DataSift Offers Several tools to Extract Value From Social Data

DataSift offers several tools to extract value from social data. The data that consumers leave on all different social networks is exploding and with DataSift it becomes possible to collect, filter and analyse this data.

DataSift offers real-time as well as historical insight to find trends on brands, business, news or financial markets. DataSift was started as TweetMeme, the highly popular Twitter news service. The founder of TweetMeme, Nick Halstead, quickly saw the impact of social media data on organizations so he created a more extensive platform to better capitalise all available data. Companies can start analysing the major social networks such as Twitter, Facebook, Youtube, Flickr but also websites as Wikipedia, Amazon or IMDb. Different filters can be created from a simple keyword filter to a more complex one using text-pattern matching and geo-location data.

Pricing varies for each data source and DataSift becomes more expensive the more complex the filter or the larger the data analysed. Every single item of the data aggregated is analysed, giving users added augmentations to better understand the results from an analysis. These range from sentiment analysis to extracting topics being mentioned. Searches can also include filters like Klout score, gender, sentiment, language etc. It may be clear that DataSift burns a lot of data. At the moment, DataSift deals with millions of data interactions a day, powers over 20.000 data streams out to customers and it processes 2-3 TB of data daily. Social networks are constantly growing and its users are sharing more and more. All this data can be very relevant for companies. For example, a mobile phone manufacturer can track how a new phone is perceived and can see the differences in region and gender.

Using this valuable information the company can adjust the offering if necessary. At the moment the historical data available from Twitter only dates back to January 2010, hopefully soon DataSift will be able to offer data dating back from before 2010 as it will then be possible to see long-term trends in the big data. At the moment Datasift has customers from all over the world, however it seems that they have not patents (pending) and have not yet won any awards. They do have a new partnership with Bit.ly that will allow their customers to measure actual engagements of their clicks. Data-startups therefore rates DataSift a 6 bits rating.