Manual CPC Won't Count Every Conversion You Can Prove. Only the Ones It Wants To.
A campaign confirmed profitable by an independent record can still read as broken on the platform's own dashboard, because the dashboard was never built to see that kind of proof.
The campaign, and the number nobody's dashboard showed
One personal-injury-law account ran $16,000 through a manual CPC campaign over 12 days. Call tracking, independent of the ad platform, confirmed 13 imported leads and 6 closed deals, one of them worth roughly $2,000,000 on its own, at a cost per closed deal of about $1,700. Google's own manual-bidding conversion column counted none of it, because those conversions were confirmed offline, on a phone call, not inside the platform's own tracked event.
A platform's reporting gap doubles as a sales pitch
A campaign producing results this strong should be easy to defend in a budget review. Instead, on manual bidding, it reads as underperforming, because the reporting column that's supposed to prove it out has no visibility into the calls that closed. The nudge that follows is predictable: switch to automated bidding, the mode the platform can actually see and optimize, rather than fix the measurement gap underneath a campaign that was already working.
The proof already existed. It just wasn't somewhere the platform could count it
The six closed deals were never actually in doubt. They were confirmed the whole time, just in a system, call tracking tied to the client's own phone lines and CRM, that the ad platform has no access to. A reconciled record built from that system, checked against what the ad account reports, would have shown the real result on day one instead of leaving it to contradict the platform's own column.