Your new post is loading...
It’s hard to extract meaning from raw data only, so we enrich our data by adding story metadata such as authorship, publication sources, entities, topic classification; social data including influencers on a story’s distribution, audience size, social velocity, and performance against peer stories. With this information, you can unpack each story and see how it is spreading. You can quickly spot events of significance in any topic, and analyze by format to find what works.
Meanwhile, our algorithms can already predict the eventual reach of newly published stories, and extract trending entities (recently, this power was used to pounce on and debunk “disinformation” spreading in the run up to the French Presidential election). Our metrics can be used to spot “white space” opportunities with an audience, by analyzing which angles on stories are over-performing with an audience but are underreported.
We also plan to use algorithms to analyze whether a story is likely to be true, false, or biased based on source and other characteristics, and are working on making our recommendations much more powerful, such as: “Here’s what your audience will like today.”...
Publishing is entering a new era, where success will be determined by how scientific you can be in your approach. A new science of content is beginning.