AMC playbook: Enhanced scoring for overlapping audiences
In short: The playbook combines AMC analyses with Amazon's persona builder API. For your own audiences it shows which other audiences, retail categories, demographics and Prime Video habits overlap. Every result comes with an affinity score and a percentage, so the strength and the spread of an overlap can be compared.
What it is about
According to Amazon the persona builder and AMC each have their own APIs and run independently; the playbook connects the two. An affinity of 5 means customers in the persona are five times more likely to belong to this audience than the average Amazon customer. A percentage of 5 means 5 percent of the customers in the persona belong to it.
Which questions the playbook answers
- How do you use the persona builder API endpoints with AMC results?
- How do you assess affinity and percentage for AMC audiences?
- Which audiences with high overlap are not part of the campaign strategy?
Flow
- Start with an exploratory query in AMC, for example on the performance of past campaigns.
- Pick audience segments from the results for more detailed scoring.
- Use the audience discovery API to get the canonical IDs of the segments.
- Query the persona builder API with these IDs: overlaps, demographics, category interests, Prime Video streaming.
- Create the audience with
POST /amc/audiences/queryand use it in Amazon DSP or sponsored ads.
Prerequisites according to Amazon
- AMC concepts and AMC SQL, basic Python and basic Amazon DSP knowledge.
- Onboarding to the API and working knowledge of the AMC and persona builder APIs are presumed.
Our take
The value lies in the question which audiences you do not address yet. Check beforehand that the resulting audience reaches the minimum size for rule-based audiences, otherwise creation fails.
Amazon documents the playbooks in technical English with SQL and Python. This page summarises what the playbook does; the queries and scripts are in the original.