An ecommerce ad works when it brings you a new customer for less than that customer is worth on the first order. Judge every creative on cost per purchase against your break-even, give it seven days, and when it misses, work down a fixed ladder to find out why before you blame the creative. Often, the ad is not the thing that is broken.
Cost per purchase, also called CAC, is what you paid to get one customer. Your break-even is your average order value minus everything it costs to deliver the order. That gap is the most you can pay for a customer and still not lose money.
Ecommerce Creative Has a Different Job Than Advertising Creative
This is the part almost every guide skips. Ecommerce creative is built to drive a specific purchase, not awareness and not brand affinity. That changes what it has to do in thirty seconds.
- It explains the product fast. The physical product is usually central to the frame, not implied.
- It proves value and reduces doubt. It answers the practical buying questions: how it is used, whether it fits, the quality, the price, whether you can be trusted, shipping, returns.
- It demonstrates. Demonstrations land hardest when someone like the target customer uses the product in a situation the viewer recognizes.
- It matches the page it sends people to. The ad and the landing page have to carry the same product, promise, price and offer, or the click goes nowhere.
- It is produced as a system, not as an asset. Multiple hooks, angles, demonstrations, proof points and offers, running at once.
Brand advertising can succeed by being remembered. Ecommerce creative succeeds by being bought from, and it gets measured against sales economics: conversion rate, acquisition cost, margin, and what a customer is worth.
One Number Decides It, and the Rest Only Explain It
That is the decision, and it is the only one that matters at the moment of keeping or killing. Every other metric in the account is diagnostic. They tell you why you got the result. An ad with a mediocre click-through rate that acquires customers at $22 against a $27.70 break-even is a good ad, and tuning it toward a prettier click-through rate is how you make it worse.
The fix: work out your break-even first, before you open the ad account, and judge every creative against that one bar. The full break-even and contribution-margin build is here. Whether the account as a whole is paying off is a separate question, covered in how to tell if your Facebook ads are actually working; this one is about which creative stays.
When an Ad Misses, Work Down This Ladder in Order
When cost per purchase comes in over your break-even, do not swap the creative yet. Go down the ladder and find the rung that is actually failing.
| Rung | What it tells you | What it means when this one is the outlier |
|---|---|---|
| CPM | How much the platform likes your content. Cheaper CPMs typically mean more engaging content | The content is not earning distribution. This is a creative problem |
| Click-through rate and cost per click | Whether the message drives the click | People are seeing it and not acting. The message or the offer framing is off |
| Video watch time and retention | Whether the hook holds attention | The opening is not earning the next three seconds. Rework the hook, not the whole ad |
| Landing page conversion rate | Whether the page, the price and the offer close what the ad opened | The ad did its job. The problem is downstream, and swapping creative will not fix it |
Great engagement with weak conversion is the most common pattern, and it almost always points past the ad. The landing page, the pricing and the offer are enormous leverage points in making an ad work better, or work at all.
Two Creatives at Day Seven
Here is the ladder doing its job. The $50 unit economics are a real example. The creative-level numbers are made up to teach.
Illustrative example
| Creative A | Creative B | |
|---|---|---|
| Cost per purchase | $24.10 | $41.80 |
| Against a $27.70 break-even | Clears it | Misses it |
| CPM | $14.20 | $11.60 |
| Click-through rate | 1.4% | 1.9% |
| Hold to three seconds | 31% | 34% |
| Landing page conversion rate | 2.4% | 0.9% |
| The call | Leave it running | Do not touch the creative |
Creative B looks worse on the only number that decides things and better on every diagnostic above it. Cheaper CPM, higher click-through, stronger hold. The ad is doing more work than Creative A and getting less for it, because the page it sends people to converts at roughly a third of the rate.
Swap that creative and you lose the best-performing content in the account while the actual problem sits untouched. That is what the ladder is for.
An Ad That Is Profitable but Barely Spending Is Telling You Something
Sometimes a creative clears break-even comfortably and simply will not spend. That is not a failure and it is not a reason to kill it.
Usually it means the ad is speaking to a smaller slice of the market that is already further down the funnel, people closer to ready. The audience is real, it is just not big.
The fix: keep it running and test the same style with higher-funnel or more educational messaging. You are looking for the version of that idea that speaks to people earlier, where the volume is.
The Testing System, as Actual Numbers
Most advice tells you to test. Here is what testing means in practice.
| The rule | The number |
|---|---|
| Days before you call a creative, one way or the other | 7 days |
| Active creatives per campaign at any time | 6 to 10 |
| New creatives every week | 4 to 10 |
| Split of new creatives | 20% new ideas, 80% iterations |
That ratio is the part people get wrong. Most of your output should be variations of proven material, not fresh concepts. New ideas are how you find the next winner; iterations are how you get paid in the meantime.
Fatigue Has a Signature, and a Rule That Feels Wrong
Creative fatigue shows up as three things moving together: CPMs rise, viewer retention falls, and cost per purchase climbs. One of those alone is noise. All three is fatigue.
It feels backwards to keep spending on something that is slipping. But the highest spender is carrying the campaign's learning and most of its volume, and cutting it before a replacement is proven takes the floor out from under your results. Replace it by out-performing it, not by switching it off.
The Mistake Almost Everyone Makes
Most brands' creative is too salesy and leans too hard on discounting. Those ads do work. That is exactly why the habit sticks.
The problem is scale. Discount-led, hard-sell creative converts people who already know you and are already close to buying. To grow you need colder traffic, and colder traffic needs different messaging: the demonstration, the problem, the practical questions answered before anyone is asked to buy.
The fix, and the best place to start: take your top performing organic posts, from your own page or from your customers' pages, and run those. They already proved they can hold attention with people who were not looking to buy.
No Format Wins, and Polish Is Not the Variable
There is no universally correct format. Each one has a job.
UGC
Best for relatability, social proof, demonstrations, and showing the product in everyday use.
Studio
Best for premium positioning, product detail, launches, and categories where quality or trust has to be established first.
Founder on camera
Works for product education, origin stories, authority, and ideas that need context to land.
Strong ecommerce brands usually run all three. The concept and the message matter more than whether the content looks polished or casual. iPhone-quality video is plenty good enough to drive millions of views, so start there and spend the money on ideas rather than on production.
The First Three Seconds Is a Principle, Not a Rule
It is real. The opening has to earn attention and stop the scroll. But it does not mean shouting, and it does not mean clickbait.
An opening earns its three seconds with a clear problem, a product demonstration, a surprising visual, a strong claim, or a familiar situation. Hooks can be seen or heard, and the strongest are both at once: a jaw-dropping first line while something visually striking happens, with a sound effect and a text overlay landing together.
Judge the opening on retention and conversion together. A sensational hook can lift views while pulling in the wrong audience, and you will see it in the cost per purchase before you see it anywhere else.
What This Looks Like When It Works
We have had several videos across Instagram, TikTok and YouTube go viral, the best hitting 15 million views, with plenty more in the millions. Those were organic views. Then we put paid spend behind the content that had already proven it could hold an audience, and that generated hundreds of thousands of dollars of profitable sales.
The order matters. The views did not cause the sales. Organic performance selected which content deserved budget, and the paid spend behind proven content is what produced the profitable sales. Some of those videos are on the Tossits case study.
Do This Now
- 01
Write down your break-even.
Average order value minus cost of delivery. Every creative decision is measured against it.
- 02
Rank your live creatives by cost per purchase.
Anything under break-even stays, whatever its other numbers look like.
- 03
Run the ladder on the misses.
CPM, then click-through, then retention, then landing page conversion rate. Find the outlier before you touch the creative.
- 04
Count your active creatives.
If you are under six per campaign, your problem is volume, not quality.
- 05
Pull your top organic posts.
From your page and your customers' pages, and queue them as this week's tests.
Want to run the call on one of your creatives? Paste this into any AI tool:
AI Prompt
Copy this and paste it into ChatGPT or Claude. It will walk you through your own numbers.
I run ecommerce ads. My average order value is [X] and my cost of delivery per order is [X] (product, shipping, returns, fees). For one creative over the last 7 days: cost per purchase [X], CPM [X], click-through rate [X], three-second hold [X], and the landing page it sends to converts at [X]. Tell me my break-even cost per customer, whether this creative stays or goes, and which rung of the ladder to investigate first.
Common Questions
Ecommerce Ad Creative FAQ
How long should I run an ad before deciding whether it works?
Seven days. Judge it on cost per purchase against your break-even, which is your average order value minus what it costs to deliver the order. If it comes in under break-even, leave it alone, whatever its other numbers look like. If it misses, work down the ladder (CPM, then click-through rate, then retention, then landing page conversion rate) before you swap the creative.
Why does my ad get plenty of clicks but few sales?
Great engagement with weak conversion almost always points past the ad. The landing page, the price and the offer are the usual suspects, and the ad and the page need to carry the same product, promise, price and offer. Check the landing page conversion rate before you touch the creative, because swapping a creative that is earning clicks removes your best content and leaves the real problem in place.
How many ad creatives should I run, and how many new ones each week?
Keep six to ten creatives active per campaign, and launch four to ten new ones every week. Make roughly 20% of them new ideas and 80% iterations of what is already working. If you are under six active creatives in a campaign, your problem is volume, not quality.
Should I turn off an ad when its performance drops?
Not until another ad is doing better, and never the highest spending ad. Fatigue is CPMs rising, retention falling and cost per purchase climbing, all three together. The answer is to launch more creatives, because the highest spender is carrying the campaign's learning and most of its volume. Replace a tired winner by out-performing it, not by switching it off.
The Bottom Line on Ecommerce Ad Creative
Ecommerce creative is built to sell a specific product, so it gets judged on cost per purchase against your break-even and nothing else. Seven days, six to ten live, four to ten new a week, and when one misses, work down the ladder before you blame the ad. Fatigue is CPMs up, retention down and cost per purchase up, all three together, and you replace a tired winner by beating it rather than by switching it off.