It's easy to get lost in the hype around big data. Developers could benefit from identifying the most common use cases of big data apps today when working with business managers and operations teams.
At the Structure Data conference, Cloudera CEO Tom Reilly said, "We are seeing things move from talk of zoo animals, to building data lakes and now seeing organizations move towards making board cases for big data applications." Reilly said most enterprises are developing big data apps that fall into three main categories: customer insight, product insight and business risk.
Improve customer insight
Customer insight applications are sometimes described as Customer 360, and are common in telecommunications, retail and banking. These types of applications are used to build profiles of customers and provide services like next best offer. These applications are able to leverage information about a customer's prior history and their profile to suit their unique needs. The goal is to improve and streamline the customer experience. Customers can be guided to useful offerings without the distraction of hundreds of options.
In telecommunications, these types of applications are being used to reduce customer churn. GoPro is using these techniques to help create social networking communities across users with similar interests.
Build a smarter service
The second class of applications is used to improve the kinds of products and services that enterprises can offer. These leverage information from customer data to make it easier to customize services for someone's unique behavior. For example, Alliance Insurance has developed an insurance product to allow drivers to pay based on their mileage. This data is gathered by special sensors connected into a car's telemetry system. This makes it possible to offer safer drivers a deeper discount.
Nike is embedding sensors into its shoes. Not only does it give the runner better data about their performance, Nike can also send out re-ordering alerts to customers when the shoe has reached its lifetime.
Cloudera is also leveraging product insight principles in its services to identify potential configuration issues. The company has developed a tool that analyzes an enterprise's configurations of the Cloudera services. Reilly said this lets them solve about 20% of their customer support cases proactively before an enterprise's operations team has noticed a problem. This costs Cloudera less than if the customer had to file a ticket first.
Business risk applications are used to improve security, identify money laundering and reduce fraud. These kinds of applications are not just used by banks. The Communications Fraud Control Associations estimates Telco Systems, a communications carrier, loses about $46 billion per year to fraud. Vodafone has implemented a big data analytics services to help reduce this risk.
One of the challenges of big data apps is that they require domain expertise to make them useful in a particular context. This makes it easier to iterate an application in a way that mitigates problems and improves workflows. For example, Cloudera is working with Argyle to customize Telco fraud apps and Amdocs to reduce Telco churn.
Another risk lies in losing market share to upstarts with better products and services. For example, Tesla's car can learn about a user's behavior to improve the user experience in ways that other auto makers have not dreamed of. Reilly noted that the first time he drove up to his house he had to manually adjust the height to pull up the driveway safely. However, the car's application learned about this, and it automatically raised the car when he got home the fourth time. Reilly said, "If you are an auto insurer and not thinking about connected cars, you risk losing market share. If you are an auto maker, you will lose customers."
How is your enterprise using big data applications? Let us know.
Is big data too big for cloud analytics?
Companies confide in big data apps as IoT rises
Keep calm and assess infrastructure requirements