as an illustration; regular deliveries push it up, and Scale guarantees none of it.

Entry rule: Prefers to begin at 40 or more employees, though each company is judged on its own.

De-identification: Anonymization happens on the seller's side, within its own systems, ahead of any extraction. PII is stripped and every data package needs the seller's sign-off before Scale receives it.

Ownership: No explicit ownership line; Scale describes the data as remaining under the seller's governance.

Out of scope, per Scale: who your employees and customers are, raw systems, credentials, source access, and anything the signed agreement leaves out.

From Scale AI's own page, opened 7 November 2026.

Labelbox Prism

Prism data licensing. Asks for longitudinal customer support data, and for support logs leans toward consumer brands with heavy ticket volume.

Published range: Roughly $50K at the small, targeted end; $1M+ for big enterprise deals; a fee paid when the license is signed, then a cut of downstream revenue.

Entry rule: Prints no headcount floor; for support logs, busy consumer-facing companies are favored.

De-identification: Prism takes on ingestion and normalization and scrubs PII on the Labelbox side; the page does not say where.

Ownership: Control of the data and its rights stays with the seller, the page says.

Exclusivity: Exclusive unless agreed otherwise, with scope and sensitivity shaping each license.

No calculator. Scoping is free and non-binding and ends in a first valuation.

From Labelbox Prism's own page, opened 7 November 2026.

Nyne

Company data partnerships. Names support tickets outright, next to accounting, CRM, dispatch and service history records.

Published range: Typical partnership put at $100K+, moving with the kind and quantity of data on offer.

Entry rule: Needs a minimum of 20 US-based, full-time W-2 employees and three-plus years in operation.

De-identification: Nyne promises de-identification before any release without saying which party carries it out.

Ownership: Ownership and rights are left to each license agreement to settle.

Offers a three-question estimator that does not show its formula.

From Nyne's own page, opened 7 November 2026.

Who is not listed here, and why?

More companies buy operational records than appear here. A buyer earns a row only when its own site names support data and we could read the terms first-hand. Mercor's pages did not open for us on 7 November 2026, so its terms are absent. Data marketplaces, where you list a product and set your own price, are a different route and sit outside this page.

When one of these pages changes, the edit goes into our terms changelog with a date.

Questions about ticket data buyers

Do these buyers pay per ticket?

None of the four publishes a per-ticket rate. Each publishes a range per company or per partnership, and each sets the final figure after scoping. Treat the ranges as the top of what is possible, not a price.

Can we license the same archive to more than one buyer?

That turns on the contract. Prism starts from exclusivity unless the parties agree otherwise, while nothing on micro1's page hands it exclusive rights. Find the exclusivity clause in each draft and read it twice.

Do we keep ownership of the tickets?

On micro1's terms the underlying data remains yours. Prism leaves control of data and rights with the seller. With Nyne the license agreement settles it, and Scale AI talks about governance staying with you rather than ownership as such.

Which buyer is best for support data?

Nobody can say in general. Floors differ (20, 30 or 40 staff; three years operating; US only or wider), and Prism favors busy consumer brands when it comes to support logs. Your fit with each set of rules settles it.

Mercor buys company data too. Where is it?

Mercor's own data pages would not open for us during the 7 November 2026 check. Until we can read its terms first-hand, this site prints none of them.

Are you one of these buyers?

No. We are an independent introduction service. We never buy, hold or resell data.

Sources behind this page