The five silent killers of paced media

Most media budgets don't blow up. They leak. The catastrophic failures — a wrong creative shipped, a targeting flag inverted — get caught in hours because they're loud. The expensive failures are quiet: they look like normal delivery in every platform UI, right up until the flight is nearly over and the make-good conversation starts.

We build an ops copilot that watches paced campaigns all day, and these are the five failure modes it catches most — with the exact signals we use. Steal the checklist even if you never touch our product.

1. Underpacing — the budget you'll never spend

A channel that can't spend its plan is dead weight, and it rarely announces itself: delivery looks “fine,” just light. By the time a weekly report flags it, the remaining days can't physically absorb the leftover budget without wrecking efficiency.

SignalCumulative spend below 80% of the planned curve is a warning; below 60% is critical. Judge each channel against its own planned-to-date, not the campaign blend — a healthy blend hides a dying line.

2. Overspend — the flight that ends early

The mirror image, and more dangerous because it feels like success. A hot line delivering at 150% of plan exhausts its budget with days of flight left — and those dark days land exactly when your client expected presence.

SignalPacing above 125% of plan deserves a bid-down; above 150% it's critical. And past the 60% mark of the flight, project end-of-flight spend from the current run-rate: if the forecast lands under 90% or over 115% of budget, act now — that's when there's still time to correct gently.

3. Invalid traffic — clicks up, conversions gone

The fraud signature is specific: click volume holds or inflates while conversion rate collapses. Bots click; they don't buy. A platform UI shows you a healthy CTR and a quietly dying CPA, and neither number alone looks alarming.

SignalConversion rate below half of the channel's expected baseline, on real volume (we require a few hundred clicks before trusting the ratio), is a pause-and-review — not a bid tweak. Estimate the wasted spend and quarantine the supply.

4. CPA drift — efficiency erodes politely

Nothing “breaks.” The channel just acquires a little worse each day — auction pressure, audience exhaustion, a competitor's new budget — until the blended CPA is far past target and no single day explains it.

SignalA channel running 30% over the campaign's CPA target (with enough conversions to mean it — we use ten) gets a proportional bid-down; 2× over target is critical. Anchor the target to something honest: your own trailing eCPA beats an aspirational number nobody validated.

5. Creative fatigue — the slow CTR bleed

Frequency does its work: the audience has seen the asset, and click-through decays week over week. It masquerades as “audience quality” problems and eats efficiency at the top of the funnel.

SignalCTR below 85% of the channel's baseline while conversion rate holds is fatigue (if conversions collapsed too, look at #3 instead). We require six-figure impression volume before trusting the read. The fix is cheap: rotate the creative.

The meta-problem: who's watching?

None of these signals is exotic. The reason they cost money is that each lives in a different report, on a different platform, and nobody's job is to compute all five, per channel, every day. That's not a diligence problem — it's an architecture problem. Checklists don't run themselves; software does.