You might notice in our upcoming release that Mattr will only show you the over-indexed interests of your influencers. This is a big, exciting change for our app. Read on to learn why over-indexed interests are more valuable to you, and how we discover them.
Why is over-indexing useful?
Over-indexing helps identify those unique or special interests that your specific group of influencers are more inclined to gravitate towards. Identifying these interests allows for more precise targeting to that group, resulting in more bang for your buck. For example, CNN may be a top interest for a particular group of people based on a strict count, but there might be a smaller, more niche media outlet that’s much more popular amongst your group than across the population as a whole– and that means more eyes on your brand.
How do we do it?
Mattr first creates a large set of the most followed interests within a report’s influencer pool, several of which might be the most followed interests for the overall population. We then analyze the probability that an interest is followed by the report pool of influencers, and compare that to all users on Twitter. That number is then used to sort and identify the most over-indexed interests for the report group, with the top 25 interests highlighted in the report.
Are there limitations?
It’s important to note that if your report contains a single influencer, or if the influencer pool is very small, results won’t be as useful because they’ll most likely reflect the least popular interests. Another thing to keep in mind is that every interest is treated the same and carries the same weight since there are no degrees of following. We may address this in the future by incorporating some data about interests that are also influencers in the report group.