Cloud BI: Business Analytics Services
Tableau Public Premium
“Customers have lots of data, it’s coming in faster than ever before, and it’s in a variety of different formats. In many cases, when people look at their data, they just see a big set of rows, columns, and numbers,” says François Ajenstat, director of product management with Tableau Software. His Tableau Server product, which essentially produces live analytics charts as a service on iPads, has been the game-changer for BI over the last two years.
“It’s hard for them to understand what they even need to do,” he continues. “Tableau helps them see their data in ways that were never possible before, all on mobile.”
Tableau’s iPad app presents live views of incoming data as workbooks, which should be very familiar to folks whose analytics experience centers around Excel. Tableau Server actually produces graphics in HTML5 format for browsers on multiple platforms. But the reason Tableau produces this iPad app is because iPad users expect an app, not a Safari bookmark. “In many cases, analytics are embedded in portals or in Web sites,” Ajenstat adds, “and they have to be optimized for mobile as well.”
Tableau Server is designed to be integrated into organizations’ existing on-premise data centers, as the mobile visualization layer their BA tools may have lacked. The visualization tool establishes direct connections with a number of DWs, including Pivotal Greenplum. Tableau has literally been giving away its visualization services for client-side browsers under the brand name Tableau Public. But to use Public, the client-side device needs to download the data that’s being visualized, which renders it not always particularly convenient for iPads.
Tableau Public Premium (formerly Tableau Digital) works somewhat differently, and this is where cloud-based service comes into the picture. With Public Premium, an integration layer is established between your data warehouse and Tableau’s servers. No data transformation takes place, so you’re not actually storing the data a second time. Tableau generates the visualization, and then produces it for the client-side app, so the iPad downloads only the renderings. You can then embed these renderings in Web pages. For example, a public company can actually embed a live visualization of its financial performance, with interactive drill-down features (and rules limiting the extent of what everybody sees) on its shareholders’ Web site.
Competitor Roambi uses a different, though still quite innovative, approach to enabling live visualization on mobile devices: Data warehouses through history have utilized so-called JDBC connector drivers to enable Java applications to deliver queries to their OLAP cubes. Roambi Pro borrows this same, pre-existing concept, employing a vast set of JDBC connectors to acquire data from the same places your existing DW would look for it, including IBM DB2 relational databases, SAP HANA in-memory databases, Salesforce repositories, and even Excel spreadsheets.
There’s a degree of integration that takes place, but not so much that you can use the integrated product as a DW in itself. The integrated data resides in a secure storage area on the Amazon S3 cloud. The Roambi Pro service accesses your data from there, and delivers your BA visualizations to your mobile device through an app. Again, the device does not download and cache the data remotely, so you can use this service more freely in low-bandwidth areas.
Using a similar connectivity model, Yellowfin delivers a similar visualization service, but this time geared more toward collaboration and sharing. Rather than re-inventing the wheel (for the twelve-thousandth time) with respect to collaboration tools, Yellowfin produces an interactive graph that triggers the database query in the background, and that’s embeddable inside most any other browser-based tool you use to communicate with co-workers — the big one showing up in the demos being Salesforce.
A Web server gathers several of these embeddable charts together to produce a customized personal dashboard, which you can use to monitor business progress in real-time from any device with a modern browser. Yellowfin’s mobile app doesn’t have to be too sophisticated, just a portal that enables one-tap access to these dashboards, and the ability to drill down through data sources to create new charts on the fly.
On the back end, Yellowfin also uses JDBC connectors to communicate with whatever the OLAP cube connects to, including other ODBC connectors to Microsoft Office apps (Excel being one), and Greenplum and Teradata DWs.
While MicroStrategy appeals to this same, new breed of BI customers who are dazzled by the sudden wealth of visualization options, its version is something of a lure to get you interested in the idea of moving the entire BI processing platform into the cloud, including the integration and ETL layer, and even the core of the DW itself.
In effect, MicroStrategy Cloud is an integration service whose payoff is a live visualization layer that plays on mobile devices. So one of the things you pay for is the direct connection to this service, which takes place over a secure VPN. If you do subscribe to full data hosting, then your data is stored on a Netezza system, and the analytics processes are run from there. MicroStrategy leverages its position as a hosting service to become a cloud service, effectively selling you access to a DW appliance maintained by its staff, off-premise.