Groceryshop 2026 Recap | Incisiv

Groceryshop 2026:
Key Themes & Strategic Insights

Las Vegas, Nevada | Incisiv

Three days, one recurring pressure point: grocery has the AI tools it needs. What it lacks is the operating model to scale them. This report captures the main themes from Groceryshop 2026.

Overview

We spent September 22–24 at Mandalay Bay with the grocers, brands and technology providers shaping the industry. Agentic commerce, GLP-1 and retail media dominated the agenda. The conversations that mattered most were about execution: making online grocery profitable, scaling AI beyond pilots, and adjusting to a consumer who is buying less and choosing more deliberately.

7%
of firms have scaled AI across the enterprise, even though access to frontier models is now table stakes
16.5%
of retailers and CPGs can quantify the ROI of AI, while 75% call it strategically critical
21%
of U.S. households now include a GLP-1 user, double the share of 16 months ago
93%
of U.S. grocery sales still happen in physical stores

Key Themes & Strategic Insights

Theme 1: Grocery's AI Gap Is an Operating-Model Gap

No one at Groceryshop questioned whether AI works. Access to models like ChatGPT and Claude is now table stakes. Scaling them is the problem. Only 7% of firms have scaled AI across the enterprise, and of every 100 AI ideas, only about 16 make it through security, IT and business reviews to testing or implementation. The most common reason projects stall is that no one can prove the return: 75% of retailers and CPGs call AI strategically critical, but only 16.5% can quantify its ROI.

The fix the panels kept coming back to was organizational, not technical. That means one accountable owner with budget authority instead of a committee, and governance written so the people using it can understand it. It also means a culture that treats failures as lessons instead of repeating them. As Mars Snacking put it: "Speed without governance is risk. Governance without speed is irrelevant."

Five pillars: Incisiv's research points to five things that close the operating-model gap: champions below the executive level, a culture of learning, single ownership, an integrated business–IT relationship, and evidence that proves value.

From customer experience to margin: CEO attention has shifted from AI for customer experience toward AI for operational efficiency, which protects thin grocery margins and funds the next round of innovation.

Three horizons: Church & Dwight plans AI across everyday productivity, workflow redesign and transformational bets, and trains employees by persona (analyst, researcher, creative, project manager and executive).

Hub and spoke: an AI council or center of excellence sets shared priorities and guardrails, while local champions and functions handle execution, so governance doesn't become a bottleneck.

Good enough to ship: Rohlik Group treats an agent as ready for production once it reaches 60–70% of human accuracy, as long as feedback loops keep it learning after launch.

Build trust before autonomy: Fruitist moved its harvest models from optional tools to the primary driver of decisions in seven months by following a "build trust, train, co-own, prove" sequence.

Key Takeaways
  • Put AI accountability with the P&L owner, not the technology team. Technology provides the tools; the business owns the outcome.
  • Before funding any AI investment, answer three questions: what will be true if it succeeds, how will it show up in a metric or the P&L, and who is responsible for delivering it?
  • Redesign the process with the people who run it. Layering AI onto an old workflow automates the old workflow.

Theme 2: Online Grocery Profitability Is a Design Decision

The online grocery panel showed three very different routes to the same goal. MC Sonae runs ecommerce on a fully loaded P&L, as if it were a standalone business, and fulfills from its existing 400-store network rather than opening dark stores that would cannibalize physical sales. Rohlik Group runs a pure-play model on 12 fulfillment centers, with baskets of about 30 units and €100, a target density of one order per 10 square feet and a 95% perfect-order rate. Shipt ties profitability to membership-driven frequency and to "preferred shopper" relationships that build trust.

None of the panelists argued that one model wins everywhere. Rohlik's rule of thumb was to pick in stores at low volume and move to dedicated fulfillment centers once a market passes 2,000–3,000 orders a day. Misfits Market showed what scale can unlock: it has grown 50x in two and a half years and built 750,000 square feet of refrigerated and frozen space. It now rents that capacity to other perishable brands, and the third-party logistics business has reached nine figures in revenue on its own.

Machine learning and shrink: one operator has brought shrink down to 0.5% of gross revenue by using machine learning to refine markdowns and ordering.

Accuracy beats speed: Rohlik measures success by perfect orders (no delays, claims or substitutions). Shipt handles inventory uncertainty by letting shoppers suggest substitutions and send photos so customers stay in control.

Basket economics: Rohlik raises order value by carrying multiple temperature zones (ambient, frozen, bakery and pharmacy) under one roof. Shipt grows share by expanding into "life moments" such as pet care and apparel.

Room to grow: ecommerce accounts for 65–70% of sales in clothing and electronics but only 20–25% in grocery, according to Misfits Market.

Key Takeaways
  • Choose a fulfillment model based on order density, not ambition. Store-pick works at low volume; dedicated fulfillment pays off only at scale.
  • Measure ecommerce on a fully loaded P&L. Anything less hides whether it really makes money.
  • Reliability (perfect orders, smart substitutions, accurate stock) drives retention more than delivery speed.

Theme 3: Agentic Commerce Is Shortening the Shelf

Kantar's research shows 60% of consumers already use AI for shopping and 45% use it daily. Shoppers are moving from keyword search to missions: instead of searching for an ingredient, they ask for a recipe or a week of meal plans, and the agent builds the basket. That pushes the brand decision much earlier in the journey and leaves fewer products in view. Euromonitor International reported that GenAI traffic has grown 376% since January 2025. Today, 60% of Misfits Market carts are filled automatically by algorithms before the customer touches them.

Retailers are taking different positions. Amazon builds its own agents and blocks third-party ones to protect its review data and traffic. Walmart runs its own agent, Sparky, while staying open to outside platforms. Brazil's Magalu put its agent, Lou, inside WhatsApp, where it converts better than the company's own app, lifts NPS and reduces service calls. Shipt has launched an integration with ChatGPT and an embedded agent that turns a photo of a meal into a shopping list.

Where AI matters most: categories that need more knowledge and personalization, such as beauty and personal care, see the biggest influence from AI referrals. Low-knowledge staples like dairy see little change.

Protect direct traffic: exclusives, better prices and add-on services give shoppers a reason to come straight to the retailer instead of starting with a third-party agent.

Own the proprietary data: returns data and deep review data are what a generic agent can't copy, and retailers need to keep them.

Retail media is exposed: when an agent skips the search box, it also skips sponsored search. Retailers will need new ad units built on intent data.

Key Takeaways
  • Make product data readable by machines and consistent with what shoppers say about you on Reddit, YouTube and elsewhere. Agents check both.
  • Decide your agent strategy deliberately: build your own, partner, or meet customers on the platforms they already use.
  • Start planning now for a retail media model that doesn't depend on sponsored search.

Theme 4: GLP-1 and Budget Pressure Are Rewriting the Value Equation

GLP-1 adoption has doubled in 16 months, and 21% of U.S. households now include a user. Grocery spending in those households drops 5.3%, and 8.2% in higher-income households. Only 17% of grocery executives expect a 5–10% hit to sales, but with 72% of U.S. adults overweight, the long-term effect could be much larger. At the same time, 80% of Americans are trying to cut spending, and one in four is targeting groceries: 56% are trading down, 49% are buying fewer items and 44% are relying on discounts.

The food and beverage panel described a consumer moving away from "hollow bites" toward purposeful eating. They buy for an outcome, eat smaller and more often, and judge value on more than price. Target has made food a core part of its business through Good & Gather and increased newness in its food assortment by 50% this year, aiming for "the best, not the biggest" assortment. Chobani anchors on "food truths" rather than fleeting trends, and uses limited drops, like a recent blueberry creamer collaboration with Target, to create excitement.

Value has more than one dimension: formulation, transparency, experience and emotional connection now sit alongside price. Shoppers read the back of the pack, and brand loyalty is declining.

Two buyers, one product: Roxberry Fizz for Kids has to win over parents, who care about nutrition, and kids, who care about taste. It does that with fun branding and strong flavor.

Innovation in months, not years: brands are using social listening, cross-category signals from sports nutrition, energy drinks and beauty, and retailer data to shorten development cycles.

The five-second rule: Target's advice to emerging brands is to make the hook clear almost instantly, find real white space, and listen more than you sell.

Key Takeaways
  • Model GLP-1 into category and assortment plans now. The early data points to a bigger effect than most executives expect.
  • Compete on outcome and proof, not just price. Shoppers are trading down and trading up in the same basket.
  • Cut innovation cycles from years to months by connecting social signals, retailer data and product development.

Theme 5: The Store Is Still Where Grocery Is Won

With 93% of U.S. grocery sales still happening in physical stores, several sessions pushed back on digital-only thinking. Retail media should prioritize in-store placement, where buying decisions are made, and can help offset margins that average about 2% across the industry. Digital signage is now among the top three store technology investments for grocers, ahead of mobile devices and loyalty programs.

Grocery executives are also unusually positive about AI's effect on their workforce: 85% view it favorably, and most expect associates to become advisors and product specialists rather than be replaced. To get there, 92% say associates need AI-enabled mobile devices.

In-store media: 58% of brand executives expect placement on retailers' properties, including in-store displays, to matter significantly more over the next five years.

Store tech priorities: digital coupons and buy online, pick up in store remain priorities, with digital signage climbing fast.

Image recognition on the shelf: CPG companies are using image recognition in stores to check brand execution and reduce time spent on store visits.

Parity across channels: MC Sonae keeps the online and in-store value proposition the same and uses its loyalty ecosystem to personalize offers across both.

Key Takeaways
  • Build retail media strategy around the store, where the decision happens, not just around digital shelves.
  • Equip associates with AI on the floor. The payoff is better service, not fewer people.
  • Treat in-store technology as part of one connected system with ecommerce and loyalty.

Closing Perspective

Across three days, one gap kept coming up: the distance between what grocers know they should do and what their organizations are built to deliver. AI is widely available but rarely scaled. Agents are reshaping discovery faster than retail media models can adjust. GLP-1 and budget pressure are changing what a basket looks like. Online grocery can be profitable, but only for operators who design for density and reliability from the start.

The grocers and brands making progress aren't the ones with the most pilots. They put one owner on every AI initiative, measure ecommerce honestly, protect the data that makes them different, and keep the store at the center of the plan.

The common thread: in grocery, the advantage now comes from operating discipline, not access to technology.

Incisiv at Groceryshop

Tapping Data Analytics and AI for Productivity and Profitability

Gaurav Pant Moderated by Gaurav Pant, Co-Founder & Chief Insights Officer, Incisiv

Dalibor Patek Brian Bell Renato Desio With Dalibor Patek, VP, Internal Systems & Process Excellence, Rohlik Group; Brian Bell, VP, Strategy & Planning, Church & Dwight; and Renato Desio, Head of Global Ecommerce, Kimberly-Clark

On Wednesday, September 23, Gaurav led a Capsule Stage panel on the operational side of AI, where Incisiv's research consistently finds the gap between pilots and P&L impact. He opened with the data: only 7% of firms have scaled AI across the enterprise, and only about 16 of every 100 approved AI ideas make it through security, IT and business reviews to implementation. Eighty percent of companies use AI for generic work, but few scale it across departments, and the biggest reason is that they can't measure the return. CEO attention has also shifted from AI for customer experience to AI for operational efficiency, which funds the next round of innovation.

Gaurav laid out Incisiv's framework for how AI matures along two dimensions: how much context it has and how much permission it has to act. With little of either, AI is a tool. With more context, it becomes a thought partner that makes recommendations. With both, it becomes a colleague that acts on its own, for example by executing markdowns. As agents rather than people start to drive demand, the accuracy and accessibility of a company's data will decide whether it gets chosen. Closing the operating-model gap takes five things:

Champions: employees below the executive level who understand the work in detail.

Learning: treating failures as lessons rather than blaming people for them.

Ownership: a single owner with budget authority, not a committee.

State: an integrated working relationship between business and IT, not a siloed or transactional one.

Evidence: baseline measures that prove value and track progress.

The panelists showed what this looks like in practice. Brian Bell described Church & Dwight's three-horizon approach: everyday AI wins, more efficient workflows, and transformational rethinking. Church & Dwight backs it with training built around five personas: analyst, researcher, creative, project manager and executive. Renato Desio argued against layering AI onto old processes, saying new workflows have to be designed together with the employees who run them.

The conversation then turned to governance. Dalibor Patek described Rohlik's hybrid model: global guardrails, local power users and champions, and sandboxes where employees can build agents without risking production systems. Agents are monitored as rigorously as any critical IT system, with escalation paths and service levels. An agent can go live at 60–70% of human accuracy, as long as feedback loops keep it learning. Brian Bell added a simple test for every AI investment: what will be true if it succeeds, how it will show up in the P&L, and who owns it.

The panel left the audience with three open questions: how retailers and CPGs can work together on shared AI tools, how brands will handle disintermediation as agents change discovery, and whether employees will trust autonomous agents with core processes.

Incisiv Community Happy Hour

On Tuesday, September 22, we brought together clients and prospects at Rí Rá for an evening of good drinks and candid conversation. People compared notes on what they were hearing on stage and what it means for the year ahead.

Session Highlights

Day 1 · Tuesday, September 22

AI-Powered Shopping: New Insights on What's Truly Driving Adoption

(Kantar, Euromonitor International, Bain)

Kantar found that 60% of consumers use AI for shopping and 45% use it daily, from the 22% who are "all in" to firm rejectors. Shopping is becoming mission-based: shoppers ask for meal plans, not ingredients. Euromonitor International reported that GenAI traffic has grown 376% since January 2025, with the most influence in categories that require more knowledge and personalization. Retailers are taking different paths: Amazon builds its own agents and blocks others, Walmart builds and partners, and Magalu goes where customers already are, on WhatsApp.

Why it matters: The shelf is getting shorter. Retailers need to protect direct traffic and proprietary data, and rethink retail media for a world where agents skip the search box.

Delivering Value, Speed, and Profitability in Online Grocery

(Rohlik Group, MC Sonae, Shipt)

Three models, three routes to profit. MC Sonae uses a fully loaded P&L and fulfills from 400 stores. Rohlik runs 12 fulfillment centers with ~€100 baskets and a 95% perfect-order rate. Shipt builds frequency through membership and trusted shoppers. One operator has cut shrink to 0.5% of gross revenue with machine learning, and Shipt previewed its ChatGPT integration and an agent that turns a photo of a meal into a shopping list.

Why it matters: Online grocery profitability depends on density, basket economics and reliability, not on any single model.

Centering Products and Assortment Around Key Dietary Preferences

(Chobani, Roxberry Fizz for Kids, Target)

Consumers are moving from "hollow bites" to purposeful eating: buying for outcomes, eating smaller and more often, and judging value on more than price. Target increased newness in food by 50% this year and aims for "the best, not the biggest" assortment. Chobani anchors on "food truths" and uses limited drops to create excitement. Roxberry Fizz for Kids has to win over both parents and kids. Brands are using social listening and retailer data to cut innovation cycles from years to months.

Why it matters: Value now means formulation, transparency and experience as well as price, and brands have about five seconds to make the case.

Day 2 · Wednesday, September 23

Tapping Data Analytics and AI for Productivity and Profitability

(moderated by Gaurav Pant, Incisiv; with Rohlik Group, Church & Dwight and Kimberly-Clark)

Only 7% of firms have scaled AI across the enterprise, and only about 16 of every 100 approved ideas reach implementation. Gaurav presented Incisiv's framework, in which AI moves from tool to thought partner to colleague as it gains context and permission to act, along with the five pillars that close the operating-model gap: champions, learning, ownership, an integrated business–IT relationship, and evidence. Panelists made the case for process redesign with employees, persona-based training and governance that enables rather than blocks. Rohlik puts agents into production at 60–70% of human accuracy and relies on feedback loops to keep improving them.

Why it matters: The barrier to AI in grocery is ownership and measurement, not technology.

AI Transformation: Establishing the Right Tech, Teams, and Privacy Guardrails

(Mars Snacking, Fruitist, Associated Supermarket Group)

75% of retailers and CPGs call AI strategically critical, but only 16.5% can quantify its ROI. Advantage now comes from proprietary context, not model access: Mars Snacking calls this "context engineering" and aims for "confident velocity," meaning speed with governance built in from the start. Fruitist combines decades of historical data with satellite imagery to forecast. Associated Supermarket Group set three guardrails for every AI use: what it is allowed to do, what data it can use, and who is accountable for the outcome.

Why it matters: Proprietary data and context are the moat. Governance done right speeds teams up.

Day 3 · Thursday, September 24

The Misfits Market Journey: From 'Ugly' Produce to Full-Scale Online Grocery

(Abhi Ramesh, Founder & CEO, Misfits Market)

Misfits Market has grown 50x in two and a half years on 750,000 square feet of refrigerated and frozen space. Its third-party logistics business, which serves other perishable brands, has reached nine figures in revenue alongside its direct grocery business. Today, 60% of Misfits carts are filled automatically by algorithms.

Why it matters: Cold-chain capacity is becoming a business in its own right, and automated baskets are already here.

Groceryshop's Grocery Industry Outlook

(Brendan Witcher, Global VP, Product & Content, Shoptalk & Groceryshop)

21% of U.S. households now include a GLP-1 user, and grocery spend in those households falls 5.3%, or 8.2% at higher incomes. 80% of Americans are trying to cut spending. With 93% of grocery sales in stores, retail media should prioritize in-store placement. Digital signage is now a top-three store tech investment, and 92% of executives say associates need AI-enabled mobile devices.

Why it matters: The basket is shrinking, and the store remains the main place to win it back.

Groceryshop 2026: Key Themes & Strategic Insights

Meet with Incisiv