An independent read of the Amazon Ads account — structure, spend, and where performance is and isn't coming from.
The short version: the account isn't missing tactics — almost everything is already here. The issue is maintenance. The clearest single example: roughly 40% of search spend sits in the worst-converting ad placement while the best-converting one is left underfunded and unmanaged.
Across Sponsored Products, Brands, Brands Video and Display, the account has accumulated campaigns since 2023. The breadth is genuinely comprehensive. The problem is that activity is concentrated in a tiny, under-tended core while the rest sits idle.
The return on spend is real — but look at where it comes from. The profitable spend is almost entirely defensive: ads placed on Marcy's own product pages to hold the space, plus automatic campaigns. That protects the brand. It does not expand it.
Own-ASIN (product) targeting returns roughly 4x–16x on spend across the strongest campaigns. This is the engine — and it's well placed.
Auto targeting on the core range pulls solid, mid-single-digit ROAS with minimal hands-on effort. A reliable base.
The manual keyword campaigns — the ones meant to win new customers — convert at close to zero. Of the live keyword campaigns, most returned no profitable sales in the window; the occasional sale reads as incidental rather than earned. In short: the wins are defence, not growth.
None of the following are exotic. They're the fundamentals of a managed account — and they're absent in the campaigns currently taking budget.
Split the same spend by where the ad showed, and the inefficiency becomes a clear, low-risk opportunity — one that requires no price change and no new creative.
| Ad placement | Share of search spend | Converts at | Return on spend |
|---|---|---|---|
| Top of Search | ~29% | Highest — ~20× Rest of Search | 4.1× |
| Product pages | ~33% | Mid — carries the defensive work | 7.3× |
| Rest of Search | ~39% | Near zero | 0.7× ▼ |
The largest share of spend is going to the weakest placement. In the single best campaign, half the budget went to Rest of Search and returned nothing, while the other half — at Top of Search — drove most of the account's orders.
A small set of search terms produce sales while the account captures only a sliver of their available impressions — in some cases under 2%. The demand is there; the account is barely showing up for it.
Shift spend toward Top of Search and the converting terms, set the placement adjustments that are currently switched off, and let Product Pages keep doing the defensive work. Same budget, better placed.
Net: roughly £1,000 is currently spent in the placement that converts worst. Redirecting it toward the placement that converts ~20× better is the cleanest efficiency gain available — and it sits entirely within day-to-day account management.
It's worth being clear about what the account already has — because the opportunity is in running these well, not in adding them.
| Capability | Status in account |
|---|---|
| Sponsored Brands Video | Present — ~19 video campaigns built (mostly paused) |
| Category targeting | Present — 80+ category targets across Brands & Display |
| Display remarketing & audiences | Present — 20 remarketing campaigns, view & purchase lookbacks |
| Placement bid adjustments | Present — set on a portion of campaigns (inconsistent) |
| Defensive / brand-shielding | Present and performing |
The recurring theme: these were set up, then left. Built once, not maintained.
This is not an account that needs more campaign types bolted on. It's an account that has been allowed to sprawl — hundreds of campaigns deep, four-fifths dormant — while the live spend runs without the routine discipline that makes Amazon Ads efficient: negatives, match-type structure, value-based bidding, focused keyword sets, and intent-matched targeting.
The fastest gains aren't new tactics. They're pruning, restructuring, and steady ongoing management of what's already here — starting with the placement and bid discipline that would move ~£1,000 of monthly-scale waste into the spots that actually convert.