Case study 03
Rebuilding Amazon Ads on its own terms — not as a copy of Google
How I took over a product that was built as a replica of Google Ads, diagnosed what was fundamentally wrong with that approach, and rebuilt it around how Amazon advertising actually works.
60%↑ onboarding
Increase in onboarding completion for Amazon advertisers
50%↑ adoption
Increase in tool usage and implementation
Ground up
From Google replica to Amazon-native tooling
Product Lead
Shipped, evolving
Amazon Ads
Sponsored Products
E-commerce
The problem
Amazon Ads was being built as if it were Google Ads — and it isn't
When this product area was handed to me, every tool built for Amazon Ads was a replica of the equivalent Google Ads tool. The terminology was wrong for Amazon advertisers. The workflows didn't reflect how Amazon campaigns are actually managed. Many tools didn't meet the quality bar of the rest of the platform. And beneath all of it was a deeper technical problem: the Amazon Ads API behaves fundamentally differently from Google's.
The result was a product that confused Amazon advertisers, produced unreliable data, and couldn't be trusted to make decisions from. Adoption was low — not because the market didn't exist, but because the product wasn't built for it.
The core insight
Amazon is not a search engine. It's an e-commerce ranking system — and ads are part of that ranking.
Google / Microsoft
Paid and organic are separate
Ad efficacy is measured against results driven directly by the sponsored placement. Organic ranking is an SEO metric, managed separately. The goal is paid traffic ROI.
Amazon
Paid ads improve organic ranking
Strong sponsored ad performance drives up a product's overall ranking — including organic. The real goal is product ranking position. Nobody buys the 25th result.
This distinction changes everything about how you measure success, structure campaigns, and what you optimise for. On Amazon, total sales volume relative to ad spend — not just ad-attributed sales — is the right metric.
Technical constraint
The API problem that made everything harder
Amazon Ads APIs are not real-time. Unlike Google's APIs which can be queried on demand, Amazon's are slow, rate-limited, and not designed for frequent polling. New campaign entities would have data for only a few weeks, while older entities had months of history — creating an imbalance that made performance comparisons meaningless.
Working collaboratively with engineering, I defined the requirements for a caching architecture: account data is downloaded daily, stored in our own data layer, and served from there — decoupling the product experience from the API's latency and rate limits entirely.
Strategy
A four-phase rebuild: evaluate, fix foundations, reach parity, then differentiate
1
Evaluate and map
Full audit of existing tools against what Amazon advertisers actually need. Three categories: keep as-is, fix, and build from scratch.
2
Fix the foundations
Overhauled onboarding, fixed broken selection experiences, corrected terminology to be Amazon-native, rebuilt the caching layer.
3
Reach meaningful parity
Rebuilt tools that had been Google replicas into Amazon-appropriate equivalents — same capability category, redesigned around Amazon's structure and workflows.
4
Differentiate
Build capabilities that place this product ahead of other providers — built around the organic/paid interplay insight most tools in the market still don't fully address.
Impact
60% better onboarding. 50% more tool adoption. A product worth building on.
Onboarding completion for Amazon advertisers increased by 60% — a direct result of fixing broken experiences and making the product feel native to how Amazon advertisers think. Tool usage and implementation increased by 50% across the board.
The product now has a coherent foundation to build on. The differentiation work in phase four is ongoing — but it's being built on a product that actually works, with a clear strategic point of view about what makes Amazon advertising different.
This is still evolving. The foundation is solid; the differentiation roadmap is what makes it competitively interesting going forward.