There's a number Meta will let you acquire customers at indefinitely. Push below it and the algorithm resists. Find it and accept it, and the scale becomes almost automatic. Here's how I think about that number.
Running accounts at a DTC home and lifestyle brand at $300k to $600k a month gives you a lot of time with the algorithm. One of the patterns that became impossible to ignore was this: certain CPA targets would perform perfectly until you tried to scale, and then they'd fall apart. Not because the creative was wrong, not because the audience was exhausted — but because we were trying to buy customers at a price Meta simply wasn't going to sustain at higher volume.
Meta's auction knows things about your category that you don't. It's processed every conversion in your vertical, across every advertiser, at every spend level. It has a model of what a buyer in your space costs to acquire, and it price-discovers against that model continuously. When you set a target that's significantly below what that model says is realistic, the system doesn't fail gracefully — it restricts delivery, stops spending efficiently, and makes every optimization look like it's working until you check the actual cost basis.
The better question isn't "how do I get my CPA lower." It's "what's the CPA Meta will let me buy customers at indefinitely, and does my unit economics work at that number." If the answer is yes, you scale into it. If the answer is no, the creative and offer problem is upstream — not in the targeting.
This reframe matters because it changes where you spend your energy. Fighting the algorithm's price floor with bid adjustments and cost caps is usually a losing game at scale. Working on what determines that floor — your creative quality, your relevance to the audience, and the offer's conversion rate — is where the real leverage is.
Most accounts take one of three approaches to CAC targets. Only one of them scales. The difference isn't discipline or sophistication — it's whether you've accepted what the algorithm is trying to tell you, or whether you're still arguing with it.
At the brand, this played out a few times across different product categories. We'd have a CPA target set based on internal margin requirements — a number that made sense on a spreadsheet. And the account would perform fine at lower spend. But when we tried to scale, delivery would degrade: inconsistent spend, rising CPMs, creative that worked at $500/day falling apart at $3k/day.
The fix, when we finally stopped trying to force the target, was to let the campaigns run broad with strong creative and observe where the algorithm settled. That number was almost always higher than the internal goal — but not so much higher that the unit economics broke. The question was whether we could build the business to accommodate it, not whether we could force Meta to ignore it.
The same pattern showed up across the lead gen accounts I managed. Home services in competitive markets — roofing, bathroom remodel, garage renovation — had floors that reflected how saturated those categories were on Meta. Fighting those floors with cost caps and aggressive targets produced inconsistent delivery. Working within them, while improving creative quality to compress costs at the margin, produced accounts that scaled predictably.
Accepting the floor doesn't mean accepting the cost passively. It means redirecting your optimization energy to the variables that actually move the floor — rather than fighting a bid strategy battle you're unlikely to win at scale.