At its "JOIN" conference in San Francisco today, Looker, a Santa Cruz-based BI vendor that been around for five years now is, fittingly, announcing the fifth major version of its product. With it comes a slew of new core BI features , but an even larger number of functional additions that flesh out the product's platform credibility.
To that end, Looker 5 is introducing an "Action Hub" for reaching out to various Software as a Service (SaaS) applications; Viz Blocks, an extensibility mechanism for Looker visualizations; Data Blocks, an integration channel for public data sets; and Applications by Looker, which are end-to-end solutions that combine dashboards along with the aforementioned Actions and Blocks.
Actions are pretty cool. In fact they remind me of a feature in Microsoft SQL Server Analysis Services of the same name. On the Microsoft side, actions provide for clicking on a data cell (either in a Microsoft tool or a custom application that uses the right APIs) and run a query, navigate to a related report, visit a Web site/URL, or execute a command line command, among others (there are 9 action types in all), where the value clicked can be passed as a parameter.
Looker adopts that concept and takes it further by extending it to entire sets of API calls to SaaS applications, like Salesforce, Google Drive, Box, DropBox, Twilio and Zapier (which can, in turn, connect to a number of other Web sites and services). Looker actions can also be launched from UI objects like buttons, rather than just data cells.
Viz Blocks and Data Blocks
Data Blocks provide another packaging mechanism, but for public data sets, rather than new viz types. The facility will launch with a public repository of weather data dating back to 1939, and more will be forthcoming, including data sets focused on demographics, economic indicators and mapping data that will allow for geographical dimensions that include county level.
Procuring data using Blocks is consistent with Looker's connector-less paradigm. Even mainstream data sources are queried via Code Blocks, so using Data Blocks for public data is consistent in philosophy and technology.
There's an app for that
Applications by Looker are a content packaging mechanism that ship related Actions, Viz Blocks, Data Blocks and dashboards in a unified fashion. Applications for marketing analytics, IT analytics and time series analytics (which Looker calls Event Analytics) will be part of the offering.
On the pure play BI side, Looker 5 adds Data Merge (for blending across heterogenous databases); Data Tools (an end-user data environment that provides for more data exploration freedom than would a dashboard); visualizations in Looker's SQL Runner tool (which lets users write raw SQL queries to gather data); security enhancements; support for Druid (an open source OLAP-on-streaming-data framework for which a number of database connectors exist); and 57 new stats functions.
That's a lot of stuff, but Looker rolls out updates every 4 weeks, so, really, Looker 5 is a roll-up of several of these rapid-release updates. The fast-ring/slow-ring approach to releases is a good one though -- it enables product teams to get much more done than they would if they had infrequent, "big bang" releases. Another great BI example of where this is working well is Microsoft's Power BI product, which issues monthly updates for its product.
Looker says it's got more than 325 employees and more than 1000 customers, which is no small feat, especially for a single-product company. Even if you're not one of those customers, you benefit, as greater competition from Looker makes the other BI players work harder and faster.
Looker 5 will roll out to customers next month.