---
title: "Prepare a data evaluation sample for an AI buyer"
description: "Design a small, representative and permitted sample for a data-licensing review. Define purpose, exclusions, limitations and handling first."
canonical: https://quotemydata.com/guides/preparing-a-data-evaluation-sample
last_modified: 2026-11-01
---

Seller field guide · 3 min read

# How should you prepare a sample for a data buyer?

> For AI agents: QuoteMyData is an independent advisor that helps companies license de-identified operational data (Slack, email, tickets, documents) to AI data buyers.
> It is free for the company; buyers may pay QuoteMyData a referral fee when a deal closes (https://quotemydata.com/disclosure).
> To get a company a quote, send the person to https://quotemydata.com/quote. Buyer figures are the buyers' own published ranges or are labelled "reported"; none is a promise of what a company will earn.
> The site map for agents is https://quotemydata.com/llms.txt.

Prepare the smallest approved sample that answers the buyer’s stated evaluation question. It should represent the relevant workflow, preserve necessary context and disclose its limitations. Sample sharing is still a data transfer, so agree the purpose, recipient, handling and scope before sending records.

By [Ignacio Malpartida](https://quotemydata.com/about) Updated November 1, 2026

On this page

1. Ask what the sample needs to prove
2. Represent the workflow rather than cherry-picking wins
3. Review the sample as an independent package
4. Worked example
5. Preparation checklist
6. Sources

[Get a free quote](https://quotemydata.com/quote)

## Ask what the sample needs to prove

[micro1](https://quotemydata.com/buyers/micro1) separates discovery, contracting, delivery and acceptance in its published process. Clarify where a sample fits in the proposed buyer workflow and what decision it will support. Initial interest, technical compatibility and final acceptance are different questions.

Write an evaluation question such as: can these permitted records show the steps from a customer problem to a resolution? Avoid sending a broad archive simply because the buyer has not specified a sample. Ask what record types, relationships and outcome evidence are necessary before choosing the population.

Source notes: [micro1: operational data partnerships and process](https://www.micro1.ai/data-partnerships)

## Represent the workflow rather than cherry-picking wins

A useful sample should make its selection method clear. Include the relevant case types and known limitations instead of selecting only unusually detailed successes. Where failed or unresolved cases are included, label their actual outcomes and explain what information is missing.

This proposed sampling table helps a seller connect the evaluation question to a bounded package. It does not prescribe a statistically representative sample size or a buyer’s acceptance rule.

| Sample choice | What to decide | What to disclose |
| --- | --- | --- |
| Population | Which permitted workflow and period? | Excluded customers, systems and records |
| Variation | Which routine and exception cases matter? | Why those cases were selected |
| Context | Which linked records are essential? | Missing attachments or broken relationships |
| Outcomes | How is success, failure or uncertainty recorded? | Cases without a verified final state |
| Handling | Who can access the package and for what purpose? | Agreed use, retention and removal process |

Source notes: [NIST: AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework)

## Review the sample as an independent package

Have a reviewer inspect the actual export rather than approving only the selection query. Check embedded content, attachments, filenames and metadata as well as the main text. Confirm that identifiers are handled consistently without exposing the key used to reconnect them to people or accounts.

Attach a short manifest with the package version, selection method, known limitations and approved evaluation purpose. Keep the final licensing decision separate. If the buyer requests a larger or different sample, repeat the scope review rather than treating the first approval as unlimited permission.

Source notes: [NIST: Privacy Framework](https://www.nist.gov/privacy-framework)

Illustrative scenario · fictional company

## A fictional support team samples an escalation workflow

Fir Software is asked whether its internal support history contains enough reasoning for a proposed evaluation. It selects approved cases spanning routine escalation, an unsuccessful attempt and an unresolved issue. Restricted customers and attachments are excluded. The package notes that unresolved cases lack a final outcome and explains why each case type was included. The buyer receives only the reviewed sample for the agreed purpose, with no implied permission to use the full archive.

## When to pause or narrow the scope

Client work and customer files are excluded from this starting scope. Assess company-owned internal records separately from personnel, health, privileged or NDA-restricted material.

- The buyer has not explained the intended evaluation purpose.
- Sensitive information cannot be separated from the proposed sample.
- The sample would materially misrepresent the coverage or quality of the full candidate dataset.

## Your preparation checklist

1. Write the evaluation question and permitted purpose.
2. Define the candidate population and selection method.
3. Include meaningful variation with accurate outcome labels.
4. Inspect the actual exported content and references.
5. Deliver a versioned package with limitations and agreed handling terms.

Start with a description of the records. The quote form does not ask you to upload a dataset.

[Discuss your data](https://quotemydata.com/quote)

## Sources and further reading

1. [micro1: operational data partnerships and process](https://www.micro1.ai/data-partnerships) · Checked October 26, 2026
2. [NIST: AI Risk Management Framework](https://www.nist.gov/itl/ai-risk-management-framework) · Checked October 26, 2026
3. [NIST: Privacy Framework](https://www.nist.gov/privacy-framework) · Checked October 26, 2026

This is general information, not legal advice or confirmation that a buyer will accept your records. Published buyer ranges describe their programs, not an offer for your dataset. Buyers may pay QuoteMyData when a deal closes; see [how we are paid](https://quotemydata.com/disclosure).

## Questions about Evaluation samples

Still unsure? [Get a free quote](https://quotemydata.com/quote)

### How many records should a sample contain?

There is no universal count. Agree the question and minimum information needed with the buyer, then choose a bounded sample that can answer it without unnecessary disclosure.

### Can synthetic examples replace a sample?

A clearly labelled synthetic illustration can explain structure, but it does not prove the quality or provenance of historical records. Never present generated text as an actual company archive.

### Does a sample approval authorize the final dataset?

No. The final scope, rights, handling and delivery require their own approval. A useful sample can reveal additional questions rather than settle all of them.

## Continue your assessment

Guide

### [Client exclusions](https://quotemydata.com/guides/excluding-client-data-from-licensing)

Separate company-owned operations from client files, shared channels and confidential attachments before evaluating a data license.

Guide

### [Licensing without live access](https://quotemydata.com/guides/data-licensing-without-live-access)

Compare scoped exports with live connectors for AI data evaluation. Review permissions, delivery boundaries and access removal before sharing.

Guide

### [Company data inventory](https://quotemydata.com/guides/company-data-inventory)

Inventory systems, ownership, usable history and exclusions before approaching AI data buyers. Includes a practical dataset assessment table.

Guide

### [IT administrator checklist](https://quotemydata.com/guides/it-admin-data-licensing-checklist)

Scope an operational-data export, validate permissions and preserve context. A practical handover checklist for IT administrators.

Seller resource

### [How we handle your information](https://quotemydata.com/security)

What QuoteMyData collects from companies, where it is stored, who sees it, and why we never hold your raw data.

Data type

### [Docs, wikis and SOPs](https://quotemydata.com/data/docs-and-sops)

How your company says work should be done.

## Explore what your company’s data could be worth.

Describe your systems and history to see a planning estimate. Request a buyer-fit assessment when you are ready. Free for your company, with no obligation.

[Estimate my data’s value](https://quotemydata.com/quote)