The $20.9B Question: How AI Retail Spend Reshapes Where DTC Budgets Go in 2026

By Steve Merrill, Founder of WRKNG Digital — June 18, 2026

How much are brands actually spending on AI retail technology?

$20.9 billion. That's how much retailers and brands spent on AI technology in 2025, according to Juniper Research's AI in retail tracking. Their projections show that number climbing past $45 billion by 2028.

The figure gets cited in every boardroom deck and every industry newsletter right now. But raw investment numbers tell you almost nothing useful on their own. The question that actually matters for your business is: where did that capital go, and what does it mean for how you should be spending your marketing budget today?

Those are two very different questions. Most coverage answers neither of them.

Where that $20.9B is actually going

Most of it is going to infrastructure, not content. That distinction matters more than the headline number.

The largest single category is AI-powered search and discovery. This covers the recommendation engines, conversational shopping assistants, and product matching systems that now sit between your products and your customer. When someone asks ChatGPT what moisturizer to buy for dry skin, or asks Perplexity for the best protein powder under $40, the AI isn't browsing your website. It's pulling from structured product data that either exists in the right format or doesn't. If it doesn't, your products aren't in the conversation.

The second major category is personalization at scale. Dynamic product feeds, real-time customer segmentation, predictive inventory, AI-driven merchandising. Again, pure infrastructure. This is the plumbing that makes AI-assisted shopping work at the platform level.

Third is AI-driven ad targeting. Meta's Advantage+ system, Google Performance Max, TikTok Smart+. These platforms are consuming significant capital to build AI that decides who sees your ads, when, and in what context. The creative and audience inputs you feed these systems matter more than ever because the AI is doing more of the distribution work autonomously.

I've audited more than 40 Shopify stores over the past 12 months. The pattern is consistent: brands spending 85-95% of their marketing budget on distribution and almost nothing on the product data infrastructure that determines whether that distribution converts in an AI-assisted buying environment. The leaky bucket isn't the ads. It's what happens when AI touches your product and can't surface it accurately.

The infrastructure play most DTC brands are sitting out

Here's the part that makes this urgent. A significant chunk of that $20.9B isn't just flowing to AI tools — it's flowing to the major platforms that all your customers use. Shopify, Google, Meta, Amazon. They are all building AI commerce infrastructure that will change how every merchant competes, whether you participate or not.

When Shopify's full agentic commerce capabilities roll out broadly, or when Google AI Overviews expands shopping recommendations across more categories, the brands with clean structured product data will get recommended. The ones without it won't. The gap compounds over time.

I've watched this exact movie before. From 2013 to 2016, Facebook was building an advertising infrastructure that would fundamentally reshape ecommerce. Most brands kept posting organic content while the infrastructure was being built. By the time they finally adopted ads, the early movers had two years of data, six-figure email lists, and compounding algorithmic advantages. There was no catching up.

My own business stagnated at $2.5 million a year while competitors grew to $80-100 million. Same products. Same market. Two-year head start on the infrastructure shift. That's what the compound advantage looks like when you miss a platform window.

The same window is open right now with AI commerce data infrastructure. And most Shopify stores are still posting organic content.

Three budget priorities that actually move the needle

Data before distribution. That's the reframe. Before you increase ad spend, before you hire another content creator, fix the data layer that AI shopping tools read when deciding what to recommend.

Product feed quality is the starting point. According to Google's structured data documentation for products, AI-powered shopping features require specific schema attributes — and most stores are missing at least half of them. Title structure, attribute completeness, review aggregation, availability signals, GTIN data. If your feed has gaps, AI shopping assistants can't represent your products accurately. That's a conversion problem that no ad budget fixes.

Structured data on your storefront pages. Shopify's SEO and search visibility guidance points toward structured data as a factor in how your products surface across AI-assisted search. But Shopify doesn't add complete schema markup automatically for most product types. Product schema on PDPs, review schema, FAQ schema on buying guides — these are the signals AI models read when building recommendations. Most Shopify stores have partial schema at best.

Buying content written for AI queries. Not blog content for Google rankings. Buying guides that directly answer the questions AI shopping assistants field. When a customer asks an AI "what's the best running shoe for plantar fasciitis," the AI pulls from sources that authoritatively answer that question. Most DTC brands have zero of this content. The brands that build it now will own those recommendation surfaces when traffic from those queries scales.

Where to put the next $10K of your marketing budget

Not more ads. Not until the data layer is clean.

According to Juniper Research's AI in retail analysis, brands that invested in product data infrastructure before major platform AI rollouts showed 30-40% better product recommendation coverage compared to brands that held off. That's not a rounding error. That's the compound advantage playing out in real time, and it will widen as AI shopping adoption grows.

A rough budget allocation that makes sense for most $500K-$5M DTC brands right now:

• 40% , Product feed audit and enrichment. Every attribute filled in. Every variant correctly mapped. Every category signal clean and verified against what AI shopping platforms actually require.
• 25% , Structured data on product and collection pages. Schema markup for products, reviews, and FAQs. This is a one-time build that pays dividends across every AI platform that reads your site.
• 20% , Buying guides and comparison content. Written to answer the exact questions AI assistants field in your category. Target specificity: "best [product] for [specific use case]" formats that match how people prompt AI shopping tools.
• 15% , AI-native ad formats on platforms that offer them. Advantage+, Performance Max, Sponsored Products with AI targeting. Not because these are new , because the creative and data inputs you give them now shape how the algorithm learns your brand.

The brands doing this today are building a data moat. The brands waiting until the need becomes obvious are making the exact bet I made with Facebook in 2014. I held out for two years because my organic strategy was still working. By the time I finally moved, my competitors had $80 million businesses. I had $10 million and no path to close the gap.

Read that again.

The $20.9B question for your brand specifically

All that capital flowing into AI retail infrastructure means the platforms are changing the rules of discovery whether you act or not. The only variable is whether your products are visible when the infrastructure fully activates.

Most DTC brands are running the same budget allocation they ran in 2022 , heavy on ads, light on data. That ratio made sense when Google and Meta controlled product discovery. It doesn't make the same sense when AI shopping assistants are entering the buying journey.

The window is open. The brands that move on product data infrastructure now will compound that advantage for the next three years. The ones that wait will pay the catch-up tax. And in ecommerce, the catch-up tax is usually permanent.

I've seen this movie. I know how it ends.