Walmart Sparky By The Numbers: Agentic Commerce Operator Playbook

Agentic commerce stopped being a research slide this month. On the Q1 FY27 earnings call, Walmart shared real numbers from Walmart Sparky, its in-house AI shopping agent, and the gap between shoppers who use Walmart Sparky and shoppers who do not is already wide enough to change how brands plan paid media, content, and attribution. If you sell anything online, the Walmart Sparky results are the clearest signal yet that AI agents are becoming a real channel, not a future bet.

Below is what Walmart Sparky just proved, what it means for marketers, and the four Walmart Sparky moves your team should make this quarter so you are ready when agentic shopping shows up in your own funnel.

What Walmart Sparky Just Proved On The Earnings Call

On the May 21 call, Walmart US CEO John Furner said Walmart Sparky shoppers spend 35 percent more per order than non-users. Weekly active users on Walmart Sparky grew more than 100 percent in the quarter. Units purchased through Walmart Sparky more than quadrupled versus the prior quarter. And Walmart Sparky itself is 40 percent smarter, measured by response quality, than it was at the start of the year. You can read the full release on the Walmart Q1 FY27 earnings page.

Walmart Sparky now lives across the Walmart website, the app, and inside physical stores. It supports Spanish speakers. New features include personalized replenishment, meal planning, and inventory-aware product suggestions that pull from local store stock. Analyst coverage from Constellation Research framed the launch as the most public proof point to date that agentic commerce moves real revenue, not just press cycles.

Two other numbers from the same call matter for paid media teams. Walmart US advertising revenue grew 36 percent year over year. Marketplace seller ad spend on Walmart Connect grew more than 50 percent. The same shoppers who lean on Walmart Sparky are sitting inside a retail media network that is scaling faster than the storefront itself.

Why Walmart Sparky Changes The Game For Brands

Most ecommerce playbooks still assume a human typing keywords into a search bar. Walmart Sparky breaks that assumption. The shopper tells Walmart Sparky what they want, the agent reads product feeds, reviews, inventory, and past behavior, and then Walmart Sparky picks. That shift moves the buying decision one layer up the stack, away from your product page and toward the data the agent can see.

That is why the 35 percent higher average order value from Walmart Sparky matters. When the agent bundles a meal plan or a replenishment list, it adds adjacent items the shopper would not have clicked on alone. If your products are not surfaced inside that Walmart Sparky bundle, you lose the basket even when your ads run on the same query. The agentic layer is where assortment, content, and feed quality become your new shelf space.

The pace of change is real too. By the end of Q4 FY26, half of all Walmart app users had already tried Walmart Sparky at least once, according to ecommerce analytics firm Epinium. That is a faster adoption curve than mobile checkout took ten years ago, and it is happening at the largest retailer in the country. Google is moving the same direction with Ask Advisor, which we covered in our Google Marketing Live 2026 recap, and Meta is restructuring its ad stack around AI agents as well, as we wrote in the Meta AI ads reorg breakdown.

The New Funnel Looks Different When Agents Pick The Products

When Walmart Sparky does the picking, the classic awareness, consideration, purchase funnel collapses. The shopper expresses intent in one sentence. The agent compresses research, comparison, and checkout into a single response. Three things change for your team right away.

First, last-click attribution breaks even harder. The Walmart Sparky visit looks like a single session, but it replaces five or six visits across search, brand site, and review pages. We unpacked this shift in how AI Mode killed last-click attribution. Your dashboards will keep showing fewer touchpoints and higher AOV per session, and that pattern is not an error, it is the new normal.

Second, your product data becomes the ad creative. Walmart Sparky is reading titles, bullets, attributes, and reviews to decide what to recommend. Brands with clean, complete feeds and rich structured data are going to win Walmart Sparky recommendations. Brands with thin product pages are going to lose them quietly. Our take on schema and AI citations is in the new SEO metric that actually matters.

Third, agent-driven sessions are harder to track without first-party data. Walmart Sparky lives inside Walmart, so Walmart owns the signal. When you advertise on Walmart Connect, you can buy your way back into that view. When you do not, you are guessing. The same logic applies to Amazon Rufus, Google AI Mode, and any other agent surface that touches your category. We covered the broader bot identity problem in how Google-Agent is changing bot authentication.

What To Tell Your Walmart Sparky Strategy Team This Week

If you sell on Walmart, treat the next 60 days as a forced sprint. Walmart Sparky is already converting your category, so you need to know whether it is recommending you or your competitor. Pull a sample of high-intent queries in your top categories and run them through Walmart Sparky yourself. Note which brands surface, in what order, and what attributes Walmart Sparky cites. That output is your new shelf audit.

Next, treat your Walmart product feed like a paid asset, not a static catalog. Update titles, bullets, and attributes weekly. Add comparison-friendly fields like serving size, capacity, compatibility, and dietary tags. Walmart Sparky uses these fields to fill bundles and replenishment lists. Missing data equals missing baskets.

Then look at Walmart Connect. With Walmart US ad revenue up 36 percent and seller ad spend up more than 50 percent, the auction is getting more competitive every quarter. Sponsored Products and Sponsored Brands still work, and they now also act as a hint to Walmart Sparky about which items deserve top placement. Pair them with retail media display to defend high-value queries and keyword categories.

If you do not sell on Walmart, the same playbook still applies. Pick the two or three agent surfaces that matter most in your category, audit your visibility there, and route budget toward the data layer the agent uses. Our guide to connecting your data to ChatGPT and Claude walks through how to feed agents directly.

Build A Reporting Layer That Sees Agent Traffic

The hardest part of agentic commerce is not the strategy, it is the measurement. Most analytics setups still bucket agent visits into direct or referral traffic and stop there. That hides the real story. Your team needs three new reports running by end of quarter.

The first report is Walmart Sparky share of voice. For your top 25 priority queries, log which brand Walmart Sparky picks, which products it ranks, and how often you appear. Refresh weekly. Treat the trend line like a stock price.

The second report is agent-influenced revenue. Use UTM tagging, server-side tracking, and post-purchase surveys to estimate how many orders had at least one Walmart Sparky or AI agent touchpoint. Our piece on dual attribution for app performance shows the same pattern for social, and the same dual-source approach works here.

The third report is feed health. Track completeness, freshness, and structured data coverage by SKU. When an agent skips your product, the cause is usually a missing attribute, not a missing ad dollar. We covered the related search side of this in the AI search terms report change.

Where Agentic Commerce Goes From Here

Walmart is not stopping at Walmart Sparky. The company also confirmed a supplier-facing agent called Marty, which will let brands and sellers query Walmart data and run merchandising tasks through natural language. That mirrors what Google is doing with Ask Advisor for advertisers, and what Meta is rolling out for agency teams. Within twelve months, the same agent layer that picks products for shoppers will pick budgets, creatives, and bid strategies for the people selling to them.

The brands that win this cycle are not waiting for a perfect playbook. They are running real Walmart Sparky audits, fixing feeds, buying retail media to defend share, and rebuilding reports to see agent traffic. That is exactly the kind of work our team does with clients every week. If your category is moving fast and you want a partner who can help you adapt, book a free consultation and we will map out your agent-readiness plan together. Let’s Grow!