Stop Ads That Lose Money

The problem with keeping campaigns that lose money

Every ad account has a campaign that loses money. And almost every account keeps it running. Four things protect it:

1. It still shows revenue. So cutting it feels like cutting revenue.
2. Someone has a stake in defending it, your media buyer, your agency.
3. The tracking is fuzzy enough to argue about.
4. And there is always one more test to run. So the campaign never gets cut, just tweaked for another quarter.

Here is the trap. A losing campaign feels like a fix-it problem. It is really a stop-doing decision. Tuning can nudge a bad campaign to slightly less bad. But slightly less bad still loses money on every click. If a full quarter of tuning did not lift it over break-even, it does not work at your costs.

Your gut gets one thing wrong. Cut the campaign and the revenue disappears. But the ad spend disappears with it. Was the spend bigger than the profit the revenue left behind? Then the cut makes you more money, not less. That pain of a dropping number keeps brands feeding losers for months.

1. An example showing you the numbers

You run a Direct-to-Consumer (D2C) outdoor gear brand. A typical order is $100. Twelve campaigns run at once across paid social and paid search.

First, the one number that sorts winners from losers: your break-even Return on Ad Spend (ROAS). ROAS is sales divided by ad spend. On this store, the five non-ad costs eat $70 of every $100 order. That leaves $30 to pay for ads. So the ads must return $100 ÷ $30 ≈ 3.3x just to break even. Nine of your campaigns sit above that line, at around 5.0x. Three sit below it, at 2.0x.

Here is one $100 order from each group, side by side.

One $100 order, a winning campaign vs a losing campaign

Per order, whole dollars. At 5.0x ROAS the ads cost $20 per $100 order; at 2.0x they cost $50.

Same product. Same customer. Same $100 on the dashboard. The winning campaign leaves $10 behind. The losing campaign needed $50 of ads to win the order. So it takes $20 out of your pocket. On every sale it ‘wins’.

The three losing campaigns bring in 4,000 orders a year. On the dashboard, that is $400,000 of tracked revenue. In the bank, it is 4,000 × -$20 = an $80,000 loss per year. It hides inside an ad account that looks healthy.

Cut the three campaigns and the dashboard drops $400,000 of revenue. It feels terrible. But $200,000 of ad spend disappears with it. So does the $80,000 loss. But the win only lasts if you hold the cut. Switch the campaigns back on ‘for one more test’ and the loss switches back on too.

2. How to cut losing campaigns and make the cut stick

Two weeks from decision to done. Most of it is one spreadsheet and one honest talk.

1. Run the break-even check on every campaign, one by one. Use the last 90 days of spend and sales to smooth out seasonal noise. Divide each campaign’s sales by its ad spend. Compare it to your break-even ROAS, 3.3x on the store above. Flag whatever sits below the line week after week.
2. Check each flagged campaign for a real repeat-purchase defense. A campaign below break-even earns a stay only one way. Its customers must come back and buy again, proven in your own data. No proof? It goes on the cut list.
3. Bring the math to your agency and expect pushback. Frame it as moving budget: ‘These three sit below our profit floor. We are moving the budget to the ones above it.’ Agencies often answer with ‘halo effects’ and ‘brand lift’. Ask for real proof. A switch-off test. A brand study. Vague claims are a defense of spend, not data.
4. Cut cleanly. Pause the campaigns. Do not ‘tune them one more time’. One more creative, one more audience, one more quarter, that pattern keeps losers alive. If 90 days of care could not lift a campaign over break-even, pausing it is the fix.
5. Move half the freed budget to winners; bank the other half. The cut frees $200,000 of yearly spend. Moving $100,000 into the winners at 5.0x buys about 5,000 extra orders. Each leaves $10, roughly $50,000 of added profit. But watch the winners for 30 to 60 days. Big budget jumps can push a campaign into weaker audiences. If their ROAS sags, bank more.

3. One warning before you act

Not every below-break-even campaign is an instant cut. Some campaigns lose money on the first order on purpose. Their customers come back and buy again. Subscriptions and daily-use products most of all. For those, the first order is the price of the relationship.

The bar is high. The campaign’s own customers must show the repeat buys, not your store average. The repeat buys must pay back the first-order loss. Within 6 to 12 months. In your own numbers. Some losing campaigns chase cheap clicks. Their customers come back less than average. Then the defense fails. Cut it.

4. Frequently asked questions

5. Quick reference: what to avoid and what to apply