Just a short time ago we were told to stay home, take staycations and stay safe. As the recent pandemic raged on, we kind of got used to the idea and many of us signed up to various viewing platforms to widen our viewing choice and drag us away from making endless loaves of banana bread.
Netflix alone added 15.77 million subscribers in the first three months of 2020 and 10.09 million between April and June, giving the company its biggest growth spurt in history as viewers turned to paid streaming services while stuck at home. And there’s just so much choice these days when it comes down to choosing what to watch.
But is there a science behind what we choose to sit down and watch, and why? Step up artificial intelligence (AI) and a new analytical report that uncovers how we make the decisions on what to watch and who influences us — it’s not who you think — and the power of the micro community on our media choices.
“The whole world is watching America, and America is watching TV,” said Sam Levenson.
Social Media and Networking
We almost all use social media to some extent, and we log on to talk about everything from breaking stories, entertainment, sports and politics to the latest on celebrity crushes. But we also use it to find information and to ask questions.
Twitter is a good example of social networking, where a one-line question will usually result in an instant answer and because everyone likes to put their two cents in. So how do we use social networking to help us decide what to watch next? We ask for recommendations.
Listening to Tweets
The data-driven media and marketing company, Horizon, has released a report that used Twitter and a selection of tweets from qualitative and quantitative research. The report input training data from that research into Generative Pretrained Transformer 3 (GPT-3) a state-of-the-art language processing AI model developed by OpenAI. Unlike algorithms, Generative AI works categorically, and can formulate the deep logical structures in natural language and summarize them for us.
How GPT-3 was Used
Using this AI technology, the data from 10,000 tweets was then analyzed in the report’s first wave, to discover how recommendations on what to watch spread amongst us in a social networking community and what are the key attributes of the shows that are recommended — all discovered by the use of the simple question, “What should I watch?” — and other versions of that in order to understand the results.
The second wave went a little deeper and answered the question: What is the effect of influence of viewing and conversational spread? The data was used from another 10,000 tweets to discover the themes discussed around a particular show and the social cohorts.
The Results of the GPT-3 Report
The results threw up three interesting observations:
- The Iceberg Effect: Recommendations are driven by social contagion, and how re-mentions drive the viewership inside small social networks. Platform-specific virality is measured by the number of times someone goes on to mention the same show.
- Everyday Influencers: When we hear influencers mentioned, we imagine celebs, reality TV stars and those that get paid small fortunes for one well-aimed insta-post. But no. The surprise here was the everyday influencers (EI) who are generally described as younger people deeply engaged in culture…that might be sports or the arts in general. These EIs tend to be left-leaning, with Twitter bio titles such as writer, author, pop culture enthusiast, film nerd, tech geek or TV watcher.
- Show Engrams: The really interesting find. What are show engrams? We know an engram is a group of cells in the brain that forms its own network with stimulation and allows us to form memories. Well, show engrams work the same. The relationships formed between viewers, shows and social media creates its own set of networks — so these are show engrams. So, if a show’s engram becomes input for understanding — in this case — audience and affinities, then this opens the gate for marketing to step in and use this information for further activation ideas. AI used to perform analysis and deliver insights into what delivers social viewership is one thing, but the same format works for any business.
Business Needs Insights
The possibilities are endless. Companies and CIOs are able to build future and insightful businesses around these AI tools to improve and streamline processes, and really know current customers and their habits. Tools like GPT-3 are here to enhance customer experience — and retention — and generate new ideas and solutions.
AI technology takes the guesswork out of marketing and product development. Every decision, from your marketing tools to how you market, which channels, how you advertise, assess your KPIs; all of it is now based on data. And the value of these information assets has never been greater.
