Ulta Just Enabled Direct Purchases Through Google Gemini: The Agentic Commerce Shift Bypassing Your DTC Site

April 23, 2026

Your customer is buying products without ever visiting your website.

Not through Amazon. Not through a social platform. Through Google's AI assistant, which just started processing transactions directly inside conversational search results.

As Digital Commerce 360 reported today, Ulta Beauty partnered with Google to enable direct product purchases through Gemini's agentic commerce functionality. A shopper can now ask Gemini "what's the best vitamin C serum for sensitive skin," receive personalized recommendations from Ulta's catalog, and complete the purchase—all without leaving the AI interface or visiting Ulta.com.

This isn't a pilot. It's live. And it represents the most significant shift in product discovery since Google Shopping launched.

For independent CPG and DTC brands, this is the moment the agentic commerce warnings become operational reality. The purchase is happening inside the AI layer. Your product page? Optional. Your carefully crafted brand story? Might never load. Your conversion rate optimization work? Irrelevant if consumers never reach your site.

The only brands that will appear in these AI-powered shopping experiences are the ones whose product data is structured, complete, and readable by AI agents right now.

The Pattern: AI Agents Are Becoming the New Storefront

Ulta's Gemini integration isn't an isolated experiment. It's part of a coordinated infrastructure buildout happening across the retail ecosystem simultaneously.

Today also brought news that Dick's Sporting Goods launched AI-powered "digital coaches" built with Adobe that provide sport-specific training advice and guide shoppers from discovery to purchase based on contextual needs—not keyword searches. A runner training for a marathon gets different shoe recommendations than someone recovering from plantar fasciitis, even if they both search "running shoes."

And Google announced a $750M commitment to help its 120,000 partners build agentic AI solutions, embedding engineers with consulting firms like Accenture, Deloitte, and PwC to accelerate enterprise AI agent development.

Connect these dots: Major retailers are deploying AI agents that make purchase decisions. Google is investing three-quarters of a billion dollars to scale this infrastructure across enterprise partners. And the transaction is moving into the AI layer, potentially bypassing traditional ecommerce sites entirely.

As we documented in our analysis of David's Bridal making ChatGPT a direct sales channel, this shift has been building for weeks. But today's Ulta launch marks the first major beauty retailer enabling native transactions inside Google's AI interface—the platform where product discovery actually happens for most consumers.

Here's what independent brands need to understand: Dick's AI coaches need rich product data to make contextual recommendations. Ulta's Gemini integration pulls from structured catalog feeds. These AI agents don't scrape your beautifully designed About page or parse your brand manifesto. They read schema markup, product attributes, and structured data feeds.

If your product information isn't structured for machine reading, you don't exist in this channel.

Why This Matters More for Independent Brands Than Big Retailers

Ulta and Dick's have dedicated teams optimizing product feeds and negotiating AI placement deals with Google. They have retail media budgets in the eight figures.

You don't.

Which means the only way independent brands compete in agentic commerce is by having fundamentally better product data than retailers—structured so precisely that AI agents prefer recommending your DTC offering over the retail alternative.

This is actually an opportunity.

Large retailers have tens of thousands of SKUs with inconsistent data quality, legacy systems, and product information managed by dozens of vendors. You have complete control over your catalog. You can implement schema markup this week. You can add comprehensive product attributes to your Google Merchant Center feed today. You can structure FAQs in a format AI agents can parse and use to answer customer questions.

The brands that move fast on data structuring will appear in AI recommendations alongside—or instead of—major retailers, because the AI doesn't care about brand size. It cares about data quality and relevance to the user's question.

But the window is closing. As AI traffic to ecommerce surged 393% in Q1 2026, the brands that structured their content early are already seeing preferential placement. The longer you wait, the more AI training data competitors accumulate, and the harder it becomes to break through.

The YouTube Commerce Layer You Can Activate This Week

While AI agents remake discovery, another native commerce channel just opened for independent brands.

Google enabled WooCommerce merchants to sell products directly through YouTube videos and Shorts via the Google for WooCommerce extension. Products sync automatically through Google Merchant Center and appear as shoppable cards in video content.

This matters because it transforms YouTube from an awareness channel into a conversion channel without requiring users to navigate to your website.

Think about your current video strategy. You probably create product demos, how-to content, unboxings, or lifestyle videos and hope viewers click through to your site. Now those same videos become direct sales channels. A viewer watching your smoothie bowl recipe video can click the tagged organic granola product and purchase without leaving YouTube.

For Shopify and BigCommerce brands, the equivalent functionality exists through Google Shopping integrations—the key is ensuring your Google Merchant Center feed is comprehensive and synced properly so products can be tagged in video content.

The convergence here is critical: AI agents are pulling from the same structured product data that powers YouTube shopping cards, Google Shopping ads, and retail media placements. Every attribute you add to your product feed—material, color, size, use case, ingredient list—simultaneously improves your visibility across multiple discovery channels.

This is why product data structure is the foundational infrastructure for modern ecommerce, not a technical afterthought. It's the content layer that feeds every AI-powered discovery experience.

What BloggedAi's Approach Solves for Product Brands

The through-line across today's developments—Ulta's Gemini commerce, Dick's AI coaches, YouTube shoppable videos—is that they all consume structured, schema-rich product content.

BloggedAi's core thesis is that AI-discoverable content isn't about gaming algorithms or keyword stuffing. It's about providing product information in the structured formats that AI agents need to recommend and sell your products accurately.

That means comprehensive Product schema markup. FAQ schema that answers the questions consumers actually ask AI assistants. Detailed attribute data in your Google Merchant Center feed. Content structured around use cases, not just features.

When Dick's AI coach recommends products for marathon training, it's pulling from structured data about cushioning, drop height, stability features, and user reviews. When Ulta's Gemini integration suggests serums for sensitive skin, it's reading ingredient lists, product descriptions, and customer ratings formatted for machine parsing.

Independent brands that structure content this way don't just improve AI discoverability—they create a competitive moat. Because once AI agents start preferring your product data quality, you appear in recommendations even when consumers don't search for your brand name specifically.

What to Do This Week: Five Tactical Actions for Independent Brands

1. Audit and Maximize Your Google Merchant Center Product Attributes

Log into Google Merchant Center and review your product feed. For every product, fill out every optional attribute that applies: material, color, size, age_group, gender, pattern, product_detail, product_highlight.

These attributes directly feed AI shopping assistants. The more complete your data, the better AI agents can match your products to specific user questions.

If you're on Shopify, WooCommerce, or BigCommerce, use your platform's Google Shopping integration to ensure product data syncs automatically and stays updated. Then manually enhance the feed with attributes your platform might not capture automatically.

2. Add Comprehensive Product Schema Markup to Every Product Page

Implement Product schema markup on your product pages if you haven't already. Include these properties at minimum:

• name (exact product name)
• description (detailed, benefit-focused, 150-300 words)
• brand (your brand name)
• sku (unique identifier)
• offers (price, currency, availability)
• aggregateRating (if you have reviews)
• image (high-quality product photos)

For Shopify users, apps like Schema Plus or SEO Manager can add this markup automatically. WooCommerce users can use Schema Pro or Rank Math. BigCommerce has built-in schema support—verify it's enabled and complete.

AI agents use this structured data to understand your products when generating recommendations. Without it, you're invisible.

3. Create AI-Optimized FAQs Using Conversational Questions

Add an FAQ section to every product page that answers questions the way real people ask AI assistants—not the way you think about your product.

Instead of "What are the product specifications?" use "Is this serum safe for sensitive skin during pregnancy?"

Instead of "What sizes are available?" use "Will this fit someone who's 5'4" and usually wears a medium?"

Use FAQ schema markup (FAQPage) so AI agents can extract these answers directly. When someone asks Gemini a question about your product category, you want your FAQ to be the source AI cites.

Shopify users: Add FAQs using an app that includes schema markup, or code it manually using JSON-LD. WooCommerce and BigCommerce users have similar plugin options.

4. Set Up YouTube Product Tagging for Existing Video Content

If you're on WooCommerce, install the Google for WooCommerce extension immediately and connect your Google Merchant Center account.

For Shopify and BigCommerce brands, ensure your Google Shopping integration is active and products are synced to Merchant Center.

Then go to YouTube Studio, select your existing videos and Shorts, and start tagging products. Prioritize videos that demonstrate products in use—how-tos, tutorials, unboxings, lifestyle content.

You're not creating new content. You're making existing content shoppable. This takes 15 minutes per video and creates a new conversion path that doesn't exist today.

5. Test Your Products in AI Shopping Assistants

Open ChatGPT, Claude, or Google Gemini and ask the questions your customers would ask. "What's the best organic protein powder for someone with dairy allergies?" "Show me non-toxic cookware under $200."

See if your products appear. If they don't, that's your baseline problem. If they do appear, evaluate the information AI provides—is it accurate? Complete? Compelling?

The description AI generates comes from your product data. If it's generic or missing key benefits, you need better structured content.

This isn't theoretical research. This is testing the channel where product discovery is actively happening right now.

The Retail Media Reality Independent Brands Can't Ignore

There's a secondary signal in today's news that independent brands need to track carefully.

Best Buy's CEO Corie Barry is stepping down, to be replaced by Jason Bonfig—the executive who currently runs Best Buy's ads and marketplace businesses.

This isn't just a succession story. It's a strategic signal: retail media is now important enough to produce the CEO.

For brands selling through retail partners, this means retail media budgets are no longer optional marketing experiments. They're tied directly to distribution, shelf placement, and merchandising priority.

If you sell through Best Buy, Target, Ulta, or other retailers with mature retail media networks, expect increasing pressure to participate in paid placement programs. The brands that don't invest in retail media will find themselves deprioritized in search results, excluded from promotional placements, and potentially losing shelf space to competitors who do invest.

But here's the critical insight for independent brands: retail media shouldn't replace owned-channel marketing. It should complement it.

You still need to build direct customer relationships through your own site. You still need email and SMS flows that convert one-time buyers into repeat customers. You still need content and SEO that drives organic discovery.

Retail media gets you visibility in partner stores. Owned-channel infrastructure gets you margin, customer data, and independence from platform fees.

The brands that will survive the next five years do both: invest in retail media where they have distribution, and simultaneously build AI-discoverable content on owned channels that captures customers researching products before they reach retail sites.

When Digital Saturation Drives Physical Diversification

One more pattern worth noting from today: DTC brand Marine Layer is fighting digital ad saturation by investing in print catalogs and experiential pop-ups instead of purely digital channels.

Why? Because customers are "clickers, not pickers"—they quickly scroll past digital ads without meaningful engagement.

As Modern Retail reported, Marine Layer sees better engagement from physical catalogs and in-person experiences than from Instagram ads or Google Shopping campaigns.

This seems counterintuitive in a story about AI-powered commerce, but it's actually the same underlying problem: consumers are overwhelmed by digital noise.

AI shopping assistants solve this by providing curated, conversational recommendations instead of endless search results. Print catalogs solve it by creating a focused, tactile browsing experience without algorithmic distraction.

Both approaches recognize that traditional digital advertising—the scroll-and-click model—is breaking down due to oversaturation and shortened attention spans.

For independent brands, the lesson isn't "abandon digital" or "go all-in on print." It's diversify discovery channels beyond the Facebook/Google duopoly that's becoming less effective every quarter.

AI-powered product discovery is one diversification path. Physical touchpoints are another. Video commerce is a third. The brands that survive won't be the ones optimizing a single channel to perfection—they'll be the ones with presence across multiple discovery modalities.

Frequently Asked Questions

What Happens When the Storefront Disappears

Here's the question independent brands should be asking: what happens when AI agents get good enough that consumers never visit product pages at all?

Right now, Ulta's Gemini integration still links to product pages for checkout. But that's a technical limitation, not a permanent design. Payment infrastructure for AI agents already exists—American Express built it weeks ago, as we covered in our analysis of agentic commerce payment systems.

The technology exists for a complete transaction inside an AI interface: discovery, comparison, selection, payment, fulfillment—without the consumer ever loading a web browser.

In that world, what's the role of your Shopify store?

I think it becomes content infrastructure. The place where you publish the structured product data, reviews, FAQs, and media that AI agents consume to represent your brand. Your site is the source of truth AI references, even if consumers don't visit it directly.

Which means your product pages need to be optimized for machines first, humans second. That's a fundamental inversion of current ecommerce best practices, which prioritize human conversion rate optimization above all else.

The brands that recognize this shift now—and start structuring content for AI consumption—will be the ones AI agents prefer to recommend. The brands that keep optimizing for human visitors who may never arrive will wonder why their traffic disappeared.

We're not predicting a future scenario. This is operational reality as of today. The question is whether your product data is ready for it.

Want to see how your product pages perform in AI search? Try BloggedAi free → https://bloggedai.com