Before speaking with a data buyer, list the records your company holds, the period they cover and the questions that need an answer. You can begin with descriptions and approximate counts while the records stay in your systems.

Download the blank company data inventory. Keep your completed copy within your company until you have approved a recipient and the information to share. Leave out customer names, private record links, passwords and account details.

Records to describe

Start with one work process whose records you understand. Examples include a support case, a transaction review, a reconciliation or a renewal. List the systems used during that work and identify which decisions and outcomes remain available.

Providers describe different needs in their public programs. Mercor lists communication tools, document systems and CRM records. micro1 names operating procedures, project histories, quality reviews and decision-making patterns. These examples can help you describe a collection; each provider still needs to assess its relevance. Mercor Data and micro1 data partnerships

Use one row per collection in the inventory. Add another row when a system, date range or set of permissions needs separate treatment. Give each row a short reference so you can describe connections between collections without inserting private links.

Dates and volume

Record the earliest and latest dates you can verify. A company's operating history can be longer than the history available in its current systems.

Use a count that describes the collection. Tickets, completed cases, messages and files measure different things. Record the unit and how you obtained the estimate. If the count is unknown, leave it blank and add the task needed to check it.

Keep complete case counts separate from file counts. Several documents may belong to one case, while a folder may include duplicate versions or blank forms. Identify known gaps before presenting a total as usable material.

Connections and system changes

Describe how records connect to the same piece of work. A case reference may connect a request, supporting documents, comments, an approval and the outcome. Record whether those connections can survive an export.

For each system change, note the old and new systems, the periods they cover and any lost links, attachments or messages. Ask the system owner which identifiers were carried across. Keep this information beside the affected collection so a buyer can understand the available history.

Follow your existing retention and legal-hold requirements when checking an archive. An inventory is a description of what is available; preparing it does not require deleting or exporting records.

Export options and effort

Ask the system owner which export formats are available and what they include. Check whether an export would preserve dates, comments, attachments, relationships and field definitions. Note any vendor limits or additional fees.

Leave preparation hours blank until someone has assessed the work. A useful next action might be to confirm whether attachments are included or whether an old archive is still accessible.

Keep export capability separate from permission to share. Any access, sample or transfer needs its own approved scope and handling arrangements.

Permissions and exclusions

Record the role responsible for each collection and the questions they need to resolve. These may concern customer contracts, employee material, confidentiality, third-party content or source-system terms. Ask the appropriate advisers to review the proposed use.

Describe restricted categories without copying sensitive examples into the inventory. For instance, record that a collection contains borrower financial information or licensed title-plant material, then identify who will review it.

Defined.ai describes separate checks of data quality, volume, metadata and licensing rights before a marketplace listing. A completed inventory supports those discussions; it cannot establish permission, buyer acceptance or a price. Defined.ai partnership process

Fictional example

A service company has approximately 1,200 support cases from 2022 to 2025. The estimate comes from its case-management report. Its current system holds cases from 2024 onwards, while an older archive holds the earlier records.

The company records two collections. It notes that old case numbers survive in the current system, but attachments from part of 2023 are missing. The support operations manager will check the archive, and the contract owner will review which customer commitments apply.

The company can resolve these questions before discussing a sample. A buyer would still need to assess the records and decide whether to make an offer.

Before an initial discussion

  • Describe the collection, work process and available period.
  • State the approximate quantity, its unit and any known gaps.
  • Identify the person who can explain the records and approve the next step.
  • List unresolved permissions and material that needs exclusion or review.
  • Confirm the proposed recipient before sharing a company profile or completed inventory.

Company size can help a provider screen inquiries, but it does not establish the depth or usefulness of the records. A buyer may need a specific kind of work, sufficient examples and the right permissions.

Use your inventory to compare your records with the published requirements of seven company-data programmes. DataPayouts can help you assess requirements and proposed terms before choosing a deal.

For the wider process, read how to sell company data for AI. You can also review the records considered for real estate brokerages, title and settlement companies, mortgage brokers and insurance agencies.