Digital Shelf Insider

The episode touches several relevant areas (agentic commerce, AEO, retail media attribution, omnichannel SKU profitability) but stays surface-level on each, rarely going beyond the obvious industry anxiety. A handful of genuine observations - LLMs ignoring images due to compute cost, the need to migrate converting copy into images - lift it above pure filler, but the conversation meanders and repeats itself without compounding into dense takeaways.

“the text needs to talk to not just the algorithm, but these LLMs...if your keyword stuffing, it's seeing the same thing over and over again and it doesn't care anymore”

“people aren't reading the text, they're looking at the images to convert. And so you have to move the, converting text from that, nobody's really searching on, from the written text into, into the images”

The dual observation that LLMs skip images due to energy cost while consumers now primarily use images to convert - requiring brands to split their content strategy - is a mildly non-obvious framing. Most other points (Bezos 'your margin is my opportunity,' agentic commerce anxiety, copy-paste Walmart mistakes) are well-worn industry talking points with no fresh angle.

“these LLMs, they are not...they're not really reading the images. It takes too much energy, too much power, too much money at the end of the day to read the image”

“remember Jeff Bezos came out and said early on, your margin is my opportunity. And people didn't take him seriously on that”

Aaron Konan is a credible connector and advisor who aggregates signal from a genuine volume of brand conversations (20 - 30 per week), but his role is networker and podcast host rather than operator who has scaled a brand P&L or run a major retail media program. He offers useful synthesis but no deep firsthand operational authority.

“I talk with 20 to 30 brands a week and uh, I get the, uh, I don't know, the privilege of kind of like consolidating all that knowledge and then resharing it out with the Network”

“I started off as a chemist by myself at the back of a lab and. And uh, then here we are chatting”

A few concrete anchors add real value - the Sensor Tower Rufus Black Friday stat and the 40+ touchpoints figure are usable data points - but the episode largely lacks named brand case studies, dollar figures, specific platform policy details, or timeline benchmarks that would let a practitioner act on the information.

“sensor tower, um, data that came out was like Black Friday. 30% of shoppers used Rufus and it ended up having a 60% higher conversion rate”

“the seven touch points, you know, in marketing before somebody buys, now that's well over 40”

The host asks broadly reasonable questions across multiple topics but never pushes for specifics, never challenges a claim, and frequently accepts vague answers with 'Yeah, yeah, yeah.' The guest ends up steering the conversation twice by asking 'what else is top of mind?' - a sign the host ceded control. The mid-episode sponsor read further disrupts momentum.

“what else is top of mind for you? This is fun. This is fun.”

“I kind of bugged you with all sorts of questions. It's from different, different areas. It was just fun.”