I have to confess that I have always liked Spotfire. The marketing may not have been as aggressive as some competitors, and it might not have had the immediate sex appeal of other products, but for anyone serious about visual analytics it is almost unbeatable. I use the word “serious” to mean complex requirements, high volume diverse data, and the need for advanced analytic capability. The latest release, Spotfire X, brings natural language query (NLQ), so users can use spoken language to formulate complex queries and visualizations. It also adds more automation to the data wrangling process, allowing data preparation tasks to be recorded, and it now offers support for real-time streaming data. The user interface has been given an overhaul, making it more attractive and productive.
What distinguishes Spotfire from much of the competition is its support for demanding requirements. It does all the usual things – charts and dashboards, but it also offers easy to use advanced analytics capabilities such as clustering and regression analysis. Users do not have to input a whole load of parameters, Spotfire does the analysis automatically. It also provides arguably the best R runtime environment in the industry with its TIBCO Enterprise Runtime for R (TERR).
Spotfire is part of a much broader offering from TIBCO, embracing data and process integration and IoT capabilities – hence the ability to analyze streaming data in Spotfire X.
If I was to have one complaint it would be that the company is not embracing AI assisted analytics aggressively enough. They have made a start with natural language query, but they need to go much further.
The analytical capability of Spotfire is tiered. Novice users or those with modest requirements need to go no further than the first tier – a drag and drop visual interface for the creation of virtually any chart, graph, map, etc. The next level is called Out-of-the-Box Advanced Analytics. It includes the ability to incorporate statistical features into visuals, look for correlations, apply regression models and sophisticated forecasts, and cluster data to find similarities between different instances within a database. Much of this is available from a sidebar menu and requires users to enter the relevant parameters.
The scope of Spotfire is enhanced considerably by its tight integration with the R statistical and analytical language. It also features a high-speed runtime environment for R (TERR – the TIBCO Enterprise Runtime for R), which overcomes the relatively poor performance of the open source platform and makes it suitable for enterprise deployment. Spotfire supports other languages such as SAS, MATLAB, KNIME, S+, and data platforms such as Spark and MapReduce databases. These capabilities mean developers can create sophisticated analytical functions that are made available to users in much the same way as any other function and are called Data Functions within Spotfire.
For very advanced functionality Spotfire can work with TIBCO StreamBase so that analytical models can be applied to real-time streaming data. This has obvious applications in the Internet of Things, and comprehensive support is provided for Apache Spark.
With its advanced analytics capabilities Spotfire is often used in businesses that have to deal with technical or financial data. Energy companies, pharmaceuticals, and firms in financial services are all examples of where Spotfire finds full usage. However as businesses in general demand more from their analysis, so Spotfire will see more widespread adoption.
The in-analysis collaboration supports live discussion right down to individual data points via a messaging interface. Visualizations can also be annotated, and annotations can also serve as discussion points.
Beyond the use of Spotfire for diagnostic and exploratory data analysis, support is also provided for text analytics, the primary purpose of which is to detect sentiment, although tasks such as categorization, finding topics within documents, and a whole host of other text analytics tasks can be performed. With its Enterprise Runtime for R Spotfire supports the execution of R scripts for predictive analytics. Other programming languages are also supported including SAS.
The real-time analytics capability supports the processing of streaming data and the detection of pre-defined events. This capability extends to complex event processing where complex conditions are detected in real-time and actions or alerts generated.
Location analytics allow business applications which benefit from location information to be created. Obvious applications include logistics, public services, retail (knowing when prospects and customers are near a store), travel, and hospitality industries.
Many businesses will use Spotfire simply for its visual analytics and data exploration capability, but it is reassuring to know that a single platform will support most types of analytics an organization might wish to perform.
Spotfire comes in two main variants:
- Spotfire Platform enables sharing and collaboration in an enterprise setting. It also embraces other forms of analytics including predictive, prescriptive, content (e.g., text) analytics, location analytics, and real-time analytics. Only a few suppliers challenge this breadth of capability – and at much higher cost.
- Spotfire Cloud provides full Spotfire functionality via a cloud-based service. A web browser interface is used, and generous cloud storage is provided for data. An enterprise version of the cloud offering is also available, and Spotfire for Amazon Cloud Services is a pay by the hour preconfigured offering.
Its smart memory and data management greatly enhance the speed of Spotfire. For more modest data sets the in-memory processing ensures very high performance. Larger data sets can be processed in-database, and a hybrid approach called On-Demand optimizes the data held in local memory and that contained in the database. This capability is unique to Spotfire.