An Amazon private label seller came to us with a familiar problem: ad spend was climbing, sales were flat, and ACoS had crept up to 48%, well past what the margins could sustain.

The Diagnosis

The account had a single sprawling campaign mixing broad match, phrase match, and auto-targeting with no meaningful separation. There were no negative keywords added in over six months, which meant budget was leaking into irrelevant search terms daily. Bids were set once at launch and never adjusted against real performance data.

The Restructure

We rebuilt the campaign structure from scratch: separate campaigns for exact match, phrase match, and broad/auto discovery, each with its own budget so high-intent exact-match traffic wasn't competing against exploratory broad match for the same dollars. We ran a two-week negative keyword harvest, pulling irrelevant search terms out of auto and broad campaigns before they could keep draining spend. Bid automation rules were set to adjust based on placement performance rather than manual guesswork.

The Results

Within 90 days, ACoS dropped from 48% to 22%, and overall revenue doubled without increasing total ad spend, the budget was simply being spent on the right search terms instead of scattered across everything.

The Takeaway

Most inflated ACoS problems aren't a bidding problem, they're a structure problem. Fixing the architecture of a PPC account before touching individual bids is almost always where the real savings come from.