Consumer personas are like horoscopes. “Nicole is careful with her spending, but under the right circumstances is willing to splurge on herself, if the mood strikes her.” Read carefully, and psychographic profiling is full of generic, unverifiable, ambiguous and often contradictory language that supports a number of interpretations.

Psychographics also mask the inherent unpredictability of our tastes and the complex ways they interact. My sister-in-law lives in an affluent suburb of Chicago. She owns a piece of Away luggage (acquired before its fall from grace) and gets newsletters from Everlane. On the surface, she is a Henry but, in reality, she is a middle-aged married mother of three who now has both brands because they relentlessly pursued her through direct mail discounts until she finally gave in.

Inferring about psychographics based on the products people buy is unreliable. People buy the same things for wildly different reasons: because there’s a discount, or because they are in different moods at different times, or because other people have them.

Equally problematic is personas’ focus on the individual, because it ignores the fact that people are social creatures. They belong to communities and are part of influence networks that they use to decide what to watch, read, buy and pay attention to.

Thanks to the Internet and its numerous influence networks, products across categories are now more susceptible to trends than to individual preferences. A show becomes popular because a lot of people watch it, and it’s entirely possible that a big chunk of the show’s audience is there, not because it reflects their interests or values but because everyone else they know is watching a show and they do not want to be left out (Netflix even unrolled the fast-forward viewing option for those people).

Instead of focusing on individuals, we should focus on their relationships and look at the communities to which they belong.

Netflix’s taste communities are a variation of this idea. This streaming platform’s 125 million global viewers are divided into 2,000 “taste clusters” that group people based on their movie and TV show preferences. At the same time, Netflix content is extensively tagged and based on these tags and their connections, divided into microgenres. Microcommunities and microgenres are then matched up.

Thanks to its banks of data, Netflix goes beyond the psychographics of its average customer. They know the relationship (“matchmaking”) between its members and its content; and also useful things like how many hours of watched content per month makes its subscribers unlikely to cancel.

If we add the social dimension here and apply Netflix’ matchmaking to connect all sorts of products, services and brands with networks of influence to which consumers belong, we get an approach that exceeds efficacy of personas. We get a dynamic portrait of individuals, their sources of influence and their most important relationships and activities. We can understand individuals better if we understand the complexity of their social networks.

When we shift our focus from an individual to their network of relationships, we start asking different questions: how the communities an individual belongs are structured; what is their dynamics; how the influence spreads within them; who are the most active and/or valuable members. This shift reveals not our inferred taste but our actual taste.

Counting how many people carry Away luggage at airport lounges, and in which permutation with other brands, tells marketers very little, aside of the fact that consumers are receptive to persistent and repeated messaging, price deals and social influence. It also focuses attention, dangerously, on the least-valuable customer: the one that’s price-sensitive and trend-susceptible and least likely to be brand-loyal.

Marketers can do better by focusing on one or more of these four scenarios:

Think about your brand in plural. Just as my Netflix isn’t your Netflix, my pair of Off-White sneakers is not your pair of Off-White sneakers. It doesn’t matter that Netflix is a platform and Off-White sneakers are a physical product: when we apply tagging systems and shopping data, each product is worn in a manner that reflects its user.

Grow through the niches. Netflix's brand isn’t its shows. It’s personalization. This positioning allows Netflix to create a global market made out of microcommunities with niche tastes. In traditional economy, brands did the opposite. In order to scale, they had to appeal to as many people as possible, and cater to as mass, generic taste as possible. Now every brand can grow (and increasingly does) through taste clusters and niches, without reducing the product differentiation that attracts them all.

There are many doors in. Netflix personalizes more than movie recommendations—it also personalizes promo images of its movies and TV shows. Personalizing of not only what is recommended, but also how, can be applied to packaging, direct mail, newsletters, email communication or paid social. The more diverse the creative execution is, the wider the potential audience interested in different aspects of the product, service or brand—and the likelier it is that someone will like it.

Target communities, not individuals. Every Netflix user belongs to three or four taste communities. Members of modern societies belong to many more. No two persons are exactly alike, even those that buy the same products, choose the same brands and like the same content. There are those who enjoy foreign movies and travel documentaries, horror and romantic comedies, Vineyard Vines and Everlane, sneakers and high jewelry. People are the communities to which they belong.