imgZine is on a mission to help publishers and organizations create engaging magazines for mobile. But crafting these beautiful zines is only half the story. Our long-term goal is also to help content owners unlock the value of their content by providing them with robust insights into their audience’s reading behavior. In this post, we’ll introduce you to some of the skillful minds that are working to make this happen.
Say hello to our data specialists
Our data team has been growing steadily since last year. In June 2012 we hired Pedro, who has a master’s degree in artificial intelligence and a keen interest in machine learning techniques. This February, we welcomed Sarunas, a true math whizz with a degree in econometrics, and a few weeks back we hired Bernard, who is just about to finish a master’s in business information technology.
Some things we’ve been working on
Data-driven learning is paramount to the success of any product. Therefore, our data team is constantly striving to refine our metrics and help customers understand what their audience truly wants.
One of our latest breakthroughs is the Sarunas triangle. The triangle compares the time users spend reading an article with the length of the article, allowing us to infer how engaging the article was.
So far, insights drawn from this metric indicate that articles of a certain length are more engaging than others, suggesting that publishers could use it to identify what the ideal length of their articles should be. However, our team is also working on making this metric more objective. We hope to achieve this by excluding anomalies and comparing article length to syntax complexity and other features that may have an effect on reading times.
The flipside of Sarunas’ triangle is Pedro’s recommender engine. The recommender engine identifies a user’s favorite topics based on a model of their behavior in order to suggest similar topics. That way, readers are more easily exposed to the subjects that interest them most, and publishers get to learn about the most popular articles and themes.
A select number of our apps already incorporates the recommender engine and version two is already being tested. Besides taking user interests into account, new releases of the recommender engine will draw on complexity and reading context to match users with the ideal content.
From key metrics to data visualization
Smart content filtering algorithms and metrics like Sarunas’ triangle have been absolutely necessary to provide our customers with added value. But we also need to keep our data intelligible.
That’s why we’re currently designing new visualization techniques, aiming to produce graphics that customers can understand and apply for actionable business intelligence. Thanks to our data team’s efforts, an experimental version of our revamped analytics dashboard is almost ready. With it, we hope to deliver richer, friendlier and more ambitious metrics to help our clients understand their readership. Stay tuned to get a sneak peek!
