How to Measure Audience Mood
Also, when and why does mood drive audience choice? And why does your audience want to escape the holidays?
Marketers and studio people never stop talking about their audiences’ moods. Most of these practitioners have never heard a scientific definition of mood, and so it’s frequent for them to ask confusing, unanswerable questions about moods to analysts. Until recently, there hasn’t been a coherent guiding framework for talking about moods in ways that might be insightful for these practitioners. I’ll define moods as a fuzzy mental measure of one’s bodily status in relation to its needs, and show how mood can be measured. Then I’ll present a framework from Dr. Annabell Halfmann’s lab in Mannheim that’s useful for thinking about when and how mood is relevant using the concept of psychological coping.
What is a mood?
Everyone knows what a mood is, but the term has a “many-splendored” definition for most people. “This film has a Wes Anderson mood.” “I am in an angry mood.” “This song has a dark mood.“ “Are you in the mood for ice cream?” are all using the term to mean something different. We will need to be more precise to start measuring things. Moods are not the same as genre or tropes. Moods tend to last longer than emotions; from minutes to hours, or even weeks. It’s harder to identify sources of moods than it is for emotions. Moods have fewer ingredients than emotions. And mood does not always drive preferences.1 Let me explain.
Moods are fuzzy predictions our brains make about the state of our body and its needs. At any moment, your audience’s nervous system is tracking oxygen, hormones, salts, hydration, glucose, and the four Fs (fighting, fleeing, feeding, and mating) within their bodies. The mind tries to summarize everything that’s going on into a more basic feeling we call mood. Before proceeding, let’s take a look at a metric for mood in the figure below.
Figure 1. Highly prolific scholar and close friend, Nick Bowman, adapted the “affect grid” as a metric for mood in the context of entertainment. Note the emotion words in the corners are for explanation only - and will be out of scope today. The real point here is that there are only two dimensions of mood: Arousal (vertical axis) and valence (horizontal axis).
Under this definition of mood in Figure 1, we have two dimensions: (a) how awake, alert, or otherwise engaged with the environment the audience is (arousal), and (b) how positive versus negative the audience feels (valence). So you can capture mood here with only two numbers. One for arousal and the other for valence. Take a look at Figure 1. Where is your mood on this grid right now?
If you’re a storyteller, you likely already have an intuition that you should be leading your audience to places all over this grid during your story. In this way, screens can manipulate moods. Entertainment can intervene to change a person’s mood, and audiences select media to improve their mood.
In terms of story metrics, you can capture the mood of content as well as the mood of audiences. Just a few years ago, capturing it in content was not possible with any kind of reliability unless you conducted a large-scale survey with audiovisual executions. But crowd-truth methods paired with artificial intelligence like LLMs, multi-modal models, and retrieval-augmented models2 can scale up the measurement of the mood-altering potential of content even just using preproduction scripts. (Though no systematic, open-science study has been published to date that I know of.) This can be incredibly useful for finding comps, or for plotting the mood in various places of the narrative arc for development insights.
But what exactly should you measure in the content?
Mood & Entertainment - Frameworks for Thinking about Mood
Quantitative entertainment researchers use mood-management theory to think about moods. I will intro the theory here and then bring in the notion of psychological coping to help guide strategic thinking about the audience's mood.
The theory essentially proposes four hypotheses: (a) Audience members feeling bad will seek stories that are more positive to enhance their mood. (b) Those who are stressed will choose calming narratives whereas those who are bored will choose more thrilling content. (c) Audiences aiming to change their mood will desire highly absorbing stories that can effectively distract them from their negative mood. (d) Finally, individuals looking to shift their mood will gravitate to stories that don't thematically echo the reasons they believe are causing their current mood (e.g., The Office might regrettably remind a viewer of their toxic coworker, to whom they attribute their stress).3
The hypotheses above clarify what you should try to measure in the content itself. Whether it’s existing audiovisual content or preproduction scripts, and whether you use audience panels, LLMs, or other AI techniques to estimate these factors, mood-management theory identifies (a) content valence, (b) arousal metrics like suspense levels, (c) absorption potential, and (d) thematics (semantic similarity) as relevant to impacting the audience’s mood. To develop the technology of story, quantitative researchers should map measured content features and their relation to these outcomes.
Mood-management theory’s underlying assumption is that mood drives all media choices, which is clearly wrong but does help to start the conversation about how mood impacts audiences. People often choose pain instead of mood improvement, and this extends to entertainment. There are dozens of scientific studies and experiments showing mixed support for those four hypotheses the theory purports.4 People of course enhance their moods with entertainment, but they also watch romantic breakup movies when they are going through a tough breakup themselves, and they watch movies about viral outbreaks when they’re in the middle of a pandemic lockdown. So how can we untangle these mood paradoxes? How can marketers think about how moods (and their associated metrics) are relevant to their title’s success?
Entertainment as Psychological Coping
If we take mood-management theory seriously, it would imply audiences want to escape negative moods all the time. But we know that’s not always true. People make choices they know will make them anxious, sad, scared, or otherwise “negatively valenced.” As a development or marketing person, you need a framework for thinking about when mood is relevant.
Artsy film people use the term “escapist” in a derogatory sense. But from a coping perspective, using entertainment for pure mood enhancement (i.e., mood escape) is an effective strategy for audiences to rest and rebuild mental resources to help them effectively approach their problem more directly later. That is, audiences engage in avoidance coping (escapism for mood enhancement) to build mental resources for subsequent approach coping (working through their problem more directly). In this sense, escapism isn’t merely hedonistic but instrumental. This theory piece on binge-watching as escapism outlines the data supporting this idea.
When thinking about it through the lens of coping, mood will impact audience choice more when the audience is stressed than when the audience is not stressed. And so marketers should (and already do) take the approach of predicting these stressful times and at least making attempts at catering to this need. Coincidentally, another friend and colleague, Allison Eden, just did a TV news appearance on her research where she brought up the point that watching nostalgic holiday movies is often an attempt at mood repair during stressful holiday activities and family events. Studio executives and researchers could focus on identifying more key aspects of their audience’s lives to better satisfy their approach and avoidance coping needs.
It’s also important to acknowledge that moods can change rapidly. For example, even in the most stressful of times, audiences will have long periods of positive mood. Of course, it’s a good strategic move to feature mood-enhancing content a bit more when the audience is stressed than in less stressful times, But it’s also strategically a bad idea to avoid producing content that could help satisfy the audience’s approach-focused coping needs even in what you may think are stressful times. This might consist of more negatively valenced or thought-provoking stories relevant to the audience. Account-planning style research on various audience segments can greatly aid such efforts.
Finally, highlighting or featuring content that trivializes a stress-inducing event, especially one in which sacred beliefs or values might be at play — e.g, a culture-wide tragedy like economic depression, political turmoil, mental health epidemic, or a school shooting — is a separate question from whether the audience is in the mood for that content. (Although the answer is usually redundant, it is not always the same answer.) In some cases, audiences will morally judge content-makers for being in bad taste. In some cases, the content’s timing is off due to the audience’s morality, not mood. It’s fair to ask whether perceptions of trivialization could be impacting audience reception in addition to mood. Audience panels are essential to this as AI cannot capture these rapidly fluctuating aspects of culture.
Conclusion
By understanding the complex interplay between culture, mood influences, and the psychological underpinnings of media choices, marketers and studio people can ask data scientists the right questions and develop a content strategy that truly resonates with their audience because it helps them cope. This approach acknowledges the nuances of mood measurement and audience psychology, paving the way for a more effective and impactful decision-making process for agents, studio executives, and other entertainment practitioners.
https://lisafeldmanbarrett.com/books/how-emotions-are-made/
https://research.ibm.com/blog/retrieval-augmented-generation-RAG
https://www.researchgate.net/profile/Richard-Huskey-2/publication/371530408_Computationally_modeling_mood_management_theory_a_drift-diffusion_model_of_people's_preferential_choice_for_valence_and_arousal_in_media/links/6488856079a72237652c3eaa/Computationally-modeling-mood-management-theory-a-drift-diffusion-model-of-peoples-preferential-choice-for-valence-and-arousal-in-media.pdf
https://www.tandfonline.com/doi/abs/10.1080/08838158409386511
https://www.amazon.com/Choice-Preference-Media-Routledge-Communication/dp/1138779350/ref=sr_1_1?crid=13SSUIIVXUV30&keywords=Choice+and+Preference+in+Media+Use+Advances+in+Selective+Exposure+Theory+and+Research&qid=1701312079&sprefix=choice+and+preference+in+media+use+advances+in+selective+exposure+theory+and+research%2Caps%2C105&sr=8-1&ufe=app_do%3Aamzn1.fos.18ed3cb5-28d5-4975-8bc7-93deae8f9840
https://onlinelibrary.wiley.com/doi/10.1002/9781118783764.wbieme0085