“I don't kill a campaign on ROAS alone. I compare it to my break-even floor, which is just 1 divided by my gross margin, and I watch whether the campaign still clears that floor when I raise the budget. A campaign that looks profitable at €2,000 can fall apart at €5,000 because the next euro returns less than the last.”

“I keep campaigns running when the marginal dollar still clears a CAC the business can actually believe in, not just the one the dashboard shows. At Cash App we built attribution and a media mix model early because we knew paid would keep getting credit for users who would have come through referrals anyway.”

“I stop looking at CTR, CPC, and CPM first. After $100M in ad spend, the only numbers that tell me if a campaign deserves more money are cost per purchase, AOV, ROAS, and total revenue. If those are healthy, I keep it running and feed it more budget; if cost per purchase is above what the business can afford, a pretty CTR will not save it.”

“I look at payback period and whether the acquired customers actually stick and expand. At DoorDash and Square we tracked cohort usage and revenue retention because a cheap sign-up means nothing if it churns in ninety days. If your paid campaign pays back in under six to twelve months and the retained cohorts deepen usage, keep funding it; if not, kill it fast and reallocate.”

“I once told a client to turn off their Brand campaign even though the ROAS looked beautiful. We tracked backend revenue and new customers, and four weeks later revenue was up and they had 42 more new customers than before. The campaign was just taking credit for organic traffic.”

“I don't kill a campaign off platform ROAS alone. I reconcile CRM revenue weekly against what Meta or Google claims, because platform numbers are directional, not ground truth. If the CRM shows a profitable LTV:CAC and the funnel isn't leaking, the campaign stays.”

“I look at TACoS first, not ACoS, because a 10% TACoS means my ads are pulling organic growth with them. If total sales are climbing, I keep funding the campaign even when the first-sale ACoS looks ugly, especially when Subscribe and Save or repeat buyers turn that initial loss into long-term profit.”

“I ran a $4M paid media year for Business Mastery 2024 with a 3.5+ ROAS target, and the only thing that kept us from panic-killing good campaigns was a hard rule: 30 days minimum before you decide. Most people pull budget the moment cost per lead spikes on day three, but on Pinterest especially, people save, share, and come back later, so the early numbers lie.”

“I killed a $40K monthly retainer once because the client was measuring clicks, not pipeline. We flipped the test to track downstream revenue per dollar spent, and three campaigns that looked expensive became the only ones worth scaling.”