How Startups Are Using Big Data Tech to Disrupt Markets
Big data has made a dramatic impact on companies all over America, but running big data programs is only one side of the puzzle. After collecting the data, companies need to analyze it. A huge part of analysis is creating visuals that explain large amounts of seemingly abstract data in a clear, concise way. It’s no surprise then that many companies are turning to data visualization tools to streamline the transformation of their business data into something more useful.
With a surge in start-up companies seeing big results in figures and customers, it’s no surprise to discover they haven’t done this all on luck alone. Companies like Pandora, Uber, Netflix and other start-ups that went big use big data to determine what their customers want more of, who their friends are, and generally what they like—all before their customers do.
Here’s how four start-ups are using big data to disruptive their respective industries:
Pandora and Netflix
Pandora uses machine learning to engage customers and help them find new music based on the music they already listen to. Machine learning generally gives computers the ability to learn a person’s musical tastes based on the choices they make. This is the same technology your Gmail uses to decide which of your emails are deemed spam and which you want to read and the technology Amazon uses to recommend the products customers are most likely to buy.
Netflix is also a huge fan of this type of big data. When a customer watches a movie, they can give the movie a rating, and based on this rating, Netflix will recommend other movies. They can also see patterns in a user’s queue to determine what types of movies they’re most likely to want to watch. This kind of machine learning can also tell lawyers what the probability is that they will win that big case.
One of the data visualization tools Pandora uses is to allow customers to decide what artists and songs they like. When a “random” song is played to a Pandora user, the user has the option to give it a thumbs up or a thumbs down. Pandora takes this information to deduce which other artists and songs the user will like. The more songs a user gives this rating to, the more information Pandora has to cross match and determine which songs a user will like.
This type of data is then sometimes sold to the music industry to determine what genres should be funded with more money and which new artists the industry needs to seek.
One of the big data techniques that Uber uses is regression analysis. Among other things, regression analysis can determine the size of a neighborhood. It can also determine what the prices of the houses should be in the neighborhood, depending on the listed prices, demographics, schools, and other information.
Uber uses this information to determine what neighborhoods will be busy on a Friday night, what areas have the most bars and clubs, and when to add “surge charges” to their customers’ bills.
Uber can also determine customer satisfaction and loyalty using this method. Both their drivers and riders are able to rate each other, and Uber can tell a lot from both the ratings. This type of big data is so sophisticated, it can also take other variables like the weather report into factor. It works best with quantitative data.
Foursquare uses a data mining technique called social network analysis. This type of big data finds the most popular person in a social network group and markets specifically to that person under the assumption that the most popular person will then market to all of their friends. Foursquare takes information based on a person’s habits, where they check in and what they like to give their customers recommendations on other places they might like. But it not only takes your likes and dislikes into consideration; it studies the patterns between people and their ties between the group and with people outside the group. It focuses on the importance of one individual in a group and the minimum number of connections between two people. It also analyzes the social structure of their customers.
In doing this, Foursquare can not only market places but products and neighborhoods to individuals or one individual directly called a “maven” in big data terms. The maven will then recommend this location to all his or her friends and the other friends will follow suite.
Zynga, the gaming site, also uses machine learning to decide what games are popular. It gives users a large choice of games, and once the user clicks on the game he or she wants to play, the big data tells Zynga what games they should make more of. It can also detect patterns between games that users play. If two games are consistently played by users, the common connection between the two games can be analyzed by using big data.
Companies today are generating more data than ever before. While it’s great to have data, it really doesn’t do much for you and your business unless you are able to use the data in some way to make quick, and useful business decisions. Data visualization tools and techniques have not only grown in popularity recently, but in quality. The necessity of these tools to quickly provide clean and simple visualizations of data in order to help businesses draw insight from their data will only increase as time marches forward.
About the author: Brigg Patten writes in the business and tech spaces. He’s a fan of podcasts, bokeh and smooth jazz. Mostly he spends his time learning the piano and watching his Golden Retriever Julian chase a stick.