Appetise has released research showing shoppers in New Zealand and Australia are willing to use AI for meal planning but resist AI product selection. The findings are based on analysis of more than 90,000 shoppers, 8.0 million product events, and 17.6 million recipe events.
The main resistance point comes when AI moves from suggesting what meal to cook to choosing which brand, pack size, or product format goes into the basket. Appetise describes this shift as a trust threshold, with shoppers more willing to accept help at recipe and category level than at the shelf-selection stage.
The study identified three main reasons shoppers override AI-generated product suggestions: switching to a cheaper option in the same category, choosing fresh instead of processed products, and changing the pack size or variety to suit their household.
The pattern suggests the main constraint is not whether shoppers believe the technology works, but how much control they want to retain over the final purchase decision.
Demographic differences also emerged. Men were 41% more likely than women, on average, to accept assisted shopping suggestions, while shoppers aged 25 to 44 accepted them at nearly twice the rate of 18- to 24-year-olds.
Household size also appeared to influence confidence in automated suggestions. People cooking for six overrode suggestions 28% of the time, compared with 18% for those cooking for two.
Buying signals
The report argues that AI shopping agents assess products differently from shoppers standing in a supermarket aisle. Instead of responding to packaging, shelf position, or promotional displays, these systems rely on price, nutritional data, ingredient lists, review scores, and past purchase behaviour.
If automated systems play a larger role in grocery decisions, that shift could change how brands compete for attention. In that environment, on-shelf presentation may matter less than the quality and precision of structured product information.
"In a few years, the grocery shopping experience will completely change as we know it today, and those that see this coming have a real opportunity in front of them," said Toby Hilliam, Co-founder and Chief Executive Officer of Appetise.
"I believe AI is going to own the entire grocery shopping journey. Not part of it, all of it. At Appetise, we're already watching it happen," Hilliam said.
He also addressed the implications for consumer goods groups trying to influence digital selection. "A machine isn't swayed by beautiful packaging or a clever campaign," he said.
"It looks for functional claims, nutritional data, and proof you match the exact need a shopper has. For the brands ready to show up that way, that's a level playing field and a genuine chance to leapfrog the category," Hilliam said.
Business pivot
The findings come as Appetise expands after changing its business model. The company began as MenuAid during pandemic lockdowns and later rebranded, moving from a consumer subscription approach to a free-to-use model funded by the sale of aggregated behavioural data to food brands.
According to Appetise, that change lifted revenue from $1.2 million to $3.5 million in six months. The company also raised $7 million in a Series A round led by Icehouse Ventures, with backing from Ten13, NZVC, K1W1, OIF Ventures, and Brand Fund.
Mason Bleakley, a Principal at Icehouse Ventures, linked the investment case to the company's two-sided model. "Appetise is pulling off the difficult feat of growing two entirely different products simultaneously; a genuinely helpful tool for making meals, and a platform that brands can make important business decisions with," said Mason Bleakley, Principal at Icehouse Ventures.
Appetise says it now has 110,000 active users and operates one of the largest food and beverage insights panels in Australia. It says that user base is larger than more traditional research panels of 18,000 respondents in Australia and 2,000 in New Zealand.
Market differences
The dataset has also highlighted differences between the two countries, according to Appetise. Australian shoppers place more weight on protein, fat content, and macronutrients than New Zealand shoppers, and are more willing to try new products instead of defaulting to discounted items.
Appetise says broader acceptance of AI-led basket decisions could have a commercial effect beyond the point of sale. When users accept AI suggestions across their basket, downstream engagement rises by 11 to 25 times across every metric the company tracks.
The figures suggest AI-assisted grocery shopping is advancing unevenly, with routine planning decisions increasingly delegated while final purchase choices remain more closely guarded by consumers.