Percollate: Making Sense of Social Influence Scores [Review]

Social influence scores: Klout, Kred and the rest. The debate rages about whether they actually mean anything or if they’re pretty much just numbers plucked out of the air. But the real acid test for any piece of data, whether it’s a social media score or an analytics metric, isn’t whether or not it’s “good” or “bad”, but simply how useful it is. In other words, can you use that metric to do something useful for your campaign: take an action, make a decision, etc.

This is where new start-up Percollate steps in. Part social media monitoring tool, part meta-search for social scores, it lets you search for Twitter users by topic and then sort and filter them using either Twitter’s data itself, or metrics from one of Klout, Kred, or PeerIndex.

This writer’s first thought on hearing about the tool was “link building”. Being able to connect with people talking about specific topics as it happens, perhaps to ask for a link or seed a piece of viral content, is SEO gold dust. Likewise, product marketers can instantly find influential people talking about particular product categories without having to trawl through profile after profile, manually assessing each for its network reach. At least in theory. How does it work in practice?

Running a Search

Before you do anything else, you need to create an account (free for two weeks) and connect to your Twitter account. All easy enough. Percollate then gives you the opportunity to run your first search.


You can search for people following or followed by a certain user, or go for the topic search. Geographic targeting is also an option (and very useful), and you can choose to see data from Klout, Kred or PeerIndex along with your results (you can’t see data for more than one at a time due to API licensing issues, unfortunately; none of them want to see their results displayed along side those of their competitors).

And, hey presto, Percollate fetches the latest 100 people tweeting within your search criteria (you can add up to another 200 people to the results, but that’s the limit). This is where the real fun begins. The basic view is a grid of user avatars:


Not particular useful in itself, but you can zoom in to check out each person in more detail:


However, the real power comes in with Percollate’s sorting and filtering abilities. For example, you can filter out anyone below a certain Klout (or other) score, with less than a certain number of total tweets or an average number of tweets per day using a set of easy-to-use sliders. Filtering by subject, location and other criteria is also possible.

Clearly, this is where things start to get useful for the marketer, letting you find the most influential users currently tweeting about any topic with a few flicks of a slider. You can click from the tool directly to any user’s profile as well.


Garbage In, Garbage Out

Of course, using any tool requires a bit of practice and experience, and Percollate is no different. English is a very imprecise language, so entering the right kind of search terms in the first place is all important. Many words overlap in meaning or are simply too vague, but others may be so specific that they return few results.

Bottom line: no result set that the tool generates will be 100 percent relevant, but multiple searches and tweaking of the filters will usually dig out decent results (assuming people talk about your area of interest on Twitter, but that’s another matter).

Power users will also be pleased to hear that Percollate has the all-important CSV export feature, allowing you to export and combine the results of multiple searches.

Final Thoughts

For any product manager, link builder, or marketer who needs to engage with key Twitter influencers (or users at any level), Percollate offers value. Being a start-up, there are some rough edges in the interface, but these will surely be smoothed over in time.

Related reading

Seven content marketing strategies you should try [With examples]
Hootsuite social trends report for 2020
Improving personalization with machine learning
Capitalizing on paid social in B2B industries