I'll be honest, I used to jump straight to the fix.
By elizabeth-greene-junglr · September 17, 2026 · Curated by George's Blog
I'll be honest, I used to jump straight to the fix.
Someone would come to me and say, "ACoS is spiking, we're losing top of search, what do we do about placements?" And I'd start optimizing placements.
Because that's what they asked for.
That's what the symptom pointed to.
It took me longer than I'd like to admit to realize I was skipping the most important step.
Now when someone tells me rankings are dropping and ACoS is spiking, the first thing I want to know is the date.
When did something change?
Because I need two snapshots.
The before and the after. Side by side.
And honestly, that comparison changes everything.
I had a brand come to us with an entire analysis.
"The issue is we're losing top of search. We need to segment, we need to do all this stuff."
And I'm looking at it thinking, the goal is figuring out why velocity dropped.
So you pull the two time periods.
You find the product that's actually having the issue, because very rarely is it every single product in the catalog and every single campaign that stopped working.
It's usually one or two products. Specific campaigns.
And then you ask the real questions.
ACoS is just a division problem, spend over sales.
If spend went up and sales didn't, you're looking at whether CPCs spiked or clicks came in without converting.
If sales dropped and spend stayed the same, you're looking at conversion rate, and you want to find which campaign had the greatest ad sales volume drop.
Find that campaign and you find the actual thing that broke.
It could be inventory.
It could be competitors running coupons, making you less competitive on the search page.
Or a new competitor.
Or a category-wide drop where your BSR hasn't actually moved relative to everyone else.
The point is, if you skip the comparison and go straight to "increase top of search on every campaign," you're going to destroy your profitability.
You're going to make ACoS spike even more.
You're solving for a symptom you haven't actually confirmed.
Still working on being more patient with that process, to be honest.