business resources
Paid Media Budgets: A Choice eCommerce Owners Get Wrong, With Shelton Powell of Cart Capital
28 Sept 2026
The choice hiding inside every ad budget
Most eCommerce owners think of paid media as a single decision: how much to spend. That framing is too simple. The real choice is where the money goes first, testing new creative, scaling a proven winner, or reinforcing retention on customers already acquired. Get that order wrong and the budget number stops mattering.
Shelton Powell, founder of Cart Capital, a Miami-based company that manages eCommerce brands for a group of partners, spends most weeks reviewing campaign performance across dozens of brands. That vantage point matters here because the mistakes look identical across accounts, regardless of product or niche. Powell has a clear view on why owners keep making the same call in the wrong order.
The default choice, and why it feels safe
Left alone, most owners put new money into scaling whatever is already working. It feels rational. The campaign has data, the return looks solid, and doubling spend seems like the low-risk move.
Cart Capital's experience running paid media across many brands points to a different pattern. Scaling a winning campaign without fresh creative behind it usually produces diminishing returns within weeks, not months. The algorithm serves the same ad to a wider, colder audience, and the numbers that looked strong at a small budget start to slip as spend increases.
"Owners see one good week and want to pour gasoline on it," Powell says. "But a campaign that works at five hundred dollars a day does not automatically work at two thousand. The audience gets thinner and the creative gets stale at the exact moment you need it to hold up."
The three places a dollar can go
Every paid media dollar in an eCommerce brand has three real destinations. Understanding what each one buys is the actual decision owners are making, whether they realize it or not.
New creative testing
This is the slowest to show results and the easiest to cut when cash is tight. It is also the only way to find the next campaign before the current one fades. Cart Capital treats creative testing as a fixed percentage of spend, not a leftover.
Scaling proven winners
This is where most extra budget goes by default. It works, but only if paired with enough creative variation to keep the audience from going stale. Scaling without variation is the fastest way to turn a good campaign into a mediocre one.
Retention and backend spend
Email, SMS, and post-purchase flows get treated as an afterthought by many owners, something to build later once the front end is proven. Powell argues that is backward. A brand spending everything on acquisition and nothing on the customers it already has is paying twice for growth it should be getting for free.
How to split the budget without guessing
The common advice tells owners to "test creative regularly" and "invest in retention," which is true and not actionable. A workable split looks more like this:
- Seventy percent to proven campaigns that are already converting, with active monitoring for fatigue signals.
- Twenty percent to new creative concepts, run at a small enough spend that a loss does not hurt.
- Ten percent to retention infrastructure: flows, segmentation, and offers built around existing customers.
The exact percentages will shift with the size of the brand and how new it is. A brand three months old needs more in testing, since it has not found its winners yet. A brand with two years of data can lean harder into retention, since the acquisition side is more predictable.
The signal owners miss
Cost per acquisition is the number most owners watch first. It is also the number that lags. By the time it rises, the underlying problem, usually creative fatigue or audience saturation, has already been in motion for a couple of weeks.
The earlier signal is frequency: how many times the same person sees the same ad. When frequency climbs past a certain point without a matching lift in sales, that campaign is on borrowed time regardless of what the cost per acquisition still shows. Cart Capital treats frequency as an early warning system, not a background metric.
What this means for a new brand
An owner just getting a store off the ground does not have the budget to run all three categories at once, and should not try to. The right sequence is testing first, since there is no winner yet to scale. Retention can wait until there is a customer list worth building flows around. Scaling comes last, once a campaign has proven itself at a small spend over a real stretch of time, not just a good weekend.
Owners who skip straight to scaling, because a friend's brand grew fast that way, usually find out later what that shortcut cost them. The brand that grows in the right order tends to still be growing a year later. The one that scales too early tends to plateau, then quietly go quiet.
