Once a month, receive latest insights, trends, analytics information and knowledge of Big Data.
Luminar Case Study
Luminar, a unit of Entravision Communications Corporation, is the first analytics and modeling provider that uniquely serves the U.S. Latino market. The company uses its proprietary technology to transform Latino consumer data into insights and business intelligence. Luminar unlocks big data and develops predictable models that reflect the behaviors and attitudes of today’s Latino customers across a wide range of vertical categories.
Typical market research studies involve three steps:
• Identify a target segment for the study
• Create a sample panel that matches that segment
• Study the sample panel and apply findings to programs for the entire population
Luminar did not want to take sample transaction from adult Latinos living in the United States; they wanted customer data on the entire population. And they did not want to store only some transactions from one or two sources; they wanted to acquire and save as many transactions as possible from as many different sources as possible. Luminar wanted to build models based on empirical insights, not sample insights.
With its data set consisting of nearly 15 million adult U.S. Latinos, Luminar was in a position to investigate some of its clients’ most challenging hypotheses about what those customers wanted. As Luminar’s data set grew in size and scope, the company was able to offer more reliable market research data. The Luminar IT team also needed a storage platform that could handle the size and complexity of the company’s growing data set.
Luminar set about building the largest U.S. Latino customer datastore it could obtain, with as many transactions as it could capture for adult U.S. Latinos. With its data set consisting of nearly 15 million adult U.S. Latinos, Luminar was in a position to investigate some of its clients' most challenging hypotheses about what customers wanted.
Luminar compared Hortonworks and other solutions for their ability to support their growing data workloads, without the risk of getting locked in with something that might limit them in the future. They appreciated the Hortonworks commitment to nurturing Hadoop in the open source community.
One of Luminar’s key reasons for choosing Hortonworks was Hortonworks Data Platform’s (HDP) interoperability with other pieces of their storage and business intelligence framework. HDP integrated easily with their data store and Amazon Cloud platform.
Before switching workloads to HDP, Luminar was ingesting 2 terabytes of new transactional data per month, from 300 raw sources. Now, Luminar appends new transactional data every day from 2,000 raw sources, which adds up to 15 terabytes per month. With Hadoop, Luminar increased the amount of data it captured by a factor of eight.
Luminar updates its data with constant ingest of new transactional data. Before Hadoop, it could take Luminar several days to ingest and join data to refresh one of their models. Now it takes only hours. This allows Luminar to tackle a greater number of projects and to process data for more clients in parallel. The company avoids an analytics bottleneck.
Luminar grew its business by delivering its data analytics services on a client-by-client basis where they gathered information in batch and then incorporated that into executive
presentations. The company is now building custom portals for select high-value clients. This will give those clients near real-time (and constantly updated) access to their most important data. Luminar sees this innovation as an important next step to scaling their business.
Share
Sign-up
Apache, Hadoop, Falcon, Atlas, Tez, Sqoop, Flume, Kafka, Pig, Hive, HBase, Accumulo, Storm, Solr, Spark, Ranger, Knox, Ambari, ZooKeeper, Oozie, Phoenix, NiFi, HAWQ, Zeppelin, Atlas, Slider, Mahout, MapReduce, HDFS, YARN, Metron and the Hadoop elephant and Apache project logos are either registered trademarks or trademarks of the Apache Software Foundation in the United States or other countries.
