Parrot Analytics’ comprehensive audience attention measurement system captures the world’s largest audience behavior dataset. We extract the signals from the noise to deliver the first globally standardized measurement of audience demand for all markets, in all languages and across all platforms and devices.
Our technology quantifies global consumer demand for TV, movies, and talent. Billions of people interact with content and talent each month and Parrot Analytics systems collect and analyze the activities, interactions and behaviors from over 2 Billion people every day, which includes exclusive first-party consumption datasets from over 350 million households globally as well as hundreds of millions of households' search, posts, reading and social interactions activity.
Our data sources include search engines, wikis and informational sites, fan and critic rating sites, social video sites, blogs and micro-blogging sites, social media platforms, peer-to-peer apps and open streaming platforms.
By measuring demand for TV, movies, and talent we are quantifying the attention economy, our system captures how much attention, engagement, desire and viewership is expressed by consumers for content and talent, all around the world.
We group the demand sources into four distinct buckets that represent different audience behaviors:
Social media: Social media activity is collected from publicly accessible social networks, photo sharing platforms and micro-blogging sites, including but not limited to Twitter and Tumblr.
Social video platforms: This bucket represents audience activity on publicly accessible video consumption and engagement platforms, including but not limited to Youtube and Daily Motion.
Research: The Research bucket represents audience activity collected from informational and fan or rating websites, including but not limited to Wikipedia and TV Maze.
Free streaming/downloading: Free streaming and downloading data refers to full-episode consumption from peer-to-peer (P2P) connected boxes, file-sharing protocols such as BitTorrent and P2P streaming applications such as Popcorn Time.
We evaluate and refine our data sources and algorithms on an ongoing basis to ensure our demand metrics are robust.
Once captured, those inputs are then weighted by time and effort, so streaming or downloading a show is a higher expression of demand than a "like" or comment related to that show.
Capturing and combining the empirical expressions of demand from the various data sources allows us to create distinct demand metrics.
Before the creation of a normalized metric, it was not possible to benchmark audience demand for content airing on different platforms on a market-by-market basis.
Benchmarking demand with the Demand Multiplier
The demand multiplier is the unit of demand for all entities, e.g. TV shows, movies and talent. The demand multiplier allows us to benchmark any piece of content, talent or portfolio against the market average.
If we rank all the titles within a market by their Demand and plot the result, a familiar long-tail distribution appears, which is a well-documented pattern observed in empirical data collection and analysis.
We call this the Demand Distribution Curve.
The Demand Distribution Curve illustrates how a TV show, movie title, and talent popularity compares to the Demand Benchmark and is divided into performance buckets, ranging from “Below Average” to “Exceptional". A show falls into one of these buckets depending on how many times more or less demand it has compared to the Demand Benchmark.
A show’s performance is market-specific, so the same show can be in the “Average” range in the United States and in the “Good” range in France.
Only 0.2% of all titles will have exceptionally high demand, whereas 64.1% have average demand and 24.4% have below average demand.