There's a gap forming in ecommerce, between brands that are using AI to drive real operational results, and brands that are still running pilots, evaluating tools, or waiting for the right moment.
That gap is getting wider. And the reason isn't budget or ambition. It's data.
New research from MIT Technology Review Insights surveyed 500 senior leaders at large US companies, CTOs, CIOs, heads of digital and AI. The findings are a useful reality check for any commerce brand thinking about where AI actually fits in their business.
Three in four large organizations now have at least one AI workflow fully in production.* That's AI built into real operations, making real decisions.
If you're in ecommerce and that number feels distant, it's worth asking why because commerce businesses sit on some of the richest operational data of any sector. Product catalogs, order flows, customer behavior, inventory signals. The raw material is there. The question is whether the foundation is ready to use it.
The clearest finding in the research: AI succeeds most often when it's applied to processes that are already well-defined and running reliably.*
For ecommerce, it's an advantage. Order management, product data enrichment, customer segmentation, returns classification. These are structured, repeatable workflows your team runs hundreds of times a week. They're exactly the kind of processes where adding an AI layer creates real, measurable impact without needing to rebuild from scratch.
The brands getting traction with AI aren't changing their operations. They're making them smarter.
Here's where most ecommerce brands run into trouble.
AI needs a complete, consistent picture of your business to make good decisions. But in most commerce stacks, data is scattered, a PIM that doesn't sync with the OMS, a CRM that's always a day behind, a product catalog that looks different depending on which system you're looking at.
These aren't new problems. Most teams have been working around them for years. But AI makes the cost of those workarounds much higher. A model making decisions based on stale or inconsistent data doesn't just produce a wrong answer, it produces a confident wrong answer, at scale.
The research is clear on this: the companies seeing the strongest AI results are the ones that treat their data as a unified asset.*
Two-thirds of the organizations in the study have no dedicated AI team.* Progress is happening through central IT, departmental ops leads, and existing digital teams.
The "we're not ready" feeling is usually less about people and more about infrastructure. Getting clear on which processes to start with, and whether your data is clean and connected enough to support them, that's the real work. And it's the kind of work that pays off beyond AI, too.
Most AI conversations in ecommerce start with "what could we do with AI?".
A better starting point is: "which of our existing workflows are well-defined enough, and connected enough, to support an AI layer right now?"
That question leads somewhere concrete and it leads there faster.
The MIT Technology Review Insights report, Bridging the operational AI gap, goes deeper on how 500 senior leaders are building the foundation for operational AI: the data infrastructure, the integration choices, and the use cases actually delivering results.
If you're thinking about where to move next on AI in ecommerce, it's the most grounded data set we've seen on the topic.
*MIT Technology Review Insights, "Bridging the operational AI gap," 2026
Datmos helps commerce brands turn data into measurable growth, combining strategy, experience design, and operational activation. If your AI ambitions are running ahead of your data infrastructure, let's talk.
