Consumers make purchase decisions based on their personal Taste. When a Sales clerk asks, “May I help you?”, invariably the consumer responds with information about his or her Taste. “I’m looking for a slender dress, ideally in a lighter color, just above the knee, not too revealing, not too conservative, under $150.”

Netflix has succeeded in creating a user experience that’s reflective of your Taste in movies, and Spotify’s Machine Learning system has the goal of learning and presenting songs and playlists that match your musical Taste. What about retail? Shouldn’t retailers be equally responsive to your Taste in apparel, footwear, and cosmetics? Given all the data that’s out there, shouldn’t this happen when you visit a store, when you browse online, or when you receive an email from a retailer?

Read more: Not All Product Recommendations Are Created Equal

Specifically, what is Taste?

When retail products are conceived the designer thinks in terms of product attributes. An attribute can be as simple as the color or price, but it can also include the type of fabric, the fit, whether its machine washable, or if there’s a prestigious logo. Imagine you were designing a car: Is it going to be sleek or boxy? Fuel efficient or high performance? A sporty eye magnet, or a sedate sedan? Apparel and other soft goods go through a similar design journey. The product designer’s goal is to create a product with the attributes desired by the target customer. If you were asked to describe your Taste in sneakers it would sound like a list of attributes: “I like running style shoes, cushioned heel, between $60 and $80, usually blue or white color, white laces, white sole, and a mainstream brand.”

We recently performed an analysis of over 5 Billion retail transactions, which showed that consumers have very weak affinities to product brands. They might like a brand, but the data indicates they also like adjacent brands. The study showed that consumers switch product brands frequently, suggesting that the purchase of a product brand is a relatively weak predictive signal for repeat purchases. Our analysis went on to indicate that customers made repeat purchases of products that had similar attributes to their prior purchases in a category. Taken in the aggregate, this data can indicate a consumer’s overall “Taste.”

Consumers don’t seem to have brand names in mind during their shopping journey, but they demonstrate very strong affinity to their unique retail Taste. Imagine walking into a major department store and wearing a pair of special glasses that make the products that fit your taste glow yellow. That’s what a truly tailored digital retail experience should feel like.

Customizing digital experiences based on taste should actually deliver that long-promised goal of data-driven Marketing: customer relevance. Studies have shown that consumers are receptive to Digital Advertising that actually speaks to their preferences. A product that accurately fits a person’s Taste should elicit the same positive response as a spot-on recommendation from a trusted friend or knowledgable merchant.

With the decline of legacy fashion publishing, consumer Taste is more fractured and opaque than ever before. If retailers are going to keep up (as opposed to surrendering to a world dominated by one monolithic e-retailer), they need to be responsive to their customers’ personal Taste the same way that other popular platforms are. Companies like Netflix and Spotify used data to tailor recommendations according to individual preferences, and retail should do the same. With the right technology properly applied, consumers get product matches that actually interest them and retailers get a much greater return on their marketing investment. If you can deliver experiences based on Taste, everybody wins.

Read more: Recommendations Are the New Dashboards