What are you optimizing your Amazon ads for?
By Liranhirschkorn · September 4, 2026 · Curated by George's Blog
What are you optimizing your Amazon ads for?
For most brands the answer is a low TACOS.
Last week an account came across my desk running a 5% TACOS. At first look it's one of the cleanest accounts I've seen. Profitable campaigns, and hitting its target consistently.
It's also leaving about $250k a month in sales on the table.
One of their core non-branded keywords converts at 33%. The campaign running it hits its daily budget by mid-afternoon and shuts off for the rest of the day. Same pattern across a dozen of their most profitable campaigns. Every day at around 1-3pm the account stops taking new customers who already decided to buy.
A TACOS target only measures one side of the equation. It tells you how much you spent relative to sales. It doesn't tell you how many sales you didn't get because you stopped spending. A 5% TACoS on $2M a month and a 7% TACOS on $2.5M a month are both "good" numbers, but one of them is a much bigger business.
Its obvious a TACOS that's too high is a problem. But a TACOS that's too low is the same problem in the other direction. It just doesn't show up as a red number on any dashboard, so nobody flags it.
When you set ad KPIs, the target should be a range, not a ceiling. Measure profit as TACOS increases and you have a reason to spend more when the data supports it instead of a reason to stop.
A few things to check on your own account:
Budget-capped campaigns that are profitable. Filter for campaigns hitting their daily budget while beating your targets. Especially non brand keyword cmpaigns. Every one of them is demand you chose not to fund.
Conversion rate by search term. Anything converting at "high" numbers (20-40) deserves a higher bid test, regardless of what the CPC looks like (especially consumable).
Total sales trend vs TACOS trend. If TACOS has been flat or falling for a year and total sales have too, the efficiency isn't a win.
Moving this account from 5% to 7% TACOS costs real money. It also adds around $3M a year in sales, and that's the number the P&L cares about.