How to Analyze Creative Content

There is a common misconception about Content Analytics. It is indeed the application of data science to content data to draw evidenced insights for multiple use cases. It is imperative to hone in on the right signals and connect with your audience amid strong competition.

But, it is about supporting the art, not directing it.

Done right, it is about giving creators the freedom to create and not hindering them. Art and science can have a successful marriage.

There are many definitions of ‘content’. Here, content is broadly defined to include long form to short form — movies, series, trailers, digital or non-digital advertising, sizzle reels, or other marketing material. The analytical approaches described can apply to any and all of these formats. That said, it can take case-by-case modification to apply them.

Artists such as script writers and producers create content by drawing on experience, intuition and of course immense creativity. They want their creation to become successful. When it is, experience and intuition can also help us develop hypotheses about what drove the success of a TV show, a movie, an advertisement or any other media content. And potentially to what extent. Was it sealed by the twist at the end of the tale that left an indelible mark? Did the advertisement hit a nerve because it echoed the cultural zeitgeist? Was it relatable only to a niche audience? In that case, how can we expand the relatability factor?

The examination also applies to content that did not fare well. It is almost always an amalgamation of drivers rather than any one feature. And it can be a worthy effort to look deeper into it to make better content decisions in the future.

Hypotheses are a great starting point to understand how to try to repeat success. But, without actually testing the hypotheses with data and analysis, we have no evidence. That is why it is important to apply Content Analytics to test hypotheses and identify best practices.

It takes two to test hypotheses and bring out actionable insights, in this order:

The first step is to define success based on the outcome you might want to influence. For example, the number or increase in impressions or conversion in advertising, ratings thresholds for linear TV, minutes or hours viewed for streaming and so on are some examples of key performance indicators that might signify success. There are many business intelligence platforms available in the market that allow you to analyze these metrics.

While these might suffice, it is also important to include some longer term ones such as repeat views for movies or successful reruns of TV shows (e.g. Friends, Seinfeld) that are indicative of longevity. Those can be more useful than short term success metrics in identifying ‘content hooks’ and revealing deeper insights for developing content that can reap returns for a longer time. Shorter term metrics can be muddied by zeitgeist effects and initial marketing spend, neither of which are usually sustainable.

The second piece is breaking apart the concept of content and measuring it as illustrated in the visual above. Identifying different measurable attributes and features of content would vary based on the hypotheses you are looking to test and your specific goal.

Let’s pick one such feature — action. At first glance, it might seem difficult to measure action in a meaningful way for data analysis. But, there are indeed multiple ways to measure it. For example, you can use different methods to quantitatively grade the level or intensity of action and evaluate its relationship with your success metric - for the audience as a whole or better yet, broken out by audience segments and demographics — if that data is available. Such analysis would shed light on how action matters in your content.

The final step then is to statistically examine the relationship between content features and success using an approach that is tailored to the question you are trying to answer. Clearly, this is a highly simplified summary of just one way to apply Content Analytics. There’s a lot more that its application and tools can achieve.

Related posts:

An essay about how many people make assumptions, don’t listen to what other people really say, and otherwise engage in debate in a sloppy manner.

Management is the backbone of any organization. A good management insurers the success of the organization. In this article you’ll learn that how and why management matter so much.

Restarts are required to affect in spring boot applications once a change is done. It will take time, if we follow traditional way of stopping and starting the application. Is is possible to reduce…