Industry Perspectives
What operators actually said.
The eight tiers came from data on one side and conversations on the other. These are the conversations — pressure-tested by the operators, analysts, and founders who lived through each tier.
Michael Sutton
AEO Solutions Expert, Yotpo
AI as Data-Informed Analyst
"
We're seeing the most innovative brands use AI less as a content shortcut and more as a trust and discovery engine. The brands winning aren't just generating more AI content, they're feeding AI systems authentic customer signals like reviews, Q&A, loyalty behavior, and real shopper sentiment. As discovery shifts from traditional search into AI-driven recommendations, brands are realizing that structured, trustworthy consumer data is what actually influences whether they get surfaced, cited, and ultimately chosen.
Nick Selman
VP of Growth, Shoplift
AI as Dashboard Chatbot
"
Our best customers are using AI in two places. First, to synthesize across data that used to live in separate dashboards (site heatmaps, reviews, support tickets, returns, post-purchase surveys) and surface sharper strategic insight in a fraction of the time. Second, to ship the resulting changes faster, whether through Shopify Sidekick or Claude paired with their own code repo. What hasn't changed is how you prove a change actually worked. That still requires real traffic splits, segmentation, and Bayesian modeling on actual shopper behavior.
"
Every platform we used shipped an AI feature in the same six-month window. Five 'smart' tools, zero shared truth. The first cross-channel view we built was the most useful day of our year.
Bobby Palmieri
Co-Founder, Lilo Social
AI as Co-Pilot
"
We're seeing brands be able to ingest data points, insights, and monitor things that prior to AI would take too much manual labor for it to be worth it, and now can happen independently and watch AI cycle these insights into Slack natively for the team to act proactively on, rather than reactively when larger KPIs start to move in other directions. This can be things like ad fatigue, placement / demographic breakdowns, or large inputs of site analytics, customer reviews and comments to align the next creative hypothesis, landing page, or CRO test.
"
I was the brand. Sheets in one tab, ChatGPT in another, copy-pasting back and forth. The day I put my last six months of data into a project and stopped re-explaining who we were, I got my Sundays back.
Andriy Boychuk
CEO, Flowium
AI as In-House Infrastructure
"
The most effective email marketing programs we're seeing right now aren't just using AI to write copy faster — they're using it to predict customer behavior before it happens. Our clients are leveraging AI-powered order predictions and autonomous outreach to catch at-risk customers before they churn, personalize at a segment-of-one level, and let their teams focus on strategy instead of execution. The brands winning in email aren't automating more — they're automating smarter.
Patrick Coddou
COO, AJF Growth
AI as Operating System
"
We've built a full-stack software development workflow for our agency that runs autonomously using Claude Managed Agents. Our team @mentions our agent in Slack the way other teams ping an employee — it pulls roadmap context, drafts responses, files tickets, and builds new features almost entirely without supervision. That only works because we gave it real access to our data and our tools. Once an agent can read and write across your actual systems, it stops being a feature and starts being a hire.