Looker Adds Capabilities to Its Datafold Engine

Business intelligence SaaS company Looker has enhanced its analytic capabilities by allowing for  persistent derived tables, which add speed, interactivity, and flexibility to data analysis. Instead of creating billions of rows of derived tables every time a user looks at a graph, the tables will update at certain times or after certain triggers, such as after 100,000 rows of data are added. This also means that graphs charting things such as "customer lifetime value," which might include multiple derived tables, don't need to be manually updated or restructured, saving time and improving the user experience.

The new Datafold Engine offering strengthens Looker's focus in data discovery, according to Keenan Rice, Looker's chief marketing officer. "A lot of traditional BI is built on reporting, things like standard financial statements, auditing, SOX [Sarbanes-Oxley Act] compliance, and complex workflows. The new world of business intelligence is around this idea of data discovery and self-service BI."

With Looker, data scientists initially create data sets using Looker's proprietary language, LookML. The language will then translate the queries into SQL, the standard language of relational databases. Unlike GUI interfaces that remove all code from the equation, Rice believes this method is preferred by most data experts. "The majority of analysts coming out of school want code, and this also makes it much more of a platform," Rice explains.

Once created, people throughout the organization, from members of sales teams to managers to marketers, can drill into the tables whenever they need to. They can look at the data in multiple ways and drill down through the data sets. Dashboards can be a mere starting point in analysis. In this way, Looker helps free data scientists from these routine requests by empowering the end user.

Conventionally, limited computational power meant that data scientists had to extract a small portion of data from a database before doing analysis. Looker allows data scientists to perform analysis without taking this step, allowing for greater flexibility. "The idea of being 100 percent in database is how the database world is thinking, but not how the BI world is thinking," says Rice. "It's showing our customers we have a lot of innovation."

Looker interacts with cloud-based data storage services such as Amazon Redshift, Amazon RDS, HP Vertica, Greenplum, and Teradata Aster. It also works with companies that store their data on-premises, which Rice estimates is roughly half of all Looker clients.

Looker, which was founded in 2011, has penetrated the market particularly strongly in Web-based companies, which generate reams of data. Its clients include Gilt, TaskRabbit, Upworthy, the upstart app HotelTonight, and collaboration software tool Mindjet.

Related Articles

Looker Improves Visualizations, In-Browser App Development

BI product update will help move Looker upmarket.

Birst Launches New Visualizer BI Platform

The Birst Visualizer incorporates visual discovery, guidance, and recommendations.

SAP Brings Analytics Platform to Mobile Devices

The latest rapid-deployment solution for business intelligence can be up in running in three weeks or less.

Deriving Value from Big Data

Diverse data sets add to complexity but present new insights.