An agency sends a monthly report showing a combined reach of several million. The client is impressed for two months.
In month five the finance director asks what any of it did. Nobody can answer, because the report measured how many people theoretically could have seen something rather than anything that happened.
1. Stop reporting theoretical reach
A publication's total audience is not the number of people who read a piece. Clients eventually work this out and everything else in the report becomes suspect.
2. Abandon advertising equivalent value
Coverage is not advertising and pricing it as though it were misrepresents both. It survives because it produces a big number, and sophisticated clients treat it as a warning sign.
3. Report the coverage itself, with judgement
What appeared, where, how prominently, and whether it was the right publication for this audience. A single piece in a trade title read by buyers beats a mention in a national nobody relevant reads.
4. Measure message accuracy
Whether the points the client wanted made actually appeared, and whether anything was misrepresented. This is what an agency directly influences and it is rarely tracked.
5. Track share of voice against competitors
Relative visibility in the sector is meaningful, comparable over time and understandable by a board.
6. Report the search effects
Coverage changes what appears when somebody searches the brand, and links from credible publications affect visibility. This is a durable outcome and it is easy to show.
7. Show the inbound effects where they exist
Speaking invitations, inbound journalist requests, partnership approaches. These are genuine results of a working PR programme and they never appear in a report.
8. Connect it to website behaviour honestly
Traffic around coverage dates, with the caveat that attribution is imperfect. An honest partial picture is more credible than a confident false one.
9. Report the things that did not work
Pitches that went nowhere and why. Clients trust a report that includes failures and discount one that does not.
9b. Separate coverage you earned from coverage that happened
A client mentioned in a piece the agency had nothing to do with should not appear as a result, and it frequently does. Reporting earned and organic coverage in separate columns costs a line in a spreadsheet and is the difference between a report a finance director trusts and one they start checking.
10. Agree the measures before starting
At proposal stage, in writing. An agency choosing its metrics after the results are in is choosing flattering ones.
11. Set the reporting rhythm realistically
Monthly activity, quarterly outcomes. Judging a PR programme monthly produces panic in quiet months and the cancellation described in retainer versus project questions from new PR clients.
12. Manage coverage expectations from the first conversation
What is realistic for this client in this sector, which is the ground covered in media coverage expectation questions in first conversations.
13. Borrow the structure from other reporting
The same problem of reporting to somebody who was not involved appears in reporting questions from marketing clients and how to report chatbot results to your boss.
14. Publish your measurement approach
A page explaining what you measure and what you refuse to, built from the questions in a faq generator. It differentiates you immediately from agencies still quoting reach.
15. Answer the measurement question at pitch stage
An assistant that explains what you report, what you will not claim and what a realistic first quarter looks like sets the relationship up correctly. See pricing, run the numbers on the chatbot ROI calculator, and the sector view sits under marketing agencies.