California-based grocery technology company Instacart has introduced Smart Shop, which uses generative AI and machine learning to create a more personalised shopping experience.

Smart Shop can make online grocery shopping more intuitive by analysing customer habits and dietary preferences to surface the most relevant products faster.

Instacart has also unveiled AI-powered Health Tags, which provide detailed and transparent nutritional information across the catalogue, and Inspiration Pages, curated destinations featuring expert-backed health recommendations and shoppable recipes. The first Inspiration Page was developed in collaboration with the American Diabetes Association (ADA, diabetes.org) and features evidence-based nutrition guidance and diabetes-friendly grocery and recipe recommendations.

Launched to coincide with National Nutrition Month, these offerings make it easier for consumers to discover relevant products based on their health and lifestyle preferences on Instacart.

“At Instacart, we want to turn the ordinary task of grocery shopping into a delightful, personalised shopping experience that takes the mental load out of finding the exact items that meet your preferences,” said Daniel Danker, chief product officer at Instacart. “By combining our new Smart Shop technology, Health Tags and Inspiration Pages, we’re not just improving online grocery shopping, we’re reimagining it, making it seamless to go from intention to action. By customising your shopping journey to match your personal health goals or fit your dietary restrictions, we can unlock possibilities that weren’t even on the table before.”

These features help consumers shop according to their dietary and household preferences, simplifying the process of finding relevant products and making more informed grocery choices.

Powered by Instacart’s catalogue of 17 million items and its proprietary dataset of millions of grocery shopping journeys, Smart Shop leverages insights into nuanced consumer behaviour. The technology applies behaviour analysis and affinity modelling to analyse real-world shopping habits, identifying patterns in user preferences with precision.

To refine this understanding, large language models (LLMs) improve personalisation by identifying the most valuable signals for predicting preferences to enhance accuracy. For example, inferring a customer’s preference for low-carb products is complex because many everyday items are naturally low in carbohydrates, making it difficult to determine true intent. Traditional machine-learning models rely on large datasets and extensive feature engineering to differentiate between routine purchases and deliberate low-carb choices.

Items such as almond flour or cauliflower rice strongly indicate a low-carb preference, whereas staples like chicken breast, though low in carbs, are less indicative of intent in this context.

LLMs simplify this process by analysing product categories and detecting intent-based patterns in user behaviour. They pinpoint deliberate low-carb choices and filter out noise from irrelevant signals such as purchases of naturally low-carb items. By automating the identification of meaningful patterns, LLMs eliminate the need for exhaustive tagging and manual feature engineering, reducing the time and effort required to create accurate, scalable predictions, while also enabling more precise and personalised recommendations, even in scenarios it hasn’t been explicitly trained on.

As users shop, Smart Shop evolves its understanding in real time, dynamically adjusting digital aisles to prioritise the most relevant products based on inferred preferences, while seamlessly adapting to changing behaviour over time. When confidence in a preference is low, Smart Shop proactively engages users with clarifying questions such as “Show more low-carb options?”. These interactions, combined with Instacart’s proprietary shopping data, ML models and LLM reasoning, continuously sharpen Smart Shop’s ability to predict user needs and refine recommendations.

To tailor the Smart Shop experience further, customers can modify their shopping preferences, selecting from 14 dietary preferences – gluten free, high protein, high fibre, lactose free, low calorie, low carb, low sodium, low sugar, low fat, organic, pescatarian, preservative free, vegan and vegetarian. Through preferences, they can explicitly share household details, such as whether they’re also shopping for a baby, toddler or pet. All these inputs enhance personalisation, helping surface items and aisles on Instacart that better align with each person’s dietary and household preferences.

The AI-driven Health Tag system scans product data at scale to extract key nutrition characteristics. This multimodal AI platform leverages vision-language models (VLMs) to extract rich product attributes from text and images to enable rapid, high accuracy tagging of nutrition claims, ingredients and dietary labels. Using this model, Instacart has tagged over 1.3 billion data points across food and beverage products in its catalogue and, today, Health Tags provide transparent easy-to-digest details on approximately half a million items across the more than 100,000 stores on the Instacart platform.

Instacart’s 30 Health Tags now include gluten free, grass fed, heart healthy, high fibre, high protein, keto, lactose free, minimally processed, pasture raised, preservative free, no artificial colours, no artificial flavours, non-GMO, vegetarian, vegan, and wild caught, helping consumers view and filter items based on their preferences, discover new products and make more informed shopping decisions.

Instacart is also introducing Inspiration Pages, which will provide consumers with curated product selections and tailored recommendations, making it easier to find foods that align with their preferences and goals.

“Instacart is making it easier for people to shop for nutritious foods that support their health goals and lifestyle,” said Sarah Mastrorocco, vice president at Instacart (www.instacart.com). “With Smart Shop technology and Health Tags, we’re giving consumers the power to personalise their experience, with tools to filter and discover the best options for their unique preferences.”