Something interesting is happening in e-commerce right now. AI agents (personal shopping assistants built into your phone, your email client, your browser) are starting to make purchases on your behalf. Not browsing suggestions. Actual purchases. They're reordering your protein powder when you're running low, comparing air filter prices across five retailers, and completing checkouts without you lifting a finger.
The natural reaction from many in the merchant community is anxiety: If AI is doing the buying, what's the point of marketing?
The answer, once you think it through, is more reassuring than you'd expect. The death of brand and marketing is not what agentic commerce delivers. What it delivers is a clearer line between two types of purchase that have always existed but were never this distinct.
The Two Purchase Modes That AI Is Pulling Apart
Think about the last ten things you bought online. Some of them you thought about. You compared options, you read reviews, you felt something when you chose one brand over another. Maybe it was a jacket you'd been eyeing for weeks. A birthday present that had to feel right. A skincare product your friend swore by.
And some of them you just... needed. Paper towels. A phone charger. The same coffee you always order. Replacement batteries. You didn't browse. You didn't feel anything about the brand. You just bought the thing.
These two modes of purchase have always coexisted. But historically, they happened through the same channel: search, a browser, your phone. The experience was similar enough that merchants treated them similarly.
Agentic commerce is splitting them apart.
Routine, predictable purchases are becoming agent territory. If you have a recurring need, a consistent brand preference, and a clear price threshold, an AI agent can handle that purchase better than you can. It's faster, it never forgets, and it will comparison-shop without you having to ask. These are the purchases where friction was always the enemy, and AI eliminates friction entirely.
Emotional, considered purchases remain human territory. The gift for someone you love. The outfit for an occasion. The product you discovered through a brand story that resonated. The limited-edition thing that made you feel like you were in on something. These are purchases driven by meaning, identity, aspiration, and connection. No AI agent is picking your engagement ring or the dress you wear to your best friend's wedding. Those decisions are too personal, too loaded with context that only you have.
The line between these modes is becoming sharper every month. And that changes what marketing actually needs to do.
What AI Shopping Agents Actually Evaluate
Here's the part merchants often miss: AI agents are a new type of customer, one with completely different decision-making signals than a human browsing your store.
A human browsing your Shopify store responds to photography, copy, social proof, brand story, emotional resonance. They're affected by your homepage hero, your founder's story, the feeling your product page evokes.
An AI agent doesn't read your brand story. It reads structured data.
When an agent is deciding which paper towel brand to order for a user, it's weighing:
- Price relative to comparable SKUs
- Review score and review volume
- Whether the product has been ordered before (and if so, whether it was returned)
- Shipping speed and reliability
- Whether a promotion is active and machine-readable
It's not watching your banner video. It's not moved by your font choices. It's parsing your product feed, evaluating your promotional data, and checking your rating. The "shelf" in an AI-driven purchase is structured metadata, and that's what you need to optimize for.
This means there are now two distinct audiences you're marketing to simultaneously:
- Humans who are making emotional, considered purchases, and who respond to brand, story, aesthetics, and trust signals
- AI agents who are making routine purchases on behalf of humans, and who respond to structured data, pricing, availability, and reviews
The merchants who win in the next few years will be the ones who optimize for both.
Why Branding Matters More, Not Less
If AI agents are handling the low-involvement purchases, then the purchases that remain human-driven are disproportionately the high-involvement ones. The ones where someone is deeply considering what to buy and who to buy from. The ones where brand, story, and emotional resonance drive the decision.
That means the human purchase decisions that remain are more valuable, not less. And competing for them requires better marketing, not less of it.
Think about what this looks like practically. Your customer isn't going to browse your store to reorder shampoo anymore; their AI agent is going to handle that automatically, assuming your product is in the right feeds. But when they're buying a wedding gift, choosing a skincare brand for the first time, or picking something for themselves as a treat, they're going to take their time. They're going to look at your brand. They're going to feel something, or they're not going to buy.
The floor for emotional purchases is getting higher. Your brand now has to earn attention in a world where agents are handling everything else. Merchants who've relied on repeat purchases without investing in the kind of brand that earns first purchases are going to feel the squeeze.
This is why the merchant playbook has to evolve in two directions at once.
Two Audiences, Two Playbooks
Most merchants are optimizing for one audience or the other. The ones who will win are optimizing for both. Here's what that looks like in practice.
Layer 1
Optimize for AI Agents
Routine purchases — respond to structured data
Sync promotions to Google Merchant Center
AI surfaces read your promotions feed, not your Shopify storefront.
Keep your product feed accurate in real time
Price mismatches and out-of-stock items get you deprioritized.
Clean up your product data
Titles, categories, GTINs — the fields agents use to match purchase intent.
Accumulate and protect your review score
Agents weigh score and volume heavily when comparing options.
Layer 2
Optimize for Human Buyers
Emotional purchases — respond to brand and story
Build a brand that earns first purchases
Photography, storytelling, content. Agents handle repeat orders once you win the first.
Create trust signals humans actually read
Genuine reviews, clear returns policy, real social proof.
Be present where discovery happens
Social, influencer, editorial, search. Discovery is the most valuable moment in the funnel.
Brand equity trains agents too
Agents learn which brands people trust — strong brand shapes agent defaults downstream.
The two layers reinforce each other: brand equity built for human buyers shapes which brands AI agents recommend for routine purchases.
Layer 1: Optimize for AI agents
AI agents don't browse. They parse. Every action below makes you more visible and more trustworthy to the systems doing the buying.
1. Sync your promotions to Google Merchant Center. Google's Universal Cart, Direct Offers, and the Universal Commerce Protocol don't read your Shopify storefront; they read your promotions feed in Google Merchant Center. If your discounts only exist in Shopify's admin, agents can't surface them. Your active sale is invisible to the very systems routing purchase intent.
2. Keep your product feed accurate in real time. An agent that recommends your product and finds it out of stock, or sees a price mismatch at checkout, will deprioritize you. Pricing, availability, and variant data need to be consistently synced, not updated manually.
3. Clean up your product data. Titles, descriptions, category assignments, GTIN codes. These are the fields agents use to match your SKUs to purchase intent. Incomplete or inconsistent data means your products don't show up in the match. See how to prepare your store for AI channels for a full checklist.
4. Accumulate and protect your review score. Agents weigh review score and review volume heavily when comparing options in the same category. A 4.2 with 400 reviews beats a 4.8 with 12. Make it easy for customers to leave reviews after purchase.
Layer 2: Optimize for human buyers
Human buyers making considered purchases are doing more research than ever. They've handed off the routine stuff to agents, which means when they do browse, they're genuinely deciding. You need to give them something to decide in your favor.
1. Build a brand that earns first purchases. Photography, storytelling, community, content. The merchants who've relied on repeat orders without investing in acquisition-stage brand are going to feel the squeeze as agents absorb the repeat layer. Your brand needs to earn the first purchase; agents will handle the rest.
2. Create trust signals humans actually read. Genuine reviews, a clear returns policy, real social proof. The person making a considered purchase is researching more carefully than ever. Every signal that reduces uncertainty helps.
3. Be present where discovery happens. Social media, influencer content, editorial coverage, search. AI hasn't replaced human curation and recommendation; if anything, as routine purchases get automated, discovery becomes the most valuable moment in the funnel.
4. Remember: brand equity trains agents too. Agents are trained on human preferences and behavior. They learn which brands people trust and return to. High brand equity doesn't just win emotional purchases today; it shapes which brands agents default to for routine purchases tomorrow.
The Merchants Who Will Struggle
The merchants caught in the middle are the ones selling products that sit in the routine category but haven't invested in the infrastructure AI agents need to find them, and who also haven't built enough brand equity to compete in the emotional purchase layer.
If you're selling commodity products with weak promotional feed coverage, an AI agent will pick your competitor. If you're selling products that should be emotionally resonant but you've never invested in your brand story, human buyers will scroll past you.
The middle ground is shrinking. You either need to be optimized for the agent layer, or you need to be extraordinary in the human layer. Ideally both, because a customer who discovers you through an emotional, considered purchase becomes a customer whose AI agent handles their repeat orders.
The Upside for Merchants Who Do Both
The toothpaste is already on autopilot. So is the phone charger, the paper towels, the repeat order your customer places every six weeks. Agents handle those now, and that's fine.
What they can't handle is the anniversary gift. The outfit for the occasion. The skincare product that felt like it was made for her. Those decisions are still made by people who feel things. The brand that earns those moments earns the agent's loyalty for everything else, too.
For merchants who optimize both layers, agentic commerce is a growth multiplier. Get your promotional data into the right feeds, and agents drive new-customer acquisition with almost no ongoing effort. Build real brand equity, and you're not just winning the considered purchases. You're training agents to default to your brand for the routine ones too.
Optimize for the agent. Build for the human. The merchants who do both will own both ends of the funnel.
PromoOS helps Shopify merchants get their promotions visible across both layers: syncing discounts to Google Merchant Center's Promotions feed so AI agents can surface them, and giving you a single place to manage the campaign calendar across your store and your AI-channel feeds. Install PromoOS on the Shopify App Store.