How to Scale Creatives With Metrics: Sales First
Learn the metrics hierarchy for deciding which creatives to scale: sales, cost per sale, checkout, and secondary metrics. A practical optimization guide.
The order metrics matter in (and why almost everyone gets it wrong)
Sales come first. Always. Then how much that sale cost. Only after that do you look at checkout, CPC, hook rate. That's the hierarchy that decides which creative moves to day two and which one dies on day one. Flip that order and you'll scale a pretty creative with great secondary metrics that doesn't pay the bills.
Anyone who runs traffic knows the feeling: you want to scale the creative with an insane hook rate and low CPC. But hook rate doesn't pay your supplier. Sales do.
The math is simple. You launch 10 creatives, let them run day one inside the niche benchmark, and at the end of the day you sit down and read what happened. Whatever sold gets absolute priority in the decision. Everything else is context.
Sales and cost per sale: the criteria that call the shots
A creative that sold on day one already earned the right to run day two. But selling isn't a free pass to scale. These are two different decisions.
Moving to day two: it sold, it advances. Done.
Moving to pre-scale: now the bar changes. The creative has to be inside your niche rates. Cost per sale and cost per checkout need to match the benchmark, or scaling turns into a loss.
Why? Because scaling isn't just bumping the budget on the same ad. In practice, here's what happens: the media buyer duplicates the campaign five times to do an initial pre-scale and gauge the potential. If the creative already sells tight at normal volume, at multiplied volume it either loses money or breaks even.
Affiliate example. The ad is selling at 1.1 ROAS on day two. Looks like everything's fine, it's selling. But for scale to work you'd need a 1.3, 1.4 ROAS. If it's already at the limit right now, at current volume, there's no cushion to absorb the natural drop in efficiency that comes when you multiply the budget. It has no scaling potential. Did it sell at a profit? Yes. Should it go to scale? No.
This is the part that fools people. Selling at a profit and having scaling potential are two different things. One thing is the creative paying for itself. Another is it surviving five copies running without sinking.
Secondary metrics: when initiate checkout saves a creative
Checkout, CPC, and hook rate come second. They're for reading the pattern of your winners and for giving a second chance to a creative that didn't sell.
The classic case: the creative didn't sell on day one, but it had a really strong initiate checkout. A strong IC means the person got all the way down there, showed buying intent, and just didn't close. It could have been a variable you can't see: payment processor, time of day, something in the funnel. The potential is still there.
That creative deserves day two even without a sale. Not for the sale that didn't come, but for the signal that the offer landed.
Now the cut. Anything way off the benchmark dies on day one, no ceremony:
- A loose IC but a sky-high cost per checkout
- CPC through the roof, way above the niche rate
- No signal that the audience engaged with the offer
These don't move to day two. Period. Letting them run is burning budget to confirm what the numbers already told you.
When does a creative move to day two?
Two ways in.
First: it sold. It sold, it advances. Doesn't matter if the other metrics are so-so, if a sale came in within benchmark, it runs again to gather more data.
Second: it didn't sell, but it has a really interesting secondary metric (the strong IC I mentioned). Here you're betting on unconfirmed potential.
What doesn't advance: the creative that's off benchmark across the board, no sale and no sign of intent. That one you cut and use the budget to test a new ad.
And there's a philosophy behind this: for me it makes more sense to test more ads and focus on the best ones than to insist on a mediocre creative just because it sold once. Test volume beats stubbornness with a tight creative.
This is where the operation starts demanding structure. Testing more ads means launching more campaigns, more variations, every day. Doing that manually in Ads Manager bottlenecks on time and repetitive config errors. Operators running high volume across multiple BMs tend to lean on tools like DirectAds' bulk upload flow with standardized naming, which launches campaigns in mass and keeps the structure consistent without all-nighters in Ads Manager.
How day two decides the scale
Day two is for confirming. You let it run to gather more metrics and see if day one was consistency or luck.
Sold on day one, sold on day two, with cost per sale inside your rate? Now you barely even look at the other metrics. What matters is settled: it sells, and it sells with enough margin to survive scale.
Sold both days but tight? Here you hold. Go back to the ROAS math: if the current margin leaves no room for the efficiency drop that scaling brings, the creative stays at normal volume. It pays for itself, but it doesn't scale.
Pattern reading comes in here too. When several winners share the same strong secondary metric, you start to understand what makes a creative work in your niche. That becomes fuel for the next batch of tests. But it's backstage analysis. The decision to scale stays anchored in sales and cost per sale.
Takeaways
- Order your metrics: sales first, cost per sale second, checkout and CPC/hook rate after. Don't flip it.
- Separate the two decisions: selling earns day two, but only hitting the niche's cost per sale and cost per checkout earns scale.
- Save the creative with a strong initiate checkout that didn't sell. Cut the one that's off benchmark across the board.
- Before scaling, check the ROAS cushion. Selling at 1.1 ROAS won't survive five copies running.
Frequently asked questions
Should a creative that sold at a profit always go to scale?
No. Selling at a profit earns day two, not scale. To scale, cost per sale and cost per checkout need to be comfortably inside your niche rates. If it sells tight, duplicating the campaign several times drops efficiency and the profit disappears.
Why look at initiate checkout if sales are what matter?
Because a high IC means the audience got close to buying and just didn't close, maybe because of a variable outside your control. That signal justifies giving a second day to a creative that didn't sell, betting the potential gets confirmed.
What's the minimum ROAS to scale as an affiliate?
It depends on the niche, but the logic is having a cushion above breakeven. If you need 1.3 or 1.4 ROAS for scale to work and the creative already runs at 1.1 at normal volume, it has no margin to absorb the efficiency drop that scaling brings.
What makes a creative get cut on day one?
Being way off the niche benchmark: sky-high CPC, absurd cost per checkout, no sale and no sign of buying intent. In those cases day two isn't worth it. Cut it and reallocate the budget to new tests.
