Sean Frank dropped a spicy take earlier this week that “ecom is dead”. If we use history as a guide, I agree with the direction of his take, but not on the impact.

Ecom isn't dead… we’re just getting a new channel. And if you look at the pattern, we were due.

Here's the thing about commerce channels: they don't replace each other, they just add to the stack. Historically, new channels emerged every 10-15 years:

But notice what happened next: the cycle compressed to ~5 years.

Social shoppers grew from 47 million to an expected 104 million by year end - that's faster adoption than any previous channel (prior to even having it fully built in).

Very important pattern here.

Each "revolutionary" channel ends up being 10-20% of the total pie, while the others keep chugging along. Catalogs still sit at around 10% of consumer spend, malls didn't die (though they had a rough patch), and physical retail is still 81% of all consumer spending. Turns out people still enjoy touching things before buying them!

I think agentic commerce is going to be even faster than social shopping for one simple reason: Agentic commerce is building on behaviors that already exist at massive scale.

I spent 5 hours digging through research on AI shopping behavior. The data tells a story that should get every DTC operator's attention. Now you know what my bedtime reading looks like.

ChatGPT processes 2.5 billion messages daily. Even though only 2.4% fall into "purchasable products," that's still 60 million shopping-related conversations every single day globally.

For perspective: it took Amazon 10-15 years to reach 60 million daily shopping interactions. ChatGPT hit that in 3 years.

But here's what caught my eye: 35% of AI users already use it for shopping research. They're using AI for price comparison (54%), review summarization (41%), and deal finding (41%).

That's not some futuristic behavior we need to wait for. That's happening right now, at scale.

The acceleration from social commerce to agentic commerce happened because:

The data reveals exactly what you'd expect from any new channel: trust drops as stakes increase.

This mirrors every other channel adoption curve. Remember when people wouldn't put their credit card info online? Now e-commerce is a $1 trillion market.

The pattern is always the same: start with low-risk transactions, build trust, expand upmarket. Amazon started with books. TikTok Shop started with $3 phone cases.

We’ll start seeing AI agents help you sell more chapstick, but may not be moving those $4,000 Rolex’s anytime soon. (Though honestly, have you seen the price of chapstick lately?).

Google just released a purchasing API that lets AI agents complete actual transactions. That's not a research project—that's infrastructure for a channel they expect to be meaningful.

When Google builds commerce infrastructure, they're not hoping it works. They're positioning for traffic they know is coming. It’s like watching your neighbor buy a snow blower in July… they know something you don’t.

Consider: Google has handled shopping queries for 20+ years. They can see the shift to conversational commerce in their data before anyone else. If they're building agent-purchasing tools, it's because they see the writing on the wall.

Every time a new channel emerges, the same question comes up: "Will this kill Shopify?"

The answer is always no, for the same reason: Shopify doesn't care about channels. They care about transactions.

When Amazon took off, brands still needed Shopify for their direct sales. When social commerce exploded, brands still needed Shopify for checkout. When marketplaces grew, brands still needed Shopify for inventory management.

Agentic commerce will be the same. AI agents will need to buy from somewhere. Shopify's bet is that "somewhere" will still be Shopify-powered stores, whether the customer clicks "buy" or an AI agent does. And they’re already investing in that bet heavily. [1] [2]

Based on the adoption patterns and current data, here's what I predict:

Next 6 months: AI research behavior continues exploding. 60 million daily shopping conversations grows to 100+ million as more people discover AI can help them comparison shop and find deals. The reality is that “Seeking Interest” probably already overlaps some, pushing it into 100+ million

6-12 months: We’re going to see the first 2 - 5 startup brands scale crazy fast with low CAC because of winning in AI search & agentic commerce as it evolves.

12+ months: Agentic commerce becomes a real channel worth optimizing for, representing 5% of sales for early adopters.

The timeline accelerates because the behavior already exists. People already ask AI to research purchases. The leap to "just buy it for me" is smaller than the leap from "never shopping online" to "trusting websites with my credit card."

Don't panic about AI agents replacing your website.

Do start thinking about AI agents (and search) as another channel that needs optimization.

Great CMO’s are all about acquisition arbitrage. It might be 6 months from now, but I think we’re going to see it acquisition arbitrage happen here.

It’s why I’d put Agentic Commerce in-between “Overhyped” and “Actually Useful”. Yes, I am cheating, but it’s my rating system. It’s worth investing some in AI search now, but most important to keep an eye out.

Three things to do this month:

This isn't about building some fancy AI shopping agent. It's about recognizing that customers are already using AI to research purchases, and positioning your brand to be found during that research.

The channel is here. The infrastructure is being built. The behavior is already happening at larger scale than we realize.

Agentic commerce will be the next inevitable addition to your channel mix. The question is really just a question of whether it’s in the next 6 months, 12 months, or 18 months.

And that's it for this week's edition. Ecommerce isn't dying, it's just getting a new sibling. One that likes emojis, and a lot of emdashes… and is maybe a bit verbose at times.