A coverage report can tell you where a story appeared, but it rarely tells you exactly how many people read it. Outlet audience figures and social impressions are estimates of opportunity to see, not proof of attention. A defensible analysis labels each measure, removes obvious duplication where possible, and connects coverage to the audience and objective that mattered in the first place.
Start with the coverage record
Ten syndicated versions of one story are ten placements, but not necessarily ten independent editorial decisions or ten distinct audiences. Keep placement counts separate from unique story counts and identify syndication in the report.
- Save the original article or broadcast reference, publication date, outlet, author, URL, and any later correction.
- Classify whether the mention is a full feature, brief citation, syndicated copy, paid placement, or owned content.
- Record geography, topic, message accuracy, and whether the intended audience is likely to use the outlet.
- Note the source of every audience estimate and the date it was obtained.
Potential reach is not actual reach
A publisher's monthly unique visitors, circulation, or broadcast audience can offer context about an outlet's scale. It does not reveal how many people opened your particular story, noticed your mention, or remembered it. Adding the full audience of every outlet will usually overstate unduplicated reach because audiences overlap and the article may occupy only a small fraction of each outlet's output.
If article-level analytics are available from a publisher, describe the metric and its limits. Page views count loads, not necessarily unique people or attentive reading. Social-platform impressions can include repeated exposures. Use 'potential audience' or 'estimated impressions' when that is what the data actually represents.
Measure relevance and response
01
Audience fit
Was the placement in a publication the target group reads, in the relevant market and language?
02
Message quality
Did the story carry the correct facts, product name, context, and a useful link or action route?
03
Observable response
Look for referral visits, branded search changes, inquiries, downloads, or survey responses, with dates and attribution limits noted.
04
Longer-term outcome
Use repeated surveys or other appropriate evidence to test awareness, understanding, trust, or behavior rather than inferring them from a placement count.
Build a report that can be checked
Group results into outputs, out-takes, outcomes, and impact, following AMEC's evaluation framework. A compact table can show the story, outlet, relevant audience, estimated opportunity to see, observed engagement, and any measured outcome. Include a baseline and time window so changes can be interpreted.
Avoid converting coverage into an advertising-equivalent dollar figure and calling it return on investment. A media mention has different editorial context, control, and effect from an ad. State what was measured, what was estimated, and what cannot yet be attributed to the coverage.