Enterprise hits and misses - Walmart hits an e-commerce milestone, while CEOs weigh in on the global market reset

Lead story - Walmart turns an e-commerce profit - but how will generative AI shake up the retail market? 

Stuart's 'Will Walmart turn the corner in e-commerce?' posts go back as far as diginomica can go. But now we have a different headline to consider: Walmart US is finally e-commerce profitable! Here's how the country's biggest retailer pulled it off. Stuart writes: 

Over the years we’ve looked in-depth at the evolution of Walmart as an omni-channel retailer and one that has become a bellwether in the space. Of note across all that time, the firn has become a notable exemplar for the digital capabilities of the omni-channel balancing act, but crucially has singularly failed to turn a profit on its e-commerce operation.

Well, that's changed now: 

No more. While the e-commerce business was losing around an estimated $1 billion a year not that long ago, the firm had been hinting that it would turn profitable this year. Now accounting for 20% of total sales, and with a goal of bringing that up to a 50% online/offline parity, Walmart US is showing e-commerce profitability in the current quarter and expects this to be reflected in full year numbers.

But... how? Stuart quotes David Guggina, EVP, Supply Chain Operations at Walmart US:

That didn't happen overnight. We've been on a decade-long journey to reshape the economics of our e-commerce business. We've made great progress in realizing benefits from using our stores of local fulfillment nodes, improving inventory position, optimizing route density and seeing customers' willingness to pay for express... We’re delivering an omni-experience for our customers like no one else, making shopping easy, fast and convenient.

I'm not a frequent Walmart shopper, but: if my local experience is an indicator, I'm seeing some advantages of that combination of store and delivery that Amazon (mostly) doesn't have. I'm seeing a speed and responsiveness that includes, crucially, the back end inventory visibility Guggina talks up here. However, I'd say the so-called "omni-experience" is still a work in progress. 

Nonetheless it's clear to see how stores-doubling-as-local-fulfillment centers can give an e-commerce edge. It's also instructive, however, to see how long it's taken Walmart to get this far. Omni-channel dexterity is not an easy thing to achieve. Speaking of 'not so easy,' how is that generative AI thing going? Stuart has that update in Walmart CEO Doug McMillon - being an 'and company' in an age of generative AI in retail. Gist: Walmart now has some practical wins to point to. Stuart quotes McMillon: 

We've used generative AI to improve our product catalog. The quality of the data in our catalog affects nearly everything we do from helping customers find and buy what they're looking for, to how we store inventory in the network, to delivering orders. We've used multiple Large Language Models to accurately create or improve over 850 million pieces of data in a catalog. Without the use of generative AI, this work would have required nearly 100 times the current headcount to complete in the same amount of time.

Prior AI experience has been of use here. More from McMillon: 

We've been using machine learning and AI for many years for customer personalization to enable our associates and to improve inventory flow. Much of that work has been driven by traditional predictive AI. In more recent years, generative AI has been playing a bigger role, and we're in the early stages of putting agentic AI to work.

Coming soon to all: your Walmart gen AI shopping agent, Sparky. Stuart:

Sparky’s aim is to help shoppers become informed buyers - a laudable ambition! - using advanced AI models to interact conversationally, understand shopping needs, suggest relevant products, answer questions, and provide recommendations based on preferences.

A few observations/nits: 

• Retail is a good gen AI/agentic testing ground. The consequences for inaccurate output are most likely manageable, and Walmart is surely doing far more to improve that 'intelligence' via data etc., than say putting a ChatGPT front end onto your site as some overly-hasty retailers have tried.
• The catalog example is a good one: it illustrates one of the best agentic roles currently (quality assurance). I suspect some final review of some content was needed, but this is a good example of where headcount reduction is less likely than serving customers better - with an overhaul that wouldn't have been done without this tech
• I'd call this "generation" rather than "creation," and I'd feel the same if humans were generating the catalog content. We're tossing around the word "creative" with a carelessness that has consequences, but that's a much longer article (or podcast). If the catalog content wins a creative writing award, get back to me...

Diginomica picks - my top stories on diginomica this week

• Why Open Banking is still not such an open-and-shut case - George delves into the pros/cons of the open banking 'movement': "I thought Open Banking was a solved problem, at least in the UK. Fulmer paints a nuanced picture of the many challenges facing established banks, startups, regulators, and consumers that must be navigated for everyone to see the benefits." Also see: George's The UK’s first profitable online bank - here's the tech side of how Atom Bank pulled it off. 

Vendor analysis, diginomica style. Spring event madness is officially here.. Let's start with our coverage of Google Cloud Next. As Derek writes in Google Cloud Next 25 - Google Cloud makes its pitch to own multi-agent orchestration in the enterprise: 

The company is clearly betting that enterprises will opt for an open, interoperable approach to AI agents rather than locking themselves into a single model ecosystem. This bet makes particular sense for Google, which lacks the enterprise software footprint of competitors like Microsoft or Salesforce, but has strengths in AI research, infrastructure, and platform services.

Indeed. Without orchestrations of agents across vendors, enterprises are going to get frustrated with the limitations of single-vendor-agentic-workflows, and fast. Can Google play a key role here? Too early to say, but I agree with Derek's comments on their positioning. Whoever claims a guiding role in cross-vendor agentic orchestration will have a strong position, perhaps reducing other software vendors to a set of API calls. Time will tell. More Google Cloud Next coverage: 

• Google Cloud Next 25 - how Lowe's retail re-invention partnership with Google Cloud continues to fix-up future growth prospects - Stuart
• Google Cloud Next 25 - CEO Thomas Kurian addresses Trump tariff uncertainty and reassures EU customers of data sovereignty - Derek

Atlassian Team '25 diginomica coverage hub - with the future of work at a crossroads, this was an important show to track. Alyx and Sarah were on the ground, with other team members piling on the virtual coverage. As Alyx wrote in Atlassian Team '25 - inside Atlassian’s System of Work - connecting strategy, execution, and the human layer:

One of the hardest questions in enterprise software isn’t “what are we building?” – it’s “why does it matter?” That disconnect – between the work being done and the strategy it’s supposed to serve – isn’t just a visibility issue. It’s a structural flaw in how most organizations plan, build, and measure progress....

Atlassian’s evolving System of Work is responding to this disconnect. Rather than building another tool for managing tasks, Atlassian argues that it’s creating a connective tissue - linking goals to execution, surfacing context across tools, and turning documentation into a live asset instead of a forgotten artifact.

There's plenty of coverage to dig into... a few more of my picks: 

• Atlassian Team '25 - ‘We want AI regulation’ says Atlassian’s legal chief - Sarah
• Atlassian Team '25 - customers share first impressions of putting the Rovo AI agent to work - Sarah
• Atlassian Team '25 - Atlassian unveils new customer service and talent apps, removes AI price tag - Phil

A few more vendor picks, without the quotables:

• Adding a journey builder for marketers - how HubSpot is building depth into its AI agents - Barb
• Canva Create - Canva takes automation up a gear with the launch of a spreadsheet and AI app builder - Phil
• Sustaining through change - how the Elizabeth Glaser Pediatric AIDS Foundation manages their global mission with Unit4 - Jon

Jon's grab bag - Stuart looks at Europe's plans for AI dominance in Something for the weekend - welcome to the AI Continent of Europe. US and Chinese travelers, please have your documentation ready for inspection...  George examines an under-the-radar global trade topic of import in As Trump 2.0's tariff tirades deafen us, have we just forgotten about the new China rare earth controls?  Finally, Stuart looks at whether agentic AI can step in for field service worker shortages in Field service workers are burning themselves out on dreary admin. Can agent tech take over?

Best of the enterprise web

My top six

• As tech navigates volatility, here's what the big finance CEOs say about the economy - Constellation's Larry Dignan getting it done again: "On the corporate side, Barnam said clients have been reacting to tariff policy changes. 'At the margin that shifts their focus away from more strategic priorities with obvious implications for the investment banking pipeline outlook towards more short-term work, optimizing supply chains and trying to figure out how they're going to respond to the current environment.'"
• Why security stacks need to think like an attacker, and score every user in real time - Louis Columbus is on the security beat again, parsing some provocative/worthwhile new data.
• Digital Transformation Is Dying—And That Might Be a Good Thing - Eric Kimberling waves a fond farewell to a buzzword past its shelf life: "What does work? Empowering employees with better training, reimagining workflows, and fostering a culture of continuous improvement." Do we have a bingo? I don't mind Also see: Eric's The Data Dilemma: Why AI Strategy Begins with Data Strategy (I may have more to say on this one next week).
• From POC to Production: Why GenAI Projects Often Stall - Useful project lessons (and gotchas) via The New Stack. Yep, governance is a big factor.
• AI isn’t ready to replace human coders for debugging, researchers say - Ars Technica adds to the debate on how much impact gen AI can have on human coding.
• LLM-Modulo framework — The power of LLM for robust planning and reasoning - This is from last year, but I think we're going to find these compound LLM systems factor heavily going forward, so I recommend checking this piece and the white paper, and following along as this research unfolds. 

Whiffs

YouTube shorts recovery program anyone? 

Via our own Hazel Scott, a very good article on the encroachment of influencer fakery: 

This gem from 404 Media rolled in at presstime - it may warrant more attention next week: I Tested The AI That Calls Your Elderly Parents If You Can't Be Bothered. See you then...

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed.