What I'm Reading, Saving, and Rethinking - August 1st, 2025

Practical marketing insights from the trenches: summarized, questioned, and ready for action.

Author

Adam Lasky
August 01, 2025

Happy Friday! Every week, I save dozens of posts, articles, and newsletters that challenge my thinking. Here's what stood out this week and why I think it's worth your time too.

Was this email forwarded to you?

🔥 This Week's Saves

Google's Measuring Effectiveness
Still timely Google’s 2019 “Three Grand Challenges” paper identifies three core obstacles to marketing measurement.

The Platform Wars Begin
Shopify and Amazon are barring AI shopping agents from their platforms, kickstarting a showdown where each platform fights for control over AI-assisted shopping.

Confirmed: ChatGPT uses Google SERP Snippets for its Answers
ChatGPT relies heavily on Google SERP snippets as its data source, revealing that Google's index acts as a fallback for ChatGPT's answers.

TL;DR: The biggest hurdles in marketing measurement are proving incrementality (causation, not correlation), measuring long-term impact with short-term data, and building unified methods that reconcile experiments, attribution, and marketing mix models. The report stresses the importance of randomized controlled experiments as the gold standard, the need to project lifetime value and brand effects into current decision-making, and the opportunity to blend multiple methods (including structured expert judgement) into a coherent, decision-ready framework.

My Thoughts: The core message is that most marketers are still mistaking correlation for causation and chasing short-term metrics at the expense of long-term growth. The real opportunity is to build systems that prove incrementality, tie present actions to future value, and unify siloed measurement approaches into one narrative executives can trust. In practice, this means leaning on experiments where possible, using advanced modeling (like causal diagrams and LTV projections) to fill gaps, and communicating uncertainty in ways that drive confident decisions rather than paralysis.

One Quote to Chew On:

"Today's advertisers can struggle to balance brand-building with sales activation. Some observers believe that the shortening tenures of marketers have resulted in more focus on quarterly or annual targets."

Measuring Effectiveness

TL;DR: Shopify updated its robots.txt to explicitly prohibit automated AI agents from checkouts, signaling a "humans only" shopping stance. Amazon blocked Google's shopping agent while promoting its own "Buy for Me" AI checkout feature. This marks the start of a major platform battle over who gets to own the AI shopping experiences. Businesses now face a fragmented AI ecosystem that may force them to adapt across platform-specific AI systems.

My Thoughts: This is both fascinating and a bit terrifying. On one hand, the idea of a universal AI shopping agent that compares across platforms sounds like an efficiency dream. On the other hand, that fragmented future where each platform requires its own agent strategy feels like a nightmare for lean sellers. And it reeks of a fortress mindset: platforms are desperate to lock down control, betting they can monetize AI shopping better than any third party. If they succeed, the power stays with them. But if they misread how open AI ecosystems evolve, they might side-step relevance altogether.

One Quote to Chew On:

"What we're witnessing is basically every e-commerce platform having the same existential crisis: 'What if these AI agents turn us into glorified vending machines?'"

TL;DR: Aleyda Solís created a brand-new webpage and did not submit it for indexing. ChatGPT (using web search mode) initially claimed the page didn't exist. After submitting it to Google Search Console, the page appeared in Google SERPs. ChatGPT then produced an answer matching Google's exact snippet. ChatGPT identified that it fetched a "cached snippet via web search," clearly drawing from Google's public snippet rather than crawling the live page itself. The implication is that Google's index remains a crucial source for ChatGPT's response engine.

My Thoughts: ChatGPT isn't a creator. It's a pattern recognition machine. The recent Google snippet test proves it doesn’t "find" answers but mirrors existing structures: once Google generated a snippet, it echoed it back almost verbatim. This shows it's less a researcher and more a remix artist, stitching together patterns that already exist rather than generating new insights. For brands, that means success in AI search isn't about hoping ChatGPT understands you, it's about making sure the right patterns exist to be recognized: clean schema, strong snippets, authoritative backlinks, and consistent messaging.

One Quote to Chew On:

"ChatGPT continues to provide an answer based on a Google Search Results snippet (one from a different query related to the page than before and is now using the meta description unlike before)."

This newsletter is for you. What marketing challenges are you facing in your marketing journey? Reply directly to this email with your questions or topics you'd like to see covered in future issues.

Until next week,

P.S. Found this helpful? Forward it to another founder who might benefit. We're all in this together.

P.P.S. Don't forget to download the Growth Marketing OS by clicking the button below.