Skip to content
PressFriendlyツ
When It's Not Working

Is the problem the agency or is it you?

Audit your own side first, because the math favors it. Replacing an agency typically costs another 60 to 90 days of onboarding and relationship ramp before the next placement lands. Fixing the three most common buyer-side failures (no news to pitch, week-long approvals, no agreed goal) costs about a week and no transition. The two failure modes look identical from the outside: three quiet months either way. The difference is that buyer-side failures repeat with the next agency.

The split is about who controls the input. You own the raw material and the decisions. The agency owns targeting, angles, and access.

Yours to fix: no news, slow approvals, no agreed goal

  • Nothing to pitch - no launch, funding, data, or notable hire in 90 days starves any program. PR amplifies news; it cannot manufacture it.
  • Slow approvals - reporters work on same-day clocks; in a recent Cision media survey, about 47% of journalists named speed of response as one of the things they most want from PR. A week of internal sign-off kills live stories no matter how good the pitch was.
  • No named goal - if you never agreed what success means, "underperforming" has no definition and the agency defaults to chasing hit counts.

Theirs to own: targeting, staffing, and honest reporting

  • Wrong reporters, recycled angles - targeting and freshness are the craft you pay for.
  • Rotating juniors, going dark - if the senior person vanished after the sales call, the agency reduced its effort, and your news flow does not explain that.
  • Reporting theater - thick monthly decks built on impressions and "momentum" instead of pitches sent and responses received.

The pitch-log test settles ambiguous cases

Ask to see every pitch from the last 60 days, with the outlet and the response. Real volume aimed at the wrong reporters is an agency problem you can renegotiate. Thin volume against a calendar with nothing on it is your problem. If they cannot produce the list at all, treat that as the finding.

If the audit points at you, fix the input and give the program one more full cycle before judging again. If it points at them, the pause-renegotiate-or-fire page in this section works through the decision, and the hard-conversation page covers how to raise it without burning the relationship.