Shopify Says It's Never Been More Excited. Forrester Says Slow Down. They're Both Right.
Shopify's president told an audience in Los Angeles this week that his company has never been more excited about anything in its history. Five days earlier, Forrester published adoption data suggesting most consumers haven't even started shopping with AI. The interesting part isn't that they disagree. It's that they're both looking at accurate data and drawing different conclusions.
What Did Shopify Say About Agentic Commerce?
Harley Finkelstein, Shopify's president, spoke at the Upfront Summit on March 16 and made two arguments worth unpacking.
First, the market expansion argument: only 18% of US retail purchases happen online. That means 82% of commerce is still offline. If AI agents make product discovery easier through natural conversation instead of keyword searches, they could pull offline buyers into digital channels for the first time. That's not capturing share of existing ecommerce. That's growing the entire addressable market.
Second, the merit argument: "Agentic is fundamentally merit-based," Finkelstein said, contrasting AI recommendations with search engines where paid placement dominates. An AI agent that knows your preferences surfaces the best product match regardless of marketing budget. For the millions of smaller merchants on Shopify, this levels a playing field that paid search and social ads have tilted toward well-funded incumbents.
Finkelstein wasn't hedging. "We're probably more excited about this particular new era of commerce than we have ever been," he said, calling agents a potential "new front door" for ecommerce sellers.
What Does Forrester's Data Actually Show?
Forrester's reality check, published March 12, came right after OpenAI pulled native checkout from ChatGPT. Their December 2025 Consumer Pulse Survey data:
- 35% of Gen Z US online adults used ChatGPT for product search in the past month
- 32% of Millennials did the same
- 23% of Gen X
- Single digits for Boomers and older
- Actually completing a purchase inside an answer engine is the least-adopted use case across every demographic
Forrester also pointed out that only about 30 Shopify merchants were live with native ChatGPT checkout before it was pulled. And the industry is "appropriating the term with endlessly different definitions" - Forrester's own agentic commerce glossary is publishing this month to try to bring some definitional clarity.
This is sobering data. But it's also a snapshot of March 2026 consumer behavior, not a prediction about December 2026.
Are Shopify and Forrester Actually Contradicting Each Other?
They're describing different timeframes and they're both right about their respective frames.
Forrester is accurate about today. Consumer adoption of AI shopping is early, concentrated in under-40 demographics, and the transaction completion piece barely exists. If you're making budget decisions based on current consumer behavior, Forrester's data says "don't over-invest yet."
Shopify is reading the infrastructure signals. In this week alone, Visa launched an agent payments program and backed Stripe's Machine Payments Protocol. Google's UCP is live. MCP moved to the Linux Foundation. Shopify's Agentic Storefronts are active. When Shopify looks at the rate of infrastructure deployment, "most excited ever" makes sense.
Every major platform shift follows the same pattern. Mobile payment infrastructure existed years before mobile commerce was mainstream. Cloud commerce platforms were live long before the majority of retailers migrated. Infrastructure leads. Adoption follows. The gap between them is where preparation happens.
What Should Retailers Actually Do Right Now?
The practical answer depends on where you are today and what you can do without betting the budget on uncertain timing.
Start with visibility. Search for your own products on ChatGPT, Google AI Mode, and Perplexity this week. If you don't appear at all, you have a data problem, not a strategy problem. Check how many of your pages Bing has indexed (ChatGPT runs on Bing). Check whether your product pages have JSON-LD markup beyond the basics.
Then focus on data depth. Shopify's own research shows AI-driven orders are up 15x since the beginning of 2025, and Gartner predicts 20% of transactions will execute through AI platforms by 2030. The retailers capturing that early wave have one thing in common: rich, structured product data. Not 5-8 attributes per product. 30+. Material, dimensions, compatibility, care instructions, certifications, use-case tags.
Add an llms.txt file to your domain root. It's a single text file that helps AI crawlers understand what your site offers. Takes an hour, costs nothing.
What I'd hold off on: custom integrations with individual AI platforms, native checkout infrastructure for agents, and agent-specific marketing campaigns. The protocols are still consolidating and the consumer adoption numbers don't justify that level of investment yet.
Why the 18% Number Matters More Than the 35% Number
Forrester's 35% Gen Z adoption rate gets the headlines because it benchmarks current behavior. But Finkelstein's 18% online retail penetration is the more important number for long-term planning.
If AI shopping agents make online purchasing more natural - through conversation instead of search, through personalized recommendations instead of browsing, through simplified one-step purchasing instead of multi-page checkout - the growth opportunity isn't just in the 18% that's already online. It's in the 82% that isn't.
That's the gap Shopify is investing against. Not today's adoption curve. Tomorrow's market expansion.
Forrester is right to ground expectations in current data. Shopify is right to build aggressively for what's coming. Retailers should be doing both: get the infrastructure ready now - structured data, AI-optimized feeds, Bing indexing - while keeping budget commitments calibrated to actual consumer behavior. When adoption does inflect, the merchants who prepared during the infrastructure phase will be the ones who capture it.
