AI Is the New Operating System for eCommerce Growth

Performance is at risk if you continue to buy media as if it were 2022, ship a few dozen creatives per month, and hope that algorithmic luck finds you. Today’s ad auctions reward volume and variation, not vanity. The brands pulling ahead are the ones feeding machines more options, faster, and letting models do what they do best: match the right message to the right person in milliseconds.

We treat AI as infrastructure. It powers how we create, publish, measure, and now, how we capture demand from AI search and agents. On the eCommerce Coffee Break podcast, I recently talked about the blueprint we’re using with modern commerce brands. Changes in Media Buying: Win with Creative Velocity New ML systems on channels like Meta and AppLovin perform best when they can choose among hundreds of creative variants, not dozens. If you launched 50 ads last month, the question is: how do you launch 500–1,000 this month without burning out your team? The answer lives outside the ad platforms. Through platform APIs, we combine AI creative iteration tools with third-party bulk publishing, allowing us to launch and learn in minutes rather than hours. The account structure should be simple, true experiments should be isolated, and volume should drive learning.

What it looks like in practice: in a recent engagement, we converted thousands of TikTok Shop affiliates into Meta-whitelisted influencer handles. From there, we used our bulk-publish stack to deploy native ads from those pages across Meta, YouTube Shorts, AppLovin, and Reddit (in the near future). The result: a step-function increase in daily revenue, higher efficiency at 3–4× spend, and a system that comfortably ships 1,000+ ads per month.

The new buzzword is AI search, and it’s already bringing in money. Search is shifting from blue links to answers. As reported by The Wall Street Journal, an estimated 5.6% of U.S. search traffic on desktop browsers in June 2025 went to an AI-powered large language model like ChatGPT or Perplexity. If you’re not visible to these LLMS, you will feel the drop in traffic before you understand why.
We built the AI Visibility Index (AIVx) to benchmark brand presence across ChatGPT, Perplexity, and Google’s AI Overviews by prompt coverage, model-by-model differences, and sentiment. It’s a practical compass for where and how to show up and what to fix first.

Reality check on attribution: traffic recommended by LLMs often lands as “direct” or “branded search” because many models cite third-party sources (e.g., listicles, Wikipedia, Reddit) rather than linking to your PDPs. Make sure that your post-purchase survey specifically mentions “Perplexity/AI Chat” as a source to reveal the impact, as we consistently see those sessions convert at a higher intent than non-branded SERP clicks. Rethinking retention: From AI Salespeople to Flows Customers are unaware of the boilerplate replenishment flows and batch-and-blast SMS. We’re testing AI-driven conversational agents across iMessage and WhatsApp that behave like a high-touch salesperson, checking in on fit, surfacing the right offer, and closing with context. The result is a more human experience and better conversion.

Our Engagement Model: Conduct an Audit First, Then Create a Plan Prior to diagnosis, we do not recommend any strategies. Every engagement starts with a full-funnel audit, starting with growth history, forecast, channel mix, new vs. returning revenue, then a sequenced roadmap. In weeks 2–4, we rebuild the account structure, deploy new creative at scale, and implement server-side tracking (CAPI) so learning compounds. The entire strategy will be operational by the end of week four. Who wins with us: brands with the budget to staff a cross-functional team (search, social, influencer, creative, PR/SEO) and a deep bench of raw creative. More inputs → more personalization → better auction outcomes.
The Dangers of Being Still If you’re shipping “tens” of creatives while your competitors ship “hundreds,” the auction will under-personalize your delivery, and your performance will decay. If you ignore AI search while Gen Z moves research into chats and agents, you’ll cede top-of-funnel word-of-mouth to others.