Illustrative arithmetic, not a customer result. Take a brand at $80,000 of monthly revenue spending $22,400 on ads. TACoS is 28%.
Step 1: find where the spend actually goes
Aggregate TACoS hides distribution. If three SKUs sit above 40% TACoS and four sit between 8% and 12%, the blended 28% describes none of them. Pausing or rebuilding the three worst removes their spend without removing much revenue, because those SKUs were converting poorly in the first place.
Step 2: watch the denominator, not the numerator
Hold spend flat at $22,400 and grow total revenue to $140,000 and TACoS lands at 16% on its own. Trim spend to $19,600 at the same revenue and it is 14%. Neither move required a better bid; both required total sales to grow faster than advertising.
Step 3: the point of the metric
TACoS improves when organic share rises, and organic share rises from ranking, reviews and conversion rate rather than from bidding. That is why TACoS is a business metric and ACoS is a campaign metric. Product-level visibility is what tells you which of your SKUs is in which situation.