InstaBrand Makes Sense of Big Social Data With Neo4j Graph Database
The company uses Neo4j to power its advanced social media influencer search engine.
Los Angeles, CA, April 5, 2016 (Newswire.com) - Pop quiz: which influencer with 100,000 to 500,000 followers has been the most influential in the Chattanooga, TN area with the hashtag #brunch amongst females in the last 90 days?
Stumped? InstaBrand isn’t.
InstaBrand, the leading influencer marketing company, recently released their influencer search engine to the public. The new search engine allows users to search by any number of demographic filters across hundreds of thousands influencers and tens of millions of social media posts. Results can be further filtered by the relevance of the post content to what the user is looking for, a level of detail previously unheard of in the influencer marketing space. Once influencers are selected the user can choose to use InstaBrand’s full-service campaign management services or work with the influencer directly. Interested brands and agencies can sign up for access to the search engine at https://instabrand.com/influencer-search.
The company’s engineers, a motley crew made up of veterans from advertising, venture-backed startups and Yahoo!, have been combining their expertise with learnings gathered by InstaBrand over years of helping brands and agencies run influencer campaigns. They knew they’d need to process massive amounts of data to power the features they had planned and one of their first challenges was choosing a primary database for all the data and relationships behind it.
“Given the connected nature of social data we knew a graph database would be the best fit,” said Micky Dionisio, CTO at InstaBrand. “As we evaluated our options and looked at the volume of data we wanted to store, coupled with the relationships and insights we wanted to draw, and the speed in which we needed from our queries - Neo4j became the obvious choice.”
As the world’s leading graph database, Neo4j enabled InstaBrand to store both the data and data relationships and offer a highly intuitive graph based search capability across millions of data records and build a highly contextual recommendation engine.
Currently, InstaBrand’s Neo4j database contains hundreds of millions of nodes spanning everything from social media accounts to posts, hashtags, mentions, locations and everything in-between. Dionisio says he could see that growing to upwards of one billion by the end of the year as the team continues to process new users and build out their proprietary graphs. The InstaBrand team uses all this data to power not just the search results but it’s recommendation engine too.
“You can really see the power of Neo4j applied to a big data set in our recommendations algorithms. Finding related hashtags or influencers would be a monumental task in a more traditional SQL database,” said Dionisio. “The nature of the graph lets us zero in on a particular item like a hashtag and then branch out to find connected posts and people in an intuitive way. And we can do all that with lightening speed.”
Speed is important to the InstaBrand team — all that data wouldn’t be nearly as valuable if users had to wait minutes for their results. The team has a healthy sense of competition around making queries as fast as they can be, often adding on suggestions to cypher queries posted in the team chat to see who can make it the most performant. All of this means users are able to get their search results in seconds while behind the scenes Neo4j is traversing tens of thousands of nodes and relationships to build the results.
“The team you assemble is just as important as the technology you choose and I’m proud of the choices we’ve made on both — we make influencer search look easy” said Dionisio. “We’re tackling the sort of data science problems few people outside of the social networks themselves are taking on and we’re absolutely crushing it every day.”
“Graphs are everywhere — and there is a huge demand for organizations to leverage the relationships within their data for valuable business insights,” said Emil Eifrem, co-founder and CEO of Neo4j. “It’s so great to see a hot startup like InstaBrand, who is at the forefront of bridging the gap between advertising and social media influencers, using Neo4j to build their search engine. I have a feeling this is just the beginning—we are anxious to see what they do next.”
For more information about InstaBrand and to get access to their influencer search engine visit: https://instabrand.com/influencer-search
InstaBrand is a leading influencer marketing company that has successfully managed over 800 influencer-driven social media campaigns for both Fortune 500 corporations and up-and-coming businesses across multiple social platforms such as Instagram, Vine, Twitter and Snapchat. Headquartered in Los Angeles, with offices in New York and Milan, Instabrand works with top brands including Calvin Klein, Pizza Hut, Samsung, Hyundai, AirBnb, Universal Studios and Verizon. For more information on InstaBrand, visit instabrand.com.