Tibco Statistica 13.2
November 2017
“Who are JMP’s competitors? (maybe we have them in narrow areas) — but there is nothing else like it.
Who is as interactive?
Who else does as great a job at DOE?
Who does analysis faster and bigger on the desktop?
Who else does Graph Builder as well?
No one, because JMP is truly unique in the crowded marketplace.” -John Sall
Company
- Statistica has been bought and sold several times in the past few years. First acquired by Dell and then sold to a private equity firm in 2014. In May of 2017, Tibco acquired Statistica after purchasing Spotfire in 2007. Tibco historically focused on service oriented architecture, business process management, and business optimization. Statistica has been around since 1984. It is a mature product with over a million users, positioned as a comprehensive, affordable product for predictive modeling, visualization, IoT, and Big Data.
- They have several products, but the analytical family of Spotfire products is positioned in their Tibco Insights Platform, which they also refer to as Augment Intelligence. Their other offerings are under the umbrella, Interconnect Everything. Categories under the Tibco Insights Platform include
- Visual Analytics (Tibco Data Virtualization, Tobco Spotfire, and Tibco Statistica)
- Reporting and Dashboards (Tibco Jaspersoft)
- Streaming Analytics (Tibco BusinessEvents, Tibco Live Datamart, and Tibco Streambase)
- Data access and connectivity is extensive and includes IoT devices with their streaming capability.
- Tibco Spotfire has a global network of resellers, systems integrators, OEM partners and solution/technology partners.
Product
- Since acquiring Statistica, to differentiate it from Spotfire, they are calling it more of a data science offering with te “ability to make analytically fueled decisions. Statistica helps you innovate and solve complex problems faster, empower more people, and infuse algorithms everywhere to ensure insights quickly turn into optimal outcomes.”
- New in version 13 and the most recent 13.2 (released in late 2016): streamlined data prep and analytic workflows, edge-scoring capability to address IoT analytics use cases, extended in-database analytics to Apache Hive, MySQL, Oracle, and Teradata, and network analytics for fraud detection.
- Statistica allows the user to export models into C, SQL or Java to deploy models in different data environments e.g., run Java models on Hadoop.
- Native in-database capability.
- Statistica includes a variety of predictive modeling and statistical algorithms, and combine with Hadoop, Lucene/SOLR search, Mahout machine learning and advanced natural language processing. Version 13 includes a new stepwise model tool that progressively recommends optimum models for users.
- Integrated with R, Python, algorithm marketplaces like Algorithmia and Apertiva, and H2O.
- Language analysis for lemmatization, stop words, parts of speech, etc.
- Network analysis.
- Support for streaming data and real-time data discovery.
- Statistica Code Deployment with a Model Converter that converts Statistica-generated PMML models to C#, Java, SAS, Teradata, SQL User Defined Function in C#, SQL stored procedure in C#. Rules, Java for MapReduce/Hadoop (beta).
- Database connectivity includes PC file formats, popular RDBMSs, SAS, Hadoop/Hive, Cloudera, and more.
- Support for semi-structured data like PDF, XML, HTML, text, and ASCII les, and export directly to Tableau and Qlik
- Statistica is Windows-only and can be run on PC’s and servers as well as in the cloud.
Target Audience
- Strong base in manufacturing and pharma, but also in other industries like financial services and healthcare with analysts and data miners across all areas of a corporation.
- Statistica is used in academia as well.
Feature Strength and JMP Response
- For customers who want algorithms like support vector machines (SVMs), multivariate adaptive regression splines (MARS), market basket analysis (also known as association and sequence analysis), network analysis, support for streaming data, and some text analysis, Statistica provides these. JMP offers several flexible nonlinear modeling methods that may perform similarly to SVMs and MARS. For JMP 13, we have added market basket analysis and some text analytics. For customers who ask for SVMs, we have other very good “universal approximator” algorithms like neural nets, decision trees, etc.
- Statistica provides several model deployment options via PMML and conversion to other languages. JMP encourages the use of SAS for model deployment.
- Tibco may sell Spotfire and Statistica together or they may try to sell Statistica to current Spotfire customers. Tibco Spotfire as a client-server offering is more of a competitor to SAS Viya. There is considerable overlap in Spotfire and Statistica. If customers already have Spotfire on the desktop, product rationalization would be a tough sell to add yet one more statistical desktop product. If Spotfire is deployed client-server, adding a desktop-only product with overlapping functionality would also be a tough sell. The products may merge and become more integrated in the future, but for now, both Spotfire and Statistica are Windows-only with a lot of overlap. JMP is distinguished with its superior DOE, Reliability, visualization, and more.
Feature Weaknesses and JMP Response
- Scant design of experiments. They have a few traditional designs. JMP is known for exceptional software implementation of and innovations in DOE.
- They have some reliability analysis, but JMP has invested more in this area.
- Though Statistica claims integration with SAS, they can’t offer the level of integration and integrated technical support that we do.
Opportunities for JMP
- With so many changes in ownership, Statistica declined in Gartner’s Magic Quadrant (this was prior to Tibco’s acquisition, but it is still for version 13.2: February 2017 reprint). This creates some uncertainty and may be an opportunity for JMP.
- The recent acquisition may prove to be analogous to IBM’s acquisition of SPSS—one more desktop statistical software package being subsumed by a larger and broader software company.
- As the Gartner reprint highlights, ease of use (even getting started) and lack of good data exploration and visualization capabilities are drawbacks. Performance issues and slowness to add requested features are also cited.
- Statistica’s data visualization capabilities are not as interactive as JMP’s. They are even integrated with Tableau and Qlik.
- Though text mining is listed as available in the enterprise edition, very little information about these capabilities is evident. JMP’s basic text analytics may differentiate it.
- JMP’s investments and innovations in DOE, Reliability, and more modern statistical methods set us apart.
- This site shows complaints about lacking customer support and slow performance.
Threats for JMP
- More complimentary products from Tibco to potentially make it stickier both for IT and for analysts.
- More mature messaging for IT than JMP currently has.
- More integration with other offerings.
Pricing
- Pricing since Tibco acquired Statistica is not currently available. This reseller has some pricing options in Euros.
- Academic pricing starts at $24.99 for a 6-month rental, $49.99 for a 12-month rental, and $142.99 for a perpetual license.
Competitive references where we’ve won against them or displaced them.
- Archer Daniels Midland
- Vishay Siliconix Itzehoe GmbH
- CEA GRENOBLE
- Dyconex AG
- BP Biofuels SD
- US Synthetic
- PCC Structurals, Inc
We also have customers listed in Salesforce who use JMP in combination with Statistica. S
