How to compare editorial outcomes
For a pitch, check the outreach record. For an interview, verify the published piece. For a podcast, check how its host measures plays. Choose a measure for one editorial format.
Measures by service
| Service and ordering unit | What to measure | Where to verify |
|---|---|---|
| A story pitch for trade editors Pitch copy and a list of target outlets Open worksheet | Documented sends If outreach is in scope, count distinct outlets with a recorded send in the period. Track replies and coverage separately. | Approved outlet list and outreach log. |
| An interview with publication Approved interview and commissioned placement Open worksheet | Interview views For video, report views for this release over a stated period. For a written version, report the outlet’s readership metric separately. | Outlet report for this interview, with date range and metric name. |
| An episode for your own podcast One release-ready episode Open worksheet | Plays after release Report plays for the episode and period using the host’s definition. When the host reports only downloads, label those separately. | Hosting analytics filtered by episode ID and date range. |
Preparing a ranking
- Identify the offer
- Record the storefront, service, offer variant and revision. For a placement, record the publishing platform and format. For production, record the intended use and delivery unit. The published price applies to that revision and scope.
- Verify the outcome
- Use completed orders for an outcome ranking. The editor verifies completion, the measured value and publicly accessible evidence with personal details removed. Record the source URL, measurement unit, start and end dates, and time zone. Present forecasts separately and label them.
- Define a comparable group
- Match the service type, ordering unit, platform, format, measure definition, measurement unit and period. A service and variant ID pair has one position across storefronts. Use a stable anonymous reference for each completed order line to prevent duplicate counts.
- Calculate positions
- Before ranking, specify the sort direction, the minimum verified order lines per variant and the reason for that threshold. Use the median for individual measurements. For a ratio, divide the sum of numerators by the sum of denominators. Multiply by 100 for a percentage. A zero denominator leaves the ratio undefined. Positions require at least two qualifying variants.
- Describe the sample
- In the article, report the period, sources, starting sample and number of variants with comparable observations. Explain exclusions, including missing evidence or conflicting values. Report zero only when it was measured. The conclusion applies to the published sample.