Influencermoth is a tool that helps you identify social influencers on Twitter. With influencermoth you’ll be able to quickly scan Twitter’s social media platform for influential users based on your keyword of choice. Influencermoth was free to use, but is currently offline (because I’ve stopped using Twitter and currently have no API credentials).

Technical specifications
Influencermoth is limited to 12 searches per 15 minutes due to Twitter API restrictions. It uses a maximum of 2,000 tweets for it’s analysis and requires a keyword that is often tweeted about in the past 9 days, as Twitter’s Search API only includes tweets from the past 6-9 days

5 months of influencer marketing research

Influencermoth was the result of my 5-month long graduation research project for We Are Blossom (A Dutch Social Content Agency) back in 2012. I researched influencer marketing and publicly available data sources. This resulted in Influencermoth: a tool to quickly scan Twitter for social influencers.

Different types of influencers

Influencermoth shows you who the main influencers are on Twitter regarding a given topic. It shows reach on one side and activity on the other. The separation between reach and activity is based on the influencer pyramid by Forrester:

The influencer pyramid. Source: Forrester.

This pyramid suggests that there are different types of influencers:

  • The social broadcaster (the top) is an influencer with many followers. They are valuable when you want to spread a message to make people familiar with a new brand name. They have a large audience.
  • The mass influencers (the one in the middle of the pyramid) have fewer followers but do have a more active conversation with these followers. Therefore, they are able to make your brand more trustworthy.

Finding the social broadcasters and mass influencers

The tool loads the data for its calculations directly from Twitter via the Twitter Search API. It works in real-time.

First, it checks the number of tweets a users posts on a topic. The more they post, the more they matter. Tthe model reduces the value of a tweet if a user keeps using the key word (e.g. 2 tweets are counted as 2, but 10 are counted as 6).

The data calculations simplified. Source: my research.

After that, it runs two types of calculations:

  • Social broadcaster: to calculate the reach of an influencer on twitter, the tool takes the number of followers into account. This data set contains the social broadcasters.
  • Mass influencers: to get the mass influencers, the ones with a lot of interaction, the tool ignores the followers but looks at the number of mentions and retweets a user gets. This data set shows you the most active and engaged users.

The interactive data visualisation

The visualisation of the data is handled with the open source D3.js library. The main purpose is to clearly separate the reach of social broadcasters and activity of mass influencers within the tool.

To do this, both lists are clearly separated, I designed a mirrored chart setup. Each type of influencer has its own bar chart and colour.

Influencermoth screenshot.

The tool also has doughnut charts below each bar chart that allow you to compare selections. This is helpful if you want to compare the reach and activity balance of selections of users that show up on both sides of the tool.

For example, if your main focus is reaching a large audience, you might want to compare two selections of influencers with an equal amount of reach, and make your selection based on their activity scores.

On the top of the tool, there’s room for a top three list. Influencermoth only shows you these if there are users that appear on both sides of the graph. These users have both reach and are active in the community. Keep in mind that because of the clear split in reach and activity, this doesn’t have to be the case.