“It’s baffling to me why existing analytics tools are so backward,” says Antoine Amann, founder and CEO of Echobox. “They are mainly descriptive, typically showing you tons of fancy graphs and charts, but they don’t actually analyse the data for you.”
His startup, an analytics tool for news publishers, turns this approach on its head. Instead of simply displaying an abundance of analytics, Echobox looks at a site’s visitor data and, using its own algorithms and machine learning, actually recommends actions to take, including editorial, to drive additional traffic. Crucially, those recommendations are written in plain English.
Although Echobox is already being used by a major new publisher, it’s managed to remain in stealth mode until now and is currently running in private beta. However, TechCrunch has been given an exclusive demo of an early version of the product — a demo that, as somebody who writes the news, left me a little overawed.
The first thing you’ll notice about Echobox is the lack of complex charts or rows of data. Instead, publishers are presented with a list of actionable recommendations and insights, generated in real-time, based on their site’s visitor data. For example, it might recommend that a specific article be shortened in length.
Shorten the length of “Illuminators” to 220 words. 75% of your readers left your blog whilst reading “Illuminators”. Usually, merely 62% leave when reading a post. Shorten the length of Illuminators from 355 words to 220 words, which is the length of your most successful posts. Potentially, split “Illuminators’ into two separate posts.
Or Echobox might let you know that a particular article is proving popular amongst American visitors.
On average, individual posts attract around 291 visitors from the United States. But “Acne Studios In Downtown LA” had 781 American visitors, which is 168% higher than normal.
Or than an image included in an article is being shared more than usual on Pinterest.
This image from “Avedon Women” was pinned significantly on Pinterest. Your average image from your blog might be pinned 1-2 times, but this image from “Avedon Women” was pinned 32 times. This is 1,500% more than usual.
Or how to improve Facebook engagement.
“The Holiday Issue” only received 420 visitors from Facebook, whereas your average post gets around 729 Facebook visitors. Most of your Facebook visitors access your blog on Thursdays around 1.30pm. Share “The Holiday Issue” around that time on Facebook to acquire new Facebook visitors.
The result is similar to having your own in-house data-scientist, providing insights from a huge pool of data that can otherwise be like “looking for a needle in a haystack”.
“Interpreting data correctly is very hard,” says Amann, who has a background in quantitative research and statistics, and has interned at the Financial Times. “It requires time and professional statistical expertise.” However, data scientists tend to think in very technical terms and often have difficulty explaining their findings to non-technical people.
Source: TechCrunch
