If you were forwarded this newsletter, and you'd like to receive it in the future, subscribe here.​

​Finding the Midnight Snacker : Agentic Marketing​

“Feel Good Fun Mix” tops my recommendations on Spotify. Spotify has created over 6,000 such labels by hand.

Amazon and Netflix attributed 35% and 75% of their revenue to their recommendation systems. This is a profound & counterintuitive shift in how we think about marketing.

In a recent case study led by Aampe, an AI agent with this type of segmentation sent far fewer messages than traditional systems while achieving better results.

Even more striking, the AI agent learned to offer smaller discounts (15% vs. 35%) to users showing higher purchase intent, automatically optimizing for both conversion and profit margin. Streak marketing, urging users to read a few pages each day, for example, works wonders too.

Today’s marketers are limited in their ability to construct a vast number of user segments - for example, the midnight snacker who prefers to receive texts very late on a weekday isn’t an obvious cluster. A barber shop found nearly 400 different user segments in their population.

Aampe is building the future of marketing - agentic marketing.

Aampe creates an agent for each user. The agent understands point of sale history (buying a coffee every morning at 745am), mobile app usage (does she pre-order before she arrives?), and engagement with previous messages.

By combining all this - ad engagement, surveys, & core business outcomes - Aampe produces effects on performance that are significant & sustained.

At the core of the product, reinforcement learning explores what a person likes - the message, the medium (email/text/notification/GIF), when they like to receive it & how these change with time - something not possible with the marketing canvases of today.

This emerging approach empowers teams to create personalized lifecycle journeys for each user across millions of users. They serve more than 50m monthly today. We’re announcing our partnership with Aampe by leading their Series A.

Led by a team of human behavior researchers at Harvard, US Army, & NATO, Aampe publishes their learning about agentic marketing here.

​