While you were away on summer holiday, several data discovery players released new versions. The highly anticipated Tableau 10 is now available for download. Qlik Sense 3.0 was released in late June and TIBCO Spotfire a month earlier. Here is a quick summary of what is new, wild and exciting for data visualization enthusiasts.
Tableau 10 Analytics Savvy
This week the Twitter-sphere was buzzing about Tableau 10 again. In case you missed the memo, Tableau 10 is now generally available for download. A virtual launch event is scheduled for August 18th at 11:00AM EST. An excellent, interactive summary of highlighted new features is also available for the time-starved.
My personal favorite Tableau 10 enhancements are in the Analytics area. Since I first saw the Analytics tabbie announced several years ago, I have been waiting for it come to life. In Tableau 10, I am finally seeing that start to happen with the addition of drag-and-drop Clustering. The new approach humbles the former R script approach. Tableau clustering uses a popular data mining algorithm called k-means to automatically identify related records. This kind of easy button intelligence is starting to show up in mainstream visual analytics tools. I hope to see even more Analytics features at the upcoming Tableau Conference.
Tableau 10 also brings more sophisticated calculations with groups and level of detail expressions. A lovely usability improvement was made to the table calculations user experience and the Highlighter feature for dimension comparisons is an absolute analytics delight. New Marimekko charts and device specific design previews are fantastic additions.
Standing ovation for Cross Database Joins and Cross Data Source Filtering that is capable of executing filters across most data source types as if they were a single unified source. This is a huge enhancement and a top ask delivered. In the past to work around it, you might have had to “Tableau-ready” – flatten your data for it to be filtered together. Now you can connect to disparate data sources, on-premises, hybrid or cloud, and avoid time consuming modeling tasks. Another performance bonus is on-demand connections that only query a data source when it is required to display data.
Speaking of data sources, I have always valued the Statistics connections (SPSS .sav, SAS .sas7bdat, R .rda and .rdata) for sharing analysis with other industry leading analytics tools. Other interesting connectors include SQL Server contained database support, Cisco Information Server for modern data virtualization/logical data warehouse architectures and Google Sheets. (Note: I didn’t realize how popular Google Docs was until I went back out on my own. Several of my clients – big and small – do not use Office.)
Other valuable enterprise related features include data source discovery favorites and smarter search, version control with Revision History for workbooks and data sources, lots of API additions, site-specific SAML authentication for regulated reporting environments, easier clustering and Kerberos authentication for Oracle. Subscriptions were expanded to deliver dashboards to other users – no more having to script with TabCmd for that common scenario. Tableau’s enterprise administration reporting and performance recorder was already best-in-class. In Tableau 10, there are related visual data source usage cues and less exciting licensing reports.
All in all, Tableau 10 is already getting a warm reception in the market. The financial analysts that I follow are reporting higher competition and the need for Tableau to up level enterprise sales yet win rates are still high. One of the big portfolio management firms noted in their research that the business still loves Tableau while BI professionals and IT are being heavily courted by mega-vendors.
I’d like to see more innovation from Tableau – I always tell them that. Since I was awed by the Kraken v8 key note demos, my expectations are always sky high. What happened to the infographic acquisition or the super cool Dato machine learning? I see potential for Tableau to expand and differentiate even in a tough market. With the mega-vendors distracted by bigger cloud battles, Tableau’s niche focus can be a strength if they invest in the right places and keep moving swiftly.
Qlik Sense 3.0 features Bubbly Data Prep
Earlier this summer, Qlik Sense 3.0 debuted. Although it is not as large of a release as Tableau 10, it does have a variety of enhancements that I will highlight. If you want a more detailed overview, watch the short Qlik Sense 3.0 video and check out the official summary.
The most popular capability in Qlik Sense 3.0 is the unique “bubbly” approach to visual data preparation that profiles the data and automatically associates it. Look out Paxata – you are not the only ones disrupting data prep with intuitive visual approaches.
Qlik Sense 3.0 also has a creative Smart Search capability and Visual Search for finding published charts. Over 20 more popular data sources including Twitter, Facebook, YouTube, MailChimp, SugarCRM, Google AdWords, and a REST connector that provides access to JSON, XML and CSV files for reporting projects in an API economy were added. NPrint for report distribution is a spectacular feature. Other new features include app styling, widgets, and time-aware charts.
From what I can tell, the recent Thoma Bravo purchase has not affected Qlik as much as I assumed it would. The last quarterly update exceeded expectations despite intense competition, QlikView/Qlik Sense decisions and lack of complete feature parity between the two offerings. The Partner and OEM side of the Qlik business seem to be shining brightly.
Spot the Differences in Spotfire 7.6
A waterfall chart was added along with new KPI charts (looks like Extended Results acquisition influence but missing indicator designs right now). Cross tables are more elegantly formatted. Nice check-box options were added to drop-down filters. Box-plot visuals have improved and Tile Map Services beautify mapping look/feel options.
Speaking of maps, auto-zoom and an awesome radius selection capability was added! This is my favorite new Spotfire feature. Using a Rectangle or Lasso selection, users can now mark data around a location on the map.
Another really great improvement is Source View for understanding data table and function relationships. Like the other data discovery vendors, more data prep features are now available including splitting and unpivoting. A big win in Spotfire 7.6 is added support for combined linked and stored data in the same data table with granular refresh options.
Spotfire 7.6 also adds data connectors to Salesforce, Google Analytics and SAP BW 7.4 along with advanced connection parameters and stored procedures in many other connectors including Amazon Redshift, Hortonworks, SQL Server, Oracle, MySQL, Pivotal Greenplum , HAWQ, PostgreSQL and Teradata.
For showcasing Spotfire on a website, check out the how to embedding video for Spotfire cloud. Although Spotfire visuals are not as easy to embed as other offerings, it is not difficult to copy URL and scripts into HTML.
When it comes to Smart Data Discovery and baking predictive analytics into the user interface, Spotfire is a pioneer with Statistics Services, TERR and predictive modeling. Although Spotfire does not get as much media coverage as the other data discovery vendors, they really have made significant user experience strides across the entire customer journey. I noticed that they added an updated community site and a plethora of other resources that did not exist two years ago. The 2014 purchase by Vista Equity seems to be good for them.