The ads math is rigged against small stores

There is a loop almost every first-time store owner runs. You post on Instagram and TikTok because posting is free. Engagement is thin, so you put a little money behind it. A hundred dollars, maybe a few hundred. Traffic arrives. Nobody buys. You conclude that ads don't work, or that you don't know some secret, and you go back to posting.

We've read hundreds of founder threads that follow this exact shape, and the wording barely changes from post to post. “I spent $100 on TikTok product ads and it doubled the traffic to my site but didn't get any sales so I ended it,” wrote one founder. Another asked why he was getting no sales after 2,000 store views and nearly $700 spent on TikTok and Reddit ads. A third kept the books on five months of it. Four thousand visitors, thirty-five carts, zero conversions.

The self-diagnosis in these threads is almost always personal. I'm missing something. Everyone else figured out the secret and I didn't. The real problem is arithmetic, and it's worth walking through slowly, because once you see it you stop blaming yourself.

The numbers nobody shows you

Start with what a customer costs. Blended customer-acquisition cost in ecommerce runs somewhere between $68 and $84 per customer, according to benchmark data compiled by MobiLoud. That average describes stores with tuned funnels, retargeting audiences, and experienced media buyers. A brand-new store with none of that machinery usually does worse, not better.

Now the price of attention. Meta CPMs, the cost of showing your ad a thousand times, averaged $13.48 in 2025, up roughly twenty percent year over year, and reached $22.98 in the fourth quarter, per acquisition statistics from Ringly and Meta benchmarks from 27five. The quarter when every retailer on earth is bidding is the quarter attention costs the most. The auction has no discount lane for beginners. Your first dollar competes with a national brand's ten-millionth dollar, at the same prices.

Put those numbers next to a $30 product and a $200 test budget and the test is over before it starts. At average acquisition costs, $200 buys you two or three customers if everything goes right. The targeting, the creative, the landing page, the offer, all of it. Two or three conversions is not a result. It's a coin flip you paid to watch.

A test that can't teach you anything

This is the part that matters. A $200 test doesn't measure whether ads work for your store. It measures whether you got lucky inside a statistical sample far too small to mean anything. If it converts, you learned nothing durable. If it doesn't, which is the usual outcome, you learned nothing either. But it feels like a verdict, and founders treat it like one. They quit the channel and tell the next thread that ads are a scam.

Agencies understand this math perfectly. It's why they ask for three to six months of committed spend before they'll promise anything. They know signal costs money and arrives slowly. Small stores don't have months of committed spend. So they run the only test they can afford, read the noise as a verdict, and walk away from the one channel that scales.

It's a structure problem, not a skill problem

Here's the context the courses leave out. Research from BIA Advisory found that 48.2 percent of small businesses prefer to run their digital advertising entirely themselves, more than double the 22.5 percent who prefer any model where they pay for help. The DIY store owner isn't competing against other DIY store owners. They're competing against staffed teams whose whole job is to watch spend daily, kill what's dying, feed what's converting, and keep the creative fresh enough that the algorithm has something to optimize.

That job is not a weekend skill. It's a function. Somebody checks the numbers every morning, moves budget a few dollars at a time, and retires an ad the day it goes stale instead of the month after. The stores that make paid work aren't smarter. They have that function staffed. Most small stores never can, because the payroll math is even worse than the ads math.

A $200 test doesn't measure whether ads work. It measures whether you got lucky in a sample of three.

What would actually fix it

The honest version of the advice nobody wants to give is that the fix isn't another course about interest targeting, and it isn't trying harder inside a budget that can't produce signal. It's getting the daily work carried by something that doesn't sleep, doesn't guess, and doesn't need the store to gamble a payroll on it, so that a small budget is at least managed the way a big one is.

That's the thing we're building. Budget that flows toward what converts and away from what doesn't. Creative that gets refreshed before it dies. Every change sitting in a queue for your OK before a dollar moves. Watched every day, because every day is how often it changes.