What to stock next: Reading shopper needs before they reach your sales data

GLP-1 is the example in this article, and the method works the same way for every category, from fashion to electronics. GLP-1 drugs such as Ozempic and Wegovy treat diabetes and obesity and reduce appetite, so the people who take them buy different food.

Key takeaways

• 22% of US households now have a current GLP-1 user, about twice as many as in October 2023 (Numerator, via Grocery Dive).
• After a year, these households spend about 4% less on groceries: more protein, fiber and functional nutrition, fewer pantry staples, pasta and bakery goods. Circana expects them to account for 35% of all food and beverage units by 2030 (Prepared Foods).
• Shopper needs show up in search, reviews and social media weeks before they show up in your sales data.

What this means for retailers: The shopper needs that shape next year's assortment are already visible in search, social media and reviews today. Retailers who read them per category can list the right products before competitors do.

One shopper need is reshaping US grocery baskets right now. According to Numerator, 22% of US households have a current GLP-1 user, about double the share from October 2023. After a year on the drug, these households spend about 4% less at the grocery store. They buy more protein, fiber, seafood and functional nutrition, and fewer pantry staples, pasta and bakery goods (Grocery Dive). Circana expects GLP-1 households to account for 35% of all food and beverage units sold by 2030 (Prepared Foods), and 55% of grocers say GLP-1 drugs already affected their bottom line (Supermarket News).

Every category has shopper needs like this that change faster than its assortment: smaller pack sizes, cleaner labels, a new flavor, a cheaper store brand option.

Why sales data shows shopper needs late

Sales data tells you what shoppers bought from what you offered. A need you don't serve leaves no trace in it, because shoppers buy elsewhere or skip the item. By the time the gap shows up as lost volume, a competitor has often filled it.

Shopper needs appear earlier in four places:

• Search trends: rising searches such as "high protein snacks" or "small portion meals".
• Ratings and reviews: what shoppers praise, complain about and ask for, at your store and at competitors.
• Social media: products and uses that spread before they reach the shelf.
• Competitor assortments: new items and store brands that others list first.

From shopper need to assortment decision

A trend helps only when it ends in a decision. Three checks get you there:

1. Is the need growing in your category? Look at several months and at the same weeks last year. A spike after one viral video fades; a need that grows for a quarter stays.

2. Do you already serve it? Map the need to your own items and to what competitors list. Often the answer is partly: one protein bar where shoppers expect five.

3. What do you add, and what makes room? Every new item needs shelf space. Pair each addition with SKUs that are losing demand, as described in SKU rationalization without losing shoppers.

For GLP-1 households this can mean more protein and fiber variants in snacks, smaller portions in frozen meals, and less space for large-format bakery items in stores where these households shop.

How AI agents read shopper needs

Reading shopper needs by hand means a category manager checks a few sources now and then. A shopper insights agent reads all of them continuously, for every category at once.

1. It collects raw data from many sources.

• Search interest over time from Google Trends, so you see demand building before it reaches competitor stores
• Social media such as TikTok, Pinterest and Instagram: posts, reels and videos that show which products and uses are picking up
• Reviews and comments: what shoppers praise, complain about and ask for
• Retailer websites, marketplaces and Google Shopping: which products, brands and store brands others list, at which prices
• Price history and competitor assortments
• Economic data such as inflation and purchase power
• Web research: articles and reports, read and connected to your assortment and strategy
• A consumer panel: people from all age groups share what shows up in their feeds, which adds the human perspective to the signals

2. It turns raw data into insights. Data providers deliver raw numbers. The agent reads the signals together, separates lasting needs from short fads and connects each one to products, brands and categories. For GLP-1 it could look like this:

Insight	What it tells you	GLP-1 example
Needs	What shoppers want and don't find yet	High-protein snacks in small portions
Trends and fads	Which signals grow and which fade	Steady growth over months, no one-off spike
Price sensitivity and purchase power	How much shoppers can and will spend	Households spend about 4% less on groceries

3. It turns insights into decisions. Every insight is checked against your own sales, margin and stock and against the market context. Assortment agents filter the insights that fit your strategy and turn them into brand and private label recommendations: what to add, what to keep and what makes room. Pricing agents check each insight against the market and propose the price move within your margin guardrails. Your category team decides what goes on the shelf.