Newsletter Digest

We asked 3,000 US consumers to rate 18 food and beverage items on two dimensions: How natural does adding protein feel? And how interested would you be in trying a high-protein version? The results organize neatly into what we're calling the Protein Feelings Matrix.

The obvious winners are obvious. Pasta, cereal, snack bars, and milk score high on both naturalness and interest. These are established categories for high-protein offerings and should continue to thrive. Kraft just launched PowerMac, a mac-and-cheese SKU with 17g of protein and 6x the fiber of the classic blue box — a move that makes complete sense given the data.

The clearest gimmicks are clear, too. Beer and ready-to-drink cocktails ranked as the most unnatural fit for protein and drew the lowest interest. These won't take over the world.

The interesting frontier is the middle. Chocolate, ice cream, chips, popcorn, water, and coffee all cluster near the neutral baseline — but protein water and coffee are actually polarizing, not indifferent. Equal forces of very positive and very negative ratings cancel out to near-zero averages. That's where brands will have to innovate and communicate hard to win.

Younger consumers (Gen Z and Millennials, ages 15–45) are significantly more receptive across every category. GLP-1 users — the growing population taking Ozempic, Mounjaro, and Zepbound — also rated items higher across both dimensions, making sense given that the medications can cause muscle loss. Starbucks customers over-indexed on protein coffee and cold foam. Sweetgreen customers were the most protein-curious of all, with Whole Foods shoppers close behind. Walmart shoppers hovered right at the overall average.

Google search volume for "protein" hit all-time records in the US and globally last month. More than half of the people trying to get more protein say they started in the past year. This trend is accelerating, not plateauing — and the brands that figure out which middle-category products can actually taste good will find the next breakout opportunity.

Memories.ai, founded by former Meta Ray-Ban glasses engineers Shawn Shen and Ben Zhou, is building what it calls a "visual memory layer" — infrastructure that lets wearables and robots store, index, and search first-person video footage so they can remember what they've seen.

The company is partnering with Nvidia, tapping Cosmos-Reason 2 for vision-language reasoning and the Metropolis stack for large-scale video search. It has raised $16M to date, including an $8M seed round in 2025 and an $8M extension led by Susa Ventures.

The core argument: AI memory tools from OpenAI, xAI, and Google are largely text-focused. Robots operating in the physical world need to store and recall what they actually "see." Memories.ai is already working with major wearable makers and sees even bigger opportunities as humanoid robots begin deploying at scale.