✦ Sparky, and the Shelf Nobody Looks AtThin moat

Walmart (WMT) — the future bets

If an agent chooses the product there is no product page, no sponsored placement and no shelf.

Walmart has built a shopping assistant called Sparky and put it inside somebody else's chatbot. That single decision contains both the opportunity and the threat of the next decade of retailing.

What an agent removesToday: a product page, with sponsored placements at the topThe advertising business that sits on it: $6.4bn, +46%Agentic: one answer, no page, no sponsored slotWalmart's hedge: Sparky inside ChatGPTWalmart's own risk factors name AI-enabled platforms as a threat to store traffic.
Being inside the assistant preserves the transaction. What does not obviously survive is the advertising margin, which exists because a human looks at a screen and chooses.

The partnership with OpenAI was announced in October 2025 and the in-platform experience followed, letting customers discover and buy Walmart items inside ChatGPT with Sparky as the commerce agent1. Walmart's reasoning is that if shopping journeys start with an assistant, the retailer that is present inside the assistant wins and the one that waits to be visited loses.

The threat is the same fact stated from Walmart's side of the till. Walmart's own risk factors warn that "the increasing role of AI-enabled platforms in product search, discovery, advertising and purchasing" could reduce traffic to its stores and clubs2. If an agent selects the product, there is no product page, no sponsored placement and no shelf — which is a direct threat to a $6.4 billion advertising business that exists entirely because a human looks at a screen and chooses.

It is also a threat to the one thing Walmart has that Amazon does not: proximity. An agent does not care that the store is four miles away, except as an input to delivery time.

The hedge is reasonable. Being inside the assistant preserves the transaction even if the shelf disappears, and Walmart's fulfilment advantage survives a change in who does the choosing. What does not obviously survive is the advertising margin.

Watch advertising growth. It was 38% in the June 2026 quarter3. Agentic commerce is not yet large enough to register; the quarter in which retail media growth decelerates sharply without an obvious retail explanation is the quarter this bet turned into a problem.

Moat trajectory: Holding steady

Being inside the assistant preserves the transaction if shopping journeys start there, which is a sensible hedge. What it does not preserve is the advertising margin, and nothing in the current numbers shows which way that resolves.

The number that tests this moat
Reported
Global eCommerce sales growth, latest quarter
23%

Sparky puts Walmart inside ChatGPT in case shopping starts with an assistant. eCommerce growing well ahead of stores shows the digital business is keeping up; watch whether advertising grows with it.

Source: Walmart Q2 FY2027 earnings release ↗
References
  1. ReportedThe partnership with OpenAI was announced in October 2025 and the in-platform experience followed, letting customers discover and buy Walmart items inside ChatGPT with Sparky as the commerce agent.
    Retail Dive - Walmart brings its Sparky commerce agent to ChatGPT as an in-platform app experience, following the October 2025 OpenAI partnership allowing customers to discover and buy Walmart items inside ChatGPT — 2026 · publ. 2026 · source ↗
  2. ReportedWalmart's own risk factors warn that "the increasing role of AI-enabled platforms in product search, discovery, advertising and purchasing" could reduce traffic to its stores and clubs.
    Walmart Form 10-K, fiscal year ended January 31, 2026 - Item 1A Risk Factors and Item 5 (the eCommerce concentration and AI-enabled platform risk factor; five-year cumulative total return with $100 invested on February 1, 2021 worth $272.28 in Walmart against $201.03 in the S&P 500 and $164.12 in the S&P 500 Consumer Discretionary Distribution and Retailing Index; the $20.0 billion November 2022 repurchase programme with $4.0 billion remaining and the new $30.0 billion authorisation approved February 2026) — FY2026 (ended January 31, 2026) · publ. March 13, 2026 · source ↗
  3. ReportedIt was 38% in the June 2026 quarter.
    Walmart second-quarter fiscal 2027 earnings release (Form 8-K exhibit 99.1) - revenue $187.9 billion +5.9% and +5.1% in constant currency; operating income +28.8% and +17.4% adjusted in constant currency; global eCommerce +23%; global advertising +38% with Walmart Connect excluding VIZIO +43%; membership fee revenue +17%; gross profit rate +96 basis points and Walmart U.S. +158; Walmart U.S. comp sales +2.6% and operating income +20.6% with operating expenses deleveraged 72 basis points; Sam's Club operating income +44.3% and International +16.6%; ROA 8.0% and ROI 15.4%; free cash flow $5.5 billion; 42.3 million shares repurchased for $5.1 billion year to date; FY27 guidance of 4.0-5.0% net sales growth, 7.0-8.5% adjusted operating income growth and $2.80-$2.87 adjusted EPS — Q2 FY2027 · publ. August 20, 2026 · source ↗
Sources
Generated September 22, 2026