> ## Content Index
> Fetch the complete content index at: https://blog.onecrewai.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# Creator reply rates do not match? Agree on the definitions first
- URL: https://blog.onecrewai.io/en/creator-operations-metrics-definitions/
- Published: 2026-09-15T13:35:30.000Z
- Updated: 2026-09-15T13:35:30.000Z
- Description: Align creator reporting with clear cohorts, denominators and cut-off dates. Separate human replies, interest, delivery and sales results using a practical definitions table.
- Author: OneCrew Team
- Tags: Creator operations, #en, #pair-27

You calculate a 30% reply rate. A teammate reports 20% for the same activity. One counted people, the other counted emails, and someone added replies from an earlier campaign.

Before judging a number, define who is included, what counts as an event and when the measurement stops. Track outreach, replies and delivery separately from sales outcomes, and state how each measure is calculated.

## Fix the cohort before calculating the rate

People first contacted this week are not the same group as every partnership active this week. Record the campaign, first-contact period, deduplication rule and cut-off date. If the metric counts people, three follow-up emails still represent one person.

Check identities across channels and teammates. Keep different campaigns distinct. Email address counts are not automatically people counts, and a reply about another activity does not belong in the current cohort.

[OneCrew](https://onecrewai.io/resources/user-guide?ref=blog.onecrewai.io) is an AI workspace for creator discovery, contact organization and outreach communication. Use the campaign group and synchronized messages as the basis for an AI-assisted list of records and contact sources. Review duplicates and scope first, so calculations begin with an agreed set of rows.

![Actual OneCrew interface: preparing a campaign lead group, using demo information.](https://storage.ghost.io/c/2d/ee/2deeacfe-d870-4a0c-a406-95a0ad33ad67/content/images/2026/09/onecrew-workflow-en-group-setup.png)

OneCrew: organize leads into campaign groups.

## Separate a reply from an expression of interest

A human reply may be a refusal or a request for information. Define what qualifies as interest, and keep automated responses and bounces separate. Once the team chooses a definition, display it with the metric and reuse it.

In a hypothetical cohort of 50 unique contacts, 15 human replies by the cut-off give a 15 ÷ 50 = 30% reply rate. If five people want to continue, 30% is not the interest rate. The example illustrates a definition, not an industry benchmark.

Review the conversation context in OneCrew. Ask AI to apply your stated categories—human reply, interested in learning more, declined or needs review—and identify supporting messages. Check ambiguous cases before calculating.

## Measure delivery against explicit commitments

Keep product experiences without a posting obligation separate from partnerships with agreed deliverables. For on-time delivery, specify which commitments have fallen due, how extensions are handled and whether the unit is a creator or a piece of content.

One possible person-based definition is: creators whose original commitments were due by the cut-off and who completed all agreed content on time, divided by all creators with original commitments due in that cohort. Show extensions separately and retain the original date and change confirmation.

Provide confirmed delivery terms and actual content links to your connected AI. Use the OneCrew correspondence to organize due, completed, extended and unverified items. Have the owner review the rows before the figures enter the report.

| Metric           | Definition to record                            |
| ---------------- | ----------------------------------------------- |
| Human reply rate | Unique contacts, reply definition and cut-off   |
| Interest rate    | Interest criteria and chosen denominator        |
| On-time delivery | Due cohort, unit and extension treatment        |
| Sales results    | Source, attribution window, refunds and overlap |

## Keep sales data attached to its source

For clicks, orders and revenue, record the reporting source, attribution window, refund treatment and measurement time. Do not simply add platform, link and discount-code results when overlap is unknown. Show them separately until duplicates can be resolved.

Ask your connected AI to compare the verified sales information you provide with the OneCrew partnership context and identify questions, such as whether reporting dates match the content’s publication date. Retain the original report instead of treating all sales in the same period as creator-driven.

Before the next report, select a cohort in OneCrew and review contact, reply and delivery evidence against shared definitions. Then calculate. You can assess progress much more confidently when everyone is counting the same thing.