DataSift is an example of a product business model. The company uses algorithms to sift through and harness the power of Human Data in real-time. Human Data covers the entire spectrum of human-generated information, including social networks, blogs, news sites and more. DataSift allows businesses to easily aggregate, filter and extract meaning from this data — gathering market insight, understanding customers more deeply, measuring results, optimizing campaigns and more – without compromising consumer trust.
In 2007, Nick Halstead developed an algorithm to categorize RSS feeds by the millions. He called it fav.or.it, and used it to rank the popularity of various web news feeds. When Twitter emerged as a dominant microblogging service, Halstead started tracking Twitter feeds to analyze tweets and retweets about particular topics, converting fav.or.it into TweetMeme. Halstead and his team realized that the technology behind TweetMeme would be useful outside of just Twitter and in 2010 DataSift was founded. Nick made the decision to build an engine for programmatically understanding what is being said within social data streams. This engine is now DataSift and the technology enables organizations to gain meaningful insight from the wealth of data sources now available to them. In December 2013, the company closed a Series C round of funding for $42M, and the company currently has more than 100 employees globally.
In March 2015, DataSift announced its partnership with Facebook, which enables brands and advertisers to gain insights from real-time aggregated and anonymous Facebook topic data, while protecting the identity of individuals. DataSift is currently the only company to offer Facebook topic data, which is anonymous and aggregated content data about specific activities, events, brand names, and other subjects that people are sharing and engaging around on the world’s largest social platform.
DataSift works with more than 20 social networks and its partner ecosystem includes Brandwatch, Sysomos, Crimson Hexagon, Meltwater and WPP as well as thousands of app developers across a variety of industries including retail, telcos, high tech consumer packaged goods and digital agencies.
Learn more about the Product Business Model
A dyadic transactional relationship where your good or service can be designed and delivered without prior interactions with the customer.
Engagement — Value Creation Proposition
Through DataSift’s partner ecosystem, which can be seen here http://www.datasift.com/partners/, marketers and agencies can take advantage of Facebook topic data. They will be able to gain a deeper understanding of the topics people are engaging in on the world’s largest social platform and use the insights to make better business decisions.
DataSift is the only company that provides state-of-the-art technology that allows companies to capture, analyze and act on all the types of human-generated data, without compromising consumer trust. DataSift’s platform unifies all data – real-time and historical – in one place, unlocking its meaning and delivering it for use anywhere in a business.
DataSift is also currently the only company to offer Facebook topic data. The partnership with Facebook leverages DataSift’s technology to aggregate, anonymize and deliver summary results from Facebook topic data, enabling developers, agencies and brands to build innovative apps that surface insights into what audiences are sharing on Facebook. This can help marketers – from researching the latest fashion trends to identifying the next big thing. All information is anonymised and aggregated, so people’s privacy is respected.
DataSift forms strategic partnerships with data providers, structuring and enriching the data, and making extracting insight from the data easy. The partnerships that DataSift makes mean that insight providers can easily access the data without having to go direct to each source, they also ensure that all data is properly licensed and that consumer’s privacy is protected. By structuring the data, DataSift allows insight providers to work with all data, regardless of source or type, using the same technology. This means that they can concentrate on providing insights for their customers.
Delivery — Value Chain
DataSift has developed a classification engine, VEDO, in-house, which allows its partners to differentiate their offerings and provide industry or function specific insights. DataSift uses CSDL language, which allows its partners to quickly and easily isolate the content and results that are of interest to their customers.
In terms of anonymised and aggregated Facebook topic data, the privacy‐first approach is enabled by PYLON ‐ DataSift’s new API. With PYLON for Facebook Topic Data, social data never leaves Facebook. It is processed within Facebook’s data centres. In addition, user identity is removed from social data before it is processed by DataSift and only summary results containing 100 or more individual’s data are delivered by DataSift. Social data is deleted after the 30‐day retention period ends and the minimum age applied for data collected for analysis.
Monetisation — Value Capture
Partners pay an undisclosed amount to access DataSift’s platform.
MarketWatch Article: http://www.marketwatch.com/story/datasift-recogniz…
Programmable Web Article: http://www.programmableweb.com/news/two-great-social-data-platforms-how-datasift-and-gnip-stack/brief/2014/02/10
Startups.co.uk Article: http://startups.co.uk/the-entrepreneur-nick-halste…
V3 Article: http://www.v3.co.uk/v3-uk/interview/2403401/facebo…
Tech World Article: http://www.techworld.com/blog/innovation-intellige…
Programmable Web Article: http://www.programmableweb.com/news/two-great-soci…
Recode Article: http://recode.net/2015/04/15/datasift-ceo-says-twi…
36KR Article: http://36kr.com/p/531693.html
Disclaimer — Written by Luwen Chen and edited by James Knuckles under the direction of Prof Charles Baden-Fuller, Cass Business School. This case is designed to illustrate a business model category. It leverages public sources and is written to further management understanding, and it is not meant to suggest individuals made either correct or incorrect decisions. © 2016 Published online 17 May 2016