The numbers behind the rollout
Albertsons Companies has disclosed some of the clearest basket size data yet on what conversational AI is actually doing to order value in grocery ecommerce. Standard conversational search drives a 10 percent increase in average basket size, while more comprehensive AI experiences, the kind that help a shopper plan a meal or match a dietary preference, drive a 26 percent increase. Those figures come from 18 months of deployment across every Albertsons banner, including Safeway, Jewel Osco, and Vons, giving the numbers a scale most retailers pushing AI pilots cannot yet match.
The company's digital shopping experiences team frames the lift as a byproduct of how the tools change search behavior rather than a result of promotions or discounting layered on top. Jill Pavlovich, senior vice president of digital shopping experiences at Albertsons, said that when shoppers use a more comprehensive experience they add more to their basket because they stop forgetting items, a distinct claim from AI simply surfacing better product recommendations at checkout. That mechanism, fewer forgotten items rather than more upsells, is a more defensible growth driver for a CFO to underwrite, since it is tied to reducing a real shopper failure rather than to pushing incremental purchases a customer did not originally intend to make.
From three tools to one
Albertsons built its AI shopping experience in pieces: Ask AI, a search bar accessible assistant; Plan AI, focused on meal and list planning; and Buy AI, aimed at completing a purchase. All three are currently accessible through the search bar in the mobile apps across every banner the company operates. That staged rollout let Albertsons test discrete use cases independently rather than betting the entire digital experience on a single, more complex assistant from day one. That sequencing gave Albertsons usage and conversion data on each function separately, which is exactly the evidence it now needs to decide how the merged assistant should prioritize among search, planning, and purchase completion.
The company is now consolidating those three tools into a single conversational assistant, a move that reflects both a maturing product and a recognition that shoppers do not think in the same categories the engineering roadmap did. Pavlovich described the underlying shift in shopper behavior as moving from spearfishing for a single item to shopping across categories once a conversational interface removes the friction of separate searches, which is the behavior driving the larger basket gains. That kind of cross category expansion is precisely what grocers have struggled to engineer through traditional merchandising and promotion tactics, which makes a conversational interface a meaningfully different lever rather than another variant of the same discounting playbook.
Why immediate lift matters more than roadmap promises
Grocery Dive's reporting on the rollout notes that Albertsons took longer to see full ROI on these tools, while basket growth appeared right away once customers used AI for recipe building and dietary preference matching. That distinction, between ROI and basket lift, is worth sitting with. A retailer can show a real behavioral change, larger baskets, cross category purchasing, well before it can prove the tools pay for themselves net of development and infrastructure costs.
For a CTO evaluating a similar build, that gap is the honest part of the story most vendor case studies skip. Albertsons is willing to publish a basket growth number before it can point to full ROI, which is a more credible sequence than claiming both benefits simultaneously from month one. It also suggests the company sees strategic value in publishing early results now, both to justify its own continued investment and to shape how competitors and investors read the category. Publishing a partial result carries some reputational risk if the full ROI case never materializes, which makes the disclosure itself a signal of how confident Albertsons is in where the numbers are heading.
Where this fits in Albertsons' broader AI strategy
The shopping assistants sit inside Albertsons' ACI Edge program, which spans four enterprise AI priorities: digital customer experience, merchandising intelligence, labor optimization, and supply chain optimization. The company has attached a specific financial target to that broader program, aiming for 200 million dollars in savings across the current and next fiscal year. The shopping assistant work is the most visible, consumer facing piece of a much larger internal AI push that touches how the company merchandises, staffs, and forecasts across its store network.
That context matters because it shows the assistant rollout functions as part of a larger, financially accountable program rather than an isolated marketing experiment. It sits inside a plan with a hard savings number attached, which means the basket growth figures are one input among several that leadership is using to justify continued AI spending across the business. Framing the assistant work this way also makes it easier for Albertsons to defend the spend to its own board, since the shopping assistants are one line item inside a portfolio with a defined financial target rather than a standalone bet judged purely on its own merits.
The category level stakes
Albertsons' results land alongside broader projections about how fast AI shopping is scaling. eMarketer estimates roughly 79.6 million people used AI for shopping in 2026, a 25 percent increase year over year, with that figure projected to reach 109.3 million by 2030. The firm also projects AI platforms could drive 13.7 percent of US retail ecommerce sales, worth roughly 225 billion dollars, by 2029. Those are industry wide projections rather than Albertsons specific figures, and they establish the scale of the shift that Albertsons' basket data represents an early, concrete data point within.
Grocery is a useful category to watch here precisely because its margins are thin and its purchase patterns are repetitive, which makes any measurable basket lift from AI more meaningful than the same lift in a category with wider margins to absorb experimentation costs. If conversational AI can move the needle on basket size in grocery specifically, the case for adopting similar tools elsewhere in retail gets considerably easier to make to a skeptical board.
