Talking about agentic commerce optimization is one thing. Seeing what it actually looks like in practice is another.
In this post we'll walk through a real before-and-after of a Shopify product listing, explain exactly what changed and why, and show the impact on how AI shopping agents evaluate the product.
This is a composite example built from patterns we see across hundreds of Shopify catalogs. The product is fictional but the issues are real — and the fixes are the same ones merchants are making right now to get their products recommended by ChatGPT, Perplexity, and other AI shopping agents.
The Product: A Ceramic Pour-Over Coffee Dripper
Let's look at a typical specialty coffee equipment store selling a ceramic pour-over brewer. The merchant makes quality products, has beautiful photography, and ranks decently on Google for "pour over coffee dripper." But when customers ask ChatGPT for coffee brewing recommendations, their product doesn't get suggested. Here's why.
Before: The Original Listing
Title: "The Morning Ritual"
Description: "Handcrafted from premium ceramic, The Morning Ritual transforms your daily coffee routine into a moment of mindful pleasure. Each piece is made by skilled artisans who pour their passion into every detail. Experience coffee the way it was meant to be enjoyed."
Tags: coffee, ceramic, handmade, gift
Metafields: None populated
Image alt text: "product image 1", "product image 2", "lifestyle shot"
This listing reads beautifully to a human. The prose is evocative, the brand voice is consistent, the emotional appeal is clear. A shopper browsing the website gets the story of the product and feels the craftsmanship.
An AI shopping agent reading this data extracts almost nothing. It knows there's something called "The Morning Ritual" that involves ceramic and coffee. It doesn't know what the product is, how it works, what it costs to ship, what size it brews, what grind it requires, or who it's for. When a customer asks ChatGPT for "a pour-over coffee dripper for making 2 cups at a time, something high quality under $60," the agent has no way to match this product to that query. Even though this product is exactly what the customer wants.
After: The AI-Optimized Listing
Title: "The Morning Ritual — Ceramic Pour-Over Coffee Dripper, 2-Cup Capacity"
Description:
"The Morning Ritual is a handcrafted ceramic pour-over coffee dripper designed for brewing 1-2 cups of specialty coffee at a time. Made from high-fired porcelain ceramic, it retains heat exceptionally well — a critical factor for even extraction when brewing pour-over coffee at home.
Specifications:
- Material: High-fired porcelain ceramic
- Capacity: 250ml (brews 1-2 cups)
- Compatible filters: #02 conical paper filters
- Compatible grind: Medium-fine
- Height: 9cm
- Diameter: 11cm at the top, 6cm at the base
- Weight: 340g
- Dishwasher safe: Yes
- Made in: Portugal
Best for:
- Specialty coffee drinkers who want full control over their brew
- Home baristas learning pour-over technique
- Anyone who prefers single-origin coffees that benefit from the clarity of paper-filtered brewing
- Gift recipients who appreciate handcrafted coffee equipment
How to use:
Place a #02 filter in the cone, rinse with hot water, add 15-20g of medium-fine ground coffee, bloom with 30ml of water at 200°F for 30 seconds, then slowly pour 240ml of water in circular motions over the next 2-3 minutes. Designed for the 1:15 coffee-to-water ratio preferred by specialty coffee enthusiasts.
Each piece is made by skilled artisans in northern Portugal and may vary slightly in glaze and finish."
Tags: pour-over, coffee-dripper, ceramic, porcelain, handmade, 2-cup, specialty-coffee, home-brewing, pour-over-v60-style, filter-coffee, single-origin-brewing, portugal-made, gift-for-coffee-lover
Metafields:
- Material: High-fired porcelain ceramic
- Capacity: 250ml
- Country of origin: Portugal
- Care instructions: Dishwasher safe, hand wash recommended for longevity
- Product type: Pour-over coffee dripper
- Target audience: Home baristas, specialty coffee enthusiasts
- Compatible accessories: #02 conical paper filters
Image alt text:
- "Ceramic pour-over coffee dripper in cream white, shown from front with filter in place"
- "Top-down view of the ceramic dripper showing conical shape and central drip hole"
- "Dripper in use on glass server, water being poured in circular motion during brewing"
- "Hand holding the dripper to show scale, approximately 11cm diameter"
What Actually Changed
The product, price, and photography didn't change. The emotional appeal of the brand is still there — the "handcrafted by artisans in Portugal" story survived. What changed is everything around the story: the facts that answer an AI agent's questions.
The title now tells the agent what the product is. "The Morning Ritual" alone was a brand name with no context. "The Morning Ritual — Ceramic Pour-Over Coffee Dripper, 2-Cup Capacity" lets an AI agent immediately categorize this as a pour-over brewer of a specific size. When a customer asks for "a 2-cup pour-over dripper," this product can now be matched against that query.
The description now contains verifiable specifications. Material, capacity, filter compatibility, grind compatibility, dimensions, weight, country of origin. An AI agent evaluating "which pour-over dripper brews 1-2 cups at a time?" can now confidently recommend this product because the answer is explicitly in the data. Previously the agent would have had to guess — and would skip the product rather than risk a wrong recommendation.
The "Best for" section frames use cases and audiences. This is one of the most important additions. AI agents match products to customer intent, and customer intent is usually expressed as a need or situation. "Home baristas learning pour-over technique" matches a query like "what's a good beginner pour-over dripper?" The original listing gave the agent nothing to match against this kind of query.
The "How to use" section provides context that humans might skip. Experienced coffee drinkers know how to use a pour-over dripper. But an AI agent doesn't assume that knowledge exists in the customer's query. When a customer asks "how do I brew pour-over coffee at home?" an agent can now point to this product and use its instructions as part of the answer — which increases the likelihood of the recommendation.
The tags are specific and functional. The original tags (coffee, ceramic, handmade, gift) describe every coffee product ever made. The new tags include specific attributes (2-cup, filter-coffee, single-origin-brewing) that help an agent filter to exactly the right matches. "Gift-for-coffee-lover" is also more targetable than just "gift."
Metafields provide structured data. This is the invisible layer that merchants often skip entirely. AI agents lean heavily on structured metadata because it's unambiguous. The material is ceramic. The capacity is 250ml. The origin is Portugal. There's no parsing or inference required. The agent just reads the facts.
Image alt text gives visual context. AI agents can't see images, but they can read descriptions of what the images show. Describing the shape, the way the product is used, and the scale gives the agent visual information it can incorporate into recommendations — "the dripper is shown in cream white ceramic" answers a customer question about colour without the agent needing to guess.
The Result: Same Product, Different Discoverability
The product hasn't changed. The customer experience on the website is identical. Someone browsing the store still sees the beautiful photography, reads the brand story, and feels the craftsmanship. The human shopper's experience is preserved.
But the AI agent's experience is completely different. A listing that was invisible to ChatGPT is now a strong candidate for every relevant query about pour-over coffee brewing. When a customer asks for "a ceramic pour-over dripper for making 1-2 cups at home, something well-made under $60," this product can now be confidently recommended.
This Is the Gap Most Shopify Merchants Haven't Closed
The majority of Shopify stores look like the "before" version. Beautiful listings optimized for human browsing, with the facts scattered across images, descriptions, and product pages in a way that only humans can piece together. AI agents can't do that piecing. They need the facts laid out explicitly.
The merchants who close this gap now — before agentic commerce becomes the dominant discovery channel — are the ones who'll own recommendations for the next few years. Not because their products are better, but because their data is more complete.
How to Do This Across Your Catalog
Manually rewriting every product in a large catalog is unrealistic. That's what AgenticLens was built to solve. It scores every product across five categories, identifies exactly what's missing, and generates AI-optimized rewrites that preserve your brand voice while adding the structured information AI agents need. You review the suggestions, edit them where you want, and publish with one click.
The first 10 products are free. Most merchants run their initial scans, see the pattern across their catalog, and understand the gap immediately.
Check your store's readiness score for free at agenticlens.io
