Quantitative data tells you how much or how often; qualitative data helps explain what something meant and why it happened. A PR report is strongest when it uses both, tied to the campaign's original objective. Counting articles alone does not establish that an audience understood a message or changed its behavior.
What each type can answer
01
Quantitative evidence
Counts, rates, and comparable measures can show release pickups, relevant coverage, visits, enquiries, survey responses, or changes over time. Record definitions, dates, denominators, and data quality.
02
Qualitative evidence
Interviews, open-ended answers, careful content coding, and customer or reporter feedback can reveal confusion, trust, objections, and context. Document who was heard and how responses were interpreted.
03
Mixed evidence
Combine a measure with its explanation: for example, a fall in support calls plus interviews that show whether the new instructions are understood. Neither type automatically proves the campaign caused the change.
Build a useful evaluation
Start with a specific question such as whether a product-safety update reached affected customers and gave them an understandable next step. Track delivery and response measures, then sample actual messages or ask people to explain what they understood. Compare findings with a baseline where one exists and note outside events that may affect the result.
AMEC's measurement principles call for qualitative and quantitative analysis together. Its framework also separates activity and outputs from audience effects and organizational impact. That separation prevents a large clipping count from being presented as a business outcome.
Common reporting traps
Do not sum potential audience figures from overlapping outlets into a precise person count. Do not label automated sentiment as a definitive human judgment without checking examples. Do not convert every qualitative comment into a percentage from a tiny, unrepresentative sample.
Show the underlying evidence, the method, and the limitations beside the headline metric. A smaller but well-explained result is more useful for the next decision than a large unqualified number.