Segments vs Audiences: The Optimal Approach to Strategic Growth

A byproduct of the ongoing convergence between marketing, insights, technology, and analytics has been new ways of doing similar things. Case in point, there is growing buzz around digital audiences being a substitute for market segmentation. Historically, market researchers used ‘segments’ while agencies used ‘audiences.’ Are they just different terms for the same consumer groups? If not, what’s the difference, and which approach should you be using?

SEGMENTS VS AUDIENCES

What are audiences, and what are they good at?

Audiences are groups of people across the population who exhibit, or are likely to exhibit, the same traits or behaviors. The digital ecosystem has facilitated an exponential growth in data as a direct result of tracking pixels, mobile device reliance, and a technology-driven society making virtually all we do observable by technology. With more data, marketing technology platforms have been able to expand the number and granularity of audiences available for targeting.

When our behaviors are enhanced with our offline data, like purchase data, companies can leverage analytics to find patterns in the data that paint a comprehensive consumer view. If data is not known about an individual, analytics can help predict that individual’s behaviors or interests based on how similar the person looks to others for whom the behavior or interest is known. Marketers use audiences to maximize marketing ROI through ad targeting, with the goal of showing ads to people who are most likely to be receptive to the ad, and therefore convert.

Strength. Audiences are largely defined by behavioral or demographic traits, so they are best for explaining what, where, and how consumers behave. For example, people who frequently shop at Target. Each audience is targetable and spans the entire digital population.

Weakness. The absence of the “why” dimension limits the strategic power of audience-based segments. Since we do not have the exact same data on all consumers, look-alike models predict characteristics (behaviors, interests, etc.) and impute scores to create audiences based on how closely a given consumer’s known behaviors, interests, and demographics resemble other consumers. For example, if we know that Person A shares 10 traits with a group of individuals who are college football fans, we can predict that Person A is also a college football fan.

What is market segmentation, and what is it good at?

Market segmentation has been around for over fifty years. It involves dividing a market into distinct sub-groups of consumers (“segments”) based on one or more characteristics. Segments generally fall into three categories: demographic (e.g., age, gender), behavioral (e.g., visitation, purchase occasion, etc.) and attitudinal (e.g., attitudes, needs). Through surveys, researchers measure the characteristics and then apply statistical analysis and strategic judgement to identify what groupings are most actionable.

Grouping consumers into segments defined by their shared attitudes and needs tends to provide the greatest insight and competitive advantage. Organizations typically leverage such classification schemes for several years to inform strategy, craft their brand positioning, evaluate tactics, and stimulate innovation. For instance, alcohol consumptions comes in many flavors – from a health-conscious segment to a premium-paying craft-beer segment. Brands have been built to serve these segments, using segmentation to prioritize the market, position their product, gather feedback on new offerings, inform acquisition strategies (like hard seltzer), and craft engaging ads.

Strength. By leveraging survey-based data, segmentations begin with the person to understand the attitudes, emotions, and behaviors that make up the complex relationship consumers have with brands. This approach puts the “person” in personalization. By knowing why people do what they do, you gain insight into how to influence them to buy your brand, or buy it more often.

Weakness. Although you can estimate each segment’s size, you only know person-level segment membership for those who have taken the survey, whereas marketers need to buy media against the full population. Without data linkages, there’s no inherent ability to predict segment membership for those who did not take the survey. And, since attitude/needs segments rarely have strong demographic skews, precision in ad targeting has been limited.

Can audiences replace segments?

Some agencies have begun using audiences as inputs to cluster analyses (the same technique commonly used in market segmentation) with the intent of forming new groups that look and sound like market segments without doing the survey. Although this approach is often faster than traditional survey-based methods, it is more prone to error.

Strength. You don’t need to spend the time or money to collect survey data. And the new audience is directly targetable with your advertising.

Weakness. You probably aren’t getting what you think you’re getting. You are modeling with already modeled data (assumptions rather than people), layering errors on top of errors, so the people you are targeting may not actually display the type and amount of differences that you think they do. Nor can you conduct primary qualitative or quantitative research with members of these digitally designed audience (i.e., there’s no Golden Questions or Typing Tool for recruiting focus groups or classifying concept test respondents).

Finally, as privacy policies become more restrictive, the breadth and depth of data available for modeling will become increasingly limited. Given that audiences rely wholly on data availability, we have seen a negative impact on the quality and precision of digital audiences. In some cases, we’ve seen audiences disappear entirely, disrupting business strategies.

Can I leverage the power of both Audiences and Segments?

The digital ecosystem allows us to leverage the strength of both approaches, serving as complements to one another.

The process is straight forward, as we see it:

  1. First, we start with people. A 1st party, deterministic evaluation of consumer needs, attitudes, and motivations, captured with a survey-based segmentation.
  2. Next, we identify and prioritize the segments that present the greatest value to your business so you can tailor your campaigns to speak directly to the personalities of each segment.
  3. Then, we use look-alike modeling to build custom audiences for targeting. The foundation of these audiences is people who explicitly told us their mindset and activities, rather than imputed assumptions around needs or behaviors.
  4. Finally, we push these custom audiences across on/offline marketing channels, empowering marketers to target each segment with personalized campaigns that communicate directly with each segment.

Business strategies and marketing tactics need to be seamlessly interconnected so you can win with consumers. By using the best that each discipline has to offer, you can better deliver on your personalization strategies by putting the person at the center of your plans and the targeted ad at the center of their feed.

Contact us to learn more