Impressions measure possible exposure. They do not measure whether a message was understood, trusted, or acted on. For government communications, the useful scorecard begins with the public outcome the campaign exists to change.

An impression is recorded when content could have appeared on a screen. The same person may count more than once. A view may last a fraction of a second. A campaign can generate a large number while no one completes the form, prepares for the storm, attends the meeting, or understands the decision. Reach is necessary, but it is not proof of impact.

Start with the outcome, not the platform

Public communications exists to help real people do real things: submit an application, schedule an inspection, assemble a storm kit, attend a hearing, understand a service change, or reject a rumor before it hardens into belief. Those actions are the outcome. They are harder to collect than an impression count and far more useful because they answer the question residents, commissioners, and reporters will ask: did it work?

We begin with three questions. Who needed to know? How will we know the right people received and understood the message? What should change next? An impression cannot answer the second question and never reaches the third. A credible measurement plan must address all three — or state clearly what it cannot prove.

Write the scorecard before you write the campaign

Impressions become the default when nobody defines success until the work is over. By then, the easiest numbers to find are the ones the platforms provide. The fix is simple: decide what you will count before anything ships. A defensible communications measurement plan names four things up front.

  • A named audience. Not “the public” — the households inside the evacuation zone, the owners facing a permit change, or the parents of rising kindergartners.
  • A defined action. The behavior the work should produce, written as something a person does rather than something a platform records.
  • A baseline. The starting point, so any later movement has context and meaning.
  • A source you can cite in public. Permit software, call-center logs, sign-up forms, web analytics, or a follow-up survey you would be comfortable naming at a microphone.

Track activity and outcomes separately. Content produced, placements earned, distribution, and reach show what the campaign did. Applications, calls, attendance, knowledge, behavior, and changes in public sentiment show what the audience did next. Both matter. Only the second group can demonstrate public impact.

Use the council-meeting test

Assume a commissioner, auditor, resident, or reporter will ask what the community received for the money. A defensible answer names the intended audience, the action or understanding the campaign sought to produce, the evidence collected, and the limits of that evidence. If the plan cannot survive that question, it is not yet a measurement plan.

Reach still matters. A message nobody encounters cannot change anything. But reach is the beginning of the answer, not the end. The stronger answer is clear and specific: this is who needed the information, this is how we reached them, this is what they did next, and this is what changed.