Age: We record the age of our viewers, and segment them into the following categories as standard: under 30s, 30-49 years old and over 50s. These segments enable us to compare performance across age groups and are customizable.
Gender: We gain insights by comparing performance across gender. Previous studies identified that the women elicit higher emotions than the men, for instance.
Duration: We also segment videos based on their duration, as we find this provides a more accurate basis for comparison: less than 20 secs, between 20-40 sec and longer than 40 secs. We’ve previously observed that longer videos tend to invoke higher emotions amongst their viewers.
Country: We have data from 187 countries, enabling us to observe cultural differences between countries. Out of those 187, we have 66 countries with statistically significant norms. Monitoring such data we can see for example that Americans generally show higher levels of emotions than other countries, closely followed by viewers in the UK.
Continent: We have records from all 7 continents in our database. This dimension can also be used to evaluate cultural differences.
Industry vertical: Currently our database contains 19 different industry verticals, including (but not limited to) Entertainment, Financial, Automotive, Food and Beverages and Personal Care. This enables clients to compare videos within or across different industries. Perhaps unsurprisingly, we’ve previously observed videos filed under Entertainment elicit the highest levels of emotion.
Brand: Our database contains over 1600 different brands at the moment. These can also be used to compare and contrast videos across brand-specific norms.
Test location: We differentiate between data collected online or in central testing locations, and can create norms according to the preferred method.
Tags: All videos in our database are tagged with relevant keywords. Currently we have over 400 unique tags in our database, all of which can be used to create custom norms. We have data on over 500 videos that were nominated or won a Cannes Lions award for instance, which we consider an indicator of creative success and which we can use as a benchmark to compare other videos against. Another 200 videos are tagged under ‘Super Bowl’, providing norms for examples of high-budget, high-level advertising.
Scene tags: All our videos are annotated with ‘scene tags’ based on an automated object recognition technique which identifies elements of each scene, second by second. We’ve collected over 4500 distinct scene tags, and over 4.5 million overall, all of which can also be used to create and compare benchmarks.
Social media statistics: We also record social media statistics for each video. It is a highly valuable dimension that can be considered as an indicator of success, and that we use to predict social media performance based on the emotional score of each video. Some of the major metrics in this category include number of YouTube likes, YouTube Views, Facebook likes. Facebook shares, Facebook comments, etc. We also collect information from other social media like Twitter or Pinterest amongst others. Our database has between 3000-4000 videos for each metric with over 6 million social data points in total.
Survey questions: We have a large database of survey questions we’ve used to establish metrics such as brand awareness, brand likability, purchase intent, ad likability, sharing intention, brand affinity, and so on. These can also be used to create custom norms.
Combining Dimensions: Norms can be calculated in different combinations of dimensions. These are some of the most common: a) Combination of Duration, Country, Age and Gender b) Combination of Duration, Country, Age, Gender and tags c) Combination of Duration, Country, Age, Gender and industry d) Combination of Duration, Country, Age, Gender, industry and Tags etc. Learn more about combining dimensions here.