Chapter 5: Understanding Audiences

Learning Objectives

After reading this chapter, you should be able to:

1. Distinguish a creator’s actual audience from follower counts and from the imagined audience the creator addresses.
2. Differentiate demographic and psychographic descriptions of an audience and explain what each is good for.
3. Apply the three lenses of audience research to gather evidence about who an audience is and what it values, and assemble the findings into an audience profile.

Eighty-Nine Guests

In the weeks before her album 1989 came out in October 2014, Taylor Swift invited a total of eighty-nine fans, in small groups, into her homes in Los Angeles, Nashville, New York, and Rhode Island, and into her London hotel suite. The guests were contacted quietly through her fan club, signed nondisclosure agreements, and handed over their phones at the door. Then the biggest pop star in the world played them her unreleased album from her own phone, song by song, explaining what each one meant and who she had written it with. While they listened, the guests ate pumpkin chocolate chip cookies she had baked herself (Barnes, 2023).

The sessions became fan-community legend, but the part that belongs in a marketing textbook is how the guest list was made. There was no contest and no lottery. Swift chose the eighty-nine herself, and she described her method on television that same month. “I would go online and I would look at their Instagram pages, or their Twitter or their Tumblr or whatever and I just kind of watched them for months and months” (Stutz, 2014). The guests could tell. Interviewed years later, attendees kept returning to how well she knew them before they had said a word. She greeted strangers by name, referenced one guest’s Tumblr URL from the front of the room, and recognized another from what she had posted about her young son (Barnes, 2023).

Consider what happened here. An artist with access to every chart, dashboard, and market report money can buy spent months on the slowest research method available, reading what her audience members said when nobody was asking them anything. She did it because what she wanted to know doesn’t appear on any dashboard: who these people were and what her music was doing in their lives. Notice, too, whom she was studying. Not the tens of millions who had streamed a song once, but the inner circle who showed up for everything.

Even she ran into a limit. This method let her know eighty-nine people, a tiny sample compared to a fan base in the millions. That limit is where you hold an advantage she doesn’t. A creator with forty regular viewers can know their audience more completely than any superstar can, but only by looking. This chapter is about the looking: what an audience is and is not, how to tell the audience you have from the one you imagine, and how to turn free, patient attention into evidence you can build a venture on.

5.1 The Audience You Imagine and the Audience You Have

Unit 1 taught you the systems: what marketing is, how ventures work, and what the platform in the middle of the map wants. Unit 2 turns to the people, and it starts by defining the word this book has been using since page one:

Definition: Audience. A creator’s audience is the set of people who choose to spend attention on the creator’s work: watching, reading, listening, and coming back for more.

Each clause of that definition is important. “Choose” recalls Chapter 2’s exchange. Attention is paid, not taken, so an audience is not a possession but people deciding, over and over, to come back. This “coming back” is what separates an audience from traffic. Chapter 4 showed that recommendation feeds deliver your work to people who never asked for it and may never think of you again. A view count blends your audience with those passersby, just as a follower count includes people the feed no longer reaches at all (Patreon, 2025).

An audience is not one crowd, either. It comes in rings: occasional viewers on the outside, regulars closer in, and at the center the devoted few who comment, share, and buy. The rings want different things from you and will support you differently. That is why Swift’s guest list was a search for the innermost ring, and why a smart audience question is usually not “what does my audience want” but “which ring am I asking about.”

Before you can study the audience you have, though, you should meet the audience you have been carrying around in your head. Every time you make something, you make it for someone. The question is who you were picturing.

Definition: Imagined audience. The imagined audience is the mental picture of who is watching that a creator writes, films, and posts for, whether or not it matches the people on the other side of the screen.

The concept comes from research on how people write in public. Studying early Twitter users, Marwick and boyd (2011) observed that people posting to an invisible crowd write to an audience they imagine. The actual readers are unknowable at the moment of composition, so the imagined ones stand in. This is not a flaw to be cured. Nobody can compose for hundreds or thousands of strangers at once. What writers have done and creators continue to do is address an ideal audience member, an imaginary interested person who is often a mirror image of themselves (Marwick and boyd, 2011).

But an imagined audience drifts. It is built from hope and habit, and left unchecked it diverges from the people who are showing up. The gap between the two is where ventures fail, with the creator serving viewers who aren’t there while the viewers who are go unserved. (The same researchers named a second problem, context collapse, in which your classmates, your grandparents, and your future employer all flatten into one feed. Chapter 8 takes that one up.)

You can’t stop imagining your audience, and you shouldn’t try. What you can do is measure the gap between the audience you imagine and the audience you have, and close it. Measuring that gap is the next section’s subject.

5.2 Three Ways of Looking

Definition: Audience research. Audience research is the set of systematic ways a creator learns who the audience is and what it values, so that decisions rest on evidence instead of guesses.

If this sounds familiar, it should. Chapter 3 taught you customer discovery, which is treating “who is this for and what do they value” as assumptions and getting out to test them before building. Audience research is customer discovery continued after launch. It is the same questions, asked now about the audience you actually have, for as long as the venture continues. Audiences change as they grow, and recommendation feeds (recall Chapter 4) keep introducing your work to strangers you didn’t choose. Every belief about your audience is a claim, and claims about audiences get the treatment Chapter 2 taught you to give claims about strategy: state the evidence, and know what would change your mind. What audience research does not require is a budget. Students hear “research” and picture survey firms. A working creator’s core research instruments cost nothing, and there are three of them:

1. Observe: What do they do? Your platform’s analytics record what the audience does. They tell you what people watch to the end, where they leave, and what they return to, search for, and buy. Behavior is the most abundant evidence you have and the hardest for anyone to fake, your polite fans included.
2. Ask: What do they tell you? You ask by posing questions in your content, sending short surveys, and above all holding the discovery conversations you learned to run in Chapter 3. Asking is the only lens that can reach the why behind a behavior, and the only one that can reach people your content hasn’t found yet.
3. Listen: What do they say unprompted? You listen wherever audience members talk to each other: comment threads, community spaces, their own accounts. This was Swift’s method, and its power is candor. People speaking to each other reveal what they would never put in a survey answer.

The reason you need all three is that each one distorts a little on its own. Observation counts everyone the algorithm delivers, audience or not. As Chapter 4 showed, a spike in a chart may describe passersby who will never return. Asking hears the engaged and the polite. The fans who answer a survey are not the regulars who watch and never comment, and people asked directly tend to be kind. Listening favors the loudest. A comment section is a small society with its own incentives, and its most active voices may speak for themselves alone.

The working rule follows from Chapter 2’s evidence habit: a claim about your audience earns trust when at least two lenses agree, and a claim only one lens supports stays labeled a hypothesis. Swift was running this framework all along: a decade of the observe lens in her charts, months of the listen lens, and an evening of the ask lens in her own living room. It also scales down. A student with forty viewers can check one analytics chart, ask five regulars several questions each, and read every comment. That is not a toy version of audience research. It is a complete pass of the discovery cycle from Chapter 3, with the three lenses supplying the test step. In Chapter 3, the answer to the value proposition’s first question, who is this for, was a hypothesis. From now on it is evidence.

5.3 Who They Are, and Why They Come

The three lenses tell you how to look. This section is about what you are looking for, and it comes in two layers.

Definition: Demographics. Demographics are the observable, countable characteristics of an audience, such as age, location, language, and occupation.

Demographics answer “who, in census terms,” and they are the easiest audience facts to get. Platform dashboards report several of them automatically, which makes them the observe lens’s natural starting point. They also drive real decisions. Where your audience lives constrains when you should publish. The language you write in and the references you can assume are demographic questions. And platform choice is a demographic decision, because the platforms are different rooms with different people in them.

The differences are measurable. Among U.S. adults, 84% report ever using YouTube, and it is the only major platform above 60% in every age group. Instagram reaches 80% of adults under 30 but only 19% of those 65 and older. Snapchat falls from 58% of adults under 30 to 4% of those 65 and older. Facebook peaks not among the young but in the 30-to-49 group, at 80% (Pew Research Center, 2025). A creator serving retirees should think hard before building on Snapchat, and “everyone is on Facebook” turns out to be true of no age group. Note the source of these statistics. Pew is a nonpartisan research center that publishes its method, a survey of 5,022 U.S. adults with address-based sampling. That is a different class of evidence from the industry counts you learned to question in Chapter 2’s Reading the Numbers feature.

What demographics cannot do is explain why anyone shows up. Two twenty-year-olds in the same city can want opposite things from the same channel, so a purely demographic description leaves you with a census of strangers. The second layer is where the explanation lives.

Definition: Psychographics. Psychographics describe why an audience shows up: its values, interests, motivations, and tastes.

Psychographics are Chapter 2’s value question, asked as audience research. What the audience is getting from your work lives in this layer, not in the age table. They are also harder to get. No dashboard reports “watches for the calm competence” or “reads for permission to take her own writing seriously.” (Platforms do infer interests from behavior at scale, and their ad systems sell on those inferences, but an inferred category is not a reason. The why still has to be learned.)

Psychographic knowledge is learned by observing what people do, asking why they do it, and listening to what they say to each other. The discipline that keeps this layer useful is specificity. A psychographic claim worth writing down states what the audience values in terms concrete enough to test. “My regulars value the repair explanations more than the reviews” can be checked against watch patterns and comments. “They like good content” can never be wrong and is therefore unhelpful.

For an example, look back at a writer you already know. When George Saunders launched Story Club, he framed the newsletter as “simulating a version of the MFA through Substack” for readers who could never be among the six students a year admitted to his Syracuse program (Stone, 2021). That sentence is a psychographic profile in miniature. It describes an audience by what they value (i.e., real craft instruction and being taken seriously as writers) rather than by any census category. Furthermore, the newsletter’s weekly reader questions and comment exchanges keep an ask-and-listen channel open that a book author of an earlier generation never had (Substack, 2026).

5.4 The Audience Profile

Everything this chapter has taught converges on a deliverable, the first component of your Unit 2 milestone.

Definition: Audience profile. An audience profile is a short, evidence-based description of a creator’s actual or intended audience: who they are, what they value, and how the creator knows.

A good profile has three qualities:

1. It is a composite. The audience profile blends many real viewers into a portrait of the kind of person who keeps showing up, rather than describing any single fan or trying to cover everyone. It may take the rings from Figure 5.1 one at a time, since your regulars and your fans usually deserve separate sentences.
2. It is evidenced. Every claim carries its source in lens terms (e.g., an analytics observation, something audience members said when asked, something they said unprompted). Any claim resting on a single lens is marked as the hypothesis it is.
3. It is provisional. The audience profile ends with open questions you have not yet answered and the lens that would answer each. This keeps the document inside Chapter 3’s discovery cycle, not filed away as finished.

Industry practice often presents a fictionalized version of this document as a “persona,” complete with a stock photo and an invented name. The term is worth knowing, but apply the evidence rule to any persona you’re handed, because one invented without evidence is the imagined audience wearing a lab coat.

Your audience profile will join a niche statement in Chapter 6 and a positioning statement in Chapter 7, and at the unit review you will assemble the three into a brand definition. Much later, the profile becomes the seed of the audience section in the media kit you will show paying partners in Chapter 18.

For a portrait of what this chapter’s habits look like sustained over a career, meet the creator who will accompany you through Unit 2.

Summary

This chapter turned from systems to people. An audience is the set of people who choose to spend attention on your work and come back for more, which makes it smaller than a follower count and different from a view count. It also comes in rings, from occasional viewers to devoted fans. The audience you address in your head, the imagined audience, is a necessary tool that drifts from the real one unless measured against it (LO1). Audience research is customer discovery continued after launch, and its three lenses cost time, not money: observe what the audience does, ask what they will tell you, and listen to what they say unprompted. Each lens distorts alone, and a claim earns trust when two lenses agree (LO3). What the lenses look for comes in two layers: demographics, the countable outer layer that dashboards report and that drives decisions like platform choice, and psychographics, the explanatory inner layer where Chapter 2’s value question lives and where claims must be specific enough to test (LO2). The chapter’s deliverable, the audience profile, binds the findings into a composite, evidenced, provisional description of who the audience is and what it values, and it opens the Unit 2 milestone that Chapters 6 and 7 will complete (LO2, LO3). The standing lesson is Swift’s: the charts and market reports money can buy answer the shallow questions, and the deep ones are answered by patient attention, which any creator reading this book can afford.

Questions for Discussion and Application

1. Swift had access to every chart her label could buy, and spent months reading fan posts anyway. What did the reading tell her that the charts couldn’t, and what could the charts tell her that reading never would? What does her approach look like for a creator with forty regular viewers, using only free tools?
2. Write three specific claims about your venture’s intended audience: one demographic, one psychographic, and one about what they value most in the kind of work you plan to make. For each claim, identify the lens that could test it, what you would look at or ask, and what result would count as contradiction. Keep the claims. The profile you write in this chapter’s Semester Project section will test them.
3. Choose a creator you follow and, using only public materials (e.g., their comment sections, community posts, and what their regulars ask for), work out what you can about their audience through the listen lens alone. Which ring of their audience do you occupy, and what claim about their audience would you need a second lens to believe?

Semester Project

Write your audience profile in half a page. Include two demographic claims and two psychographic claims about your venture’s actual or intended audience, plus one statement of what that audience values most. Tag each claim with its evidence in lens terms. Close with two open questions and the lens that would answer each. If you have a live audience of any size, use at least two lenses on it. If you don’t yet, build the profile of a comparable creator’s visible audience from public materials only. No accounts, posting, or spending required. The profile is the first component of the Unit 2 milestone, and the niche statement and positioning statement of the next two chapters build on it.

References