To examine the feasibility and implementation of an optimal defaults intervention designed to align grocery purchases with a diet recommended for people with or at-risk for type 2 diabetes.

This was a 5-week pilot randomised trial with three groups: in-person grocery shopping, shopping online and shopping online with ‘default’ carts. Participants were asked to shop normally in Week One, according to group assignment in Weeks Two–Four (intervention period), and as preferred in Week Five. All groups received diabetes-friendly recipes via email each intervention week.

Participants grocery shopped in person or online. Grocery receipt forms, enrolment information and exit surveys were collected remotely and used to assess feasibility and implementation.

Sixty-five adults with or at-risk for type 2 diabetes.

Sixty-two participants completed the exit survey and fifty-five submitted receipts all 5 weeks. Forty utilised recipes, 95 % of whom indicated recipes were somewhat or very useful. Orange chicken, quesadillas and pork with potato and apples were the most liked recipes. Most Defaults group participants accepted at least some default cart items. Recipes with the highest default acceptance were whole grain pasta and chicken, quesadillas with black beans and chicken with olives. Participants’ primary concerns about the intervention were costs associated with online shopping, inability to select preferred foods and some recipes including ingredients household members would not eat.

The study had high retention, data were successfully collected remotely and the intervention was acceptable to most participants. Tailoring recipes to household preferences may be beneficial in future studies.