Sell Workflow DataGet a quote

What counts as workflow data?

Workflow data is the record of how work moves through a company: tickets and how they were resolved, project and sprint logs, approval chains, QA checks, SOPs and the decisions written down along the way. Buyers value the sequence and the outcome, not the documents alone.

Sheet last verified 7 November 2026, against every buyer's own page.

Which record types do buyers name?

Workflow record types, the systems they live in, and the buyers that name them
Record typeWhere it usually livesWhat a buyer looks forNamed by
Support tickets and their resolution historyHelp desks such as Zendesk, Freshdesk, IntercomThe full thread, each status change, who escalated and how it closedmicro1, Scale AI, Nyne, Labelbox Prism
Project, sprint and bug-tracking logsJira, Asana, LinearTickets linked to commits, reviews and releases; planning notesmicro1, Scale AI, Nyne
CRM and sales activitySalesforce, HubSpotDeal stages, activity logs, handoffs between teamsmicro1, Scale AI, Nyne
Accounting and ERP recordsQuickBooks, Xero, NetSuiteInvoices, estimates, payables and receivables, approval trailsmicro1, Scale AI, Nyne
SOPs, process docs and knowledge basesConfluence, Notion, shared drivesWritten procedures that match what the tickets show was donemicro1, Scale AI
Jobs, dispatch and calendarsField service and scheduling toolsJob records from booking to completionNyne, Scale AI
Chat and email threadsSlack, Microsoft Teams, emailWork conversations tied to the tasks they decidedScale AI
Agent trajectoriesLogs of AI agents doing real tasksStep-by-step records of an agent completing workmicro1

"Named by": buyers whose own pages (opened 11/07/2026) list the category. Tools are examples of where such records sit, not endorsements. Disclosure: buyers in this column may later pay us an introduction fee when a company we introduced signs; the column is the same either way.

Value drivers

What makes one company's records worth more than another's?

Length of history matters first. A help desk with six years of tickets shows how processes changed; one with six months shows a snapshot.

Links between systems matter next. Labelbox Prism lists cross-system interaction data as a category of its own: the support ticket that became a bug, the bug that became a release, the release note the customer saw. Records that can be joined are rarer than records that cannot.

Outcomes matter. A ticket that records how it was closed, and whether the customer came back, is a complete example. One that stops at "assigned" is not.

Then structure and language. micro1 does not prioritize data that is not primarily in English, and every buyer's pipeline runs faster on exports with consistent fields.

Exclusions

What is usually left out?

  • Credentials, raw system access and source access. Scale AI lists these as things it does not take.
  • Employee and customer identities, also on Scale AI's list. Other buyers remove them during de-identification.
  • Anything outside the signed agreement. Only the record sets named in the contract move.
  • Material that belongs to your clients, unless the client contract allows licensing it.
  • Customer personal details, unless contract and statute both still permit it once de-identified.
  • Categories you choose to keep back. Nyne says companies can exclude categories they pick.
Sources behind this sheet (5)

Questions

Questions about what qualifies

Is a folder of documents workflow data?

Partly. Documents on their own are reference material. They become workflow data when they are tied to the work that used them: the ticket that cited the SOP, the approval that changed the plan.

How many years of records do buyers want?

Nyne wants a business at least three years old. micro1 fixes no minimum, though it tells applicants a longer record usually fits better. Scale AI and Prism print no number.

Do HR records count?

Scale AI names HR records among the categories it takes. They are also the records most likely to contain employees’ personal data, so expect the heaviest de-identification and the most legal review.

Do records have to be in English?

micro1 says data that is not primarily in English is not prioritized. The other three buyers do not publish a language rule.

Can a company that has closed sell its records?

Sometimes. micro1 has a distinct track for defunct firms, limited by its referral rules to ones that operated in ten or more countries. After a wind-down, investors or an acquirer may own the rights, so confirm title first.

Get a quote

Takes about two minutes. Give a ZIP code, a reply address and a few facts about your records; the reply names each buyer whose printed criteria your company looks set to clear.

Describe the records; never attach or paste them. No data, exports, screenshots or samples belong in this form.

Plain disclosure: introductions are our business, and the company we introduce you to may pay us.

1/3

Where is the company based?
Get a quote