Sample34%
Lower ACoS
US supplement brand
Crowded generic searches cut, ingredient searches scaled.
Read case study01 · The challenge
Where it started
The brand came to us with a 38% ACoS on campaigns that ignored buyer intent. Spend was going to searches that never sold, at hours that rarely sold.
Brand, category and research searches sat in the same campaigns, so one bid had to fit every kind of shopper. The fix was structure, not a bigger budget.
Day 0 numbers
- ACoS38%
- ROAS2.6
- Terms with spend, no sales312
- Campaign set-upMixed intent
What the audit found
- Campaigns ignored buyer intent
Brand, category and research searches shared campaigns and bids.
- 312 search terms with spend and no sales
Budget went to searches that had never turned into an order.
- Budgets spread evenly across the day
Spend ran through the hours that rarely sell, leaving less for the hours that do.
02 · The approachSample timing
What we changed
5 steps, in the order they went live.
- Week 1
Audit every search term
We read the search term reports line by line and sorted each search by buyer intent: brand, category, competitor and research.
- Weeks 1-2
Rebuild campaigns around intent
One goal per campaign. Exact match for the searches that sell; phrase and broad match only to find new ones, each with its own bid.
- Weeks 2-4
Cut 312 wasted search terms
Every term that spent without a sale went in as a negative, so the budget moved to searches that convert.
- Weeks 5-8
Move spend to the hours that sell
Bids went up in the hours that convert and down in the hours that rarely do, so the budget lasted the whole selling day.
- Weeks 9-13
Scale what sells
With ACoS holding, budget moved to the campaigns that sell. Every change was checked in the weekly report.
03 · The results
What moved
The main numbers at each checkpoint, then the full before-and-after.
- Measured
- Sample checkpoint
| Metric | Day 0 | Day 90 | Change |
|---|---|---|---|
| Monthly sales (index) | 100 | 240 | +140% |
| ACoS | 38% | 24% | −14 pts |
| ROAS | 2.6 | 4.2 | +58% |
| TACoSSample | 14% | 9% | −5 pts |
| Search terms with spend, no sales | 312 | Blocked | 312 negatives |
04 · The honest partSample
What did not work
Our first dayparting pass cut bids too hard after 9pm, and late-evening orders dipped for a week. We rolled the cut back and moved to smaller weekly steps, checked against each day’s orders.
05 · Client voice
In their words
We’ve been working with Amplivus for over a year. Sales increased, ACoS improved noticeably, and their experience really shows.
06 · Method
How we measured
- Source
Seller Central and Amazon Ads reports for the dates shown.
- Baseline
Day 0 numbers taken before any change went live.
- Total sales
TACoS and total sales reported, not just ACoS.
- Approved
Published with the client’s written permission.
Results depend on category, margin, stock and starting point. Your free audit shows what is realistic for your account.
