Dashboards are similar to visualization tools, in that they provide easy-to-understand graphical presentations of data. But their purpose and user base are different. The primary goal of visualization tools like Tableau and QlikView is to allow data analysts to explore large data sets to discover patterns through visual analysis. Dashboards are designed to allow monitoring of a number of key metrics to ensure that everything is going to plan. They are really Corporate Performance Management tools that measure progress towards key performance metrics and overall organizational performance. Dashboards are all about measuring performance and providing real-time data to managers to ensure that KPIs are met. Related to this are scorecards, which monitor overall progress for executive staff. Most full-stack BI tools allow companies to build their own dashboards, but a number of pure-play dashboard tools exist, some of which are surprisingly sophisticated in their range of capabilities.
Reviews of iDashboards on TrustRadius indicate that the product has excellent graphic elements and chart types, but requires technical assistance to set up and does not easily connect to application data sources like CRMs.
Other significant vendors in this category include:
- Dundas has been selling its Dashboard product since 2010, and it unifies data visualization with dashboard design. It is based on Microsoft Silverlight technology and integrates particularly well with the Microsoft BI stack.
- Yellowfin is an emerging dashboard vendor that is particularly focused on creating mobile dashboards.
The distinction between Business Intelligence and Predictive Analytics is that BI is usually considered descriptive, i.e., looking at what happened in the past to understand business drivers, while predictive analysis is about finding hidden patterns in data using complex mathematical models to predict future outcomes. The emergence of big data platforms like Hadoop—used for processing enormous quantities of data to find patterns—and very fast in-memory analytics products has resulted in some blurring of the lines between big data and predictive analytics.
Increasingly, BI vendors are entering the predictive analytics space either through development or acquisition. For example, IBM acquired SPSS in 2009 and has integrated this technology with the Cognos BI product suite. Information Builders has built its own predictive product based on Revolution Analytics R, and Netezza is building predictive capabilities directly into its data warehouse. SAS has gone in the opposite direction and has built BI capabilities into its predictive analytics suite. SAP's 2013 acquisition of KXEN, a predictive analytics tool, and its subsequent integration with the SAP visualization tool Lumira and the rest of the Business Objects Suite, is the most recent example of this phenomenon.
The major issue with predictive analytics software is that, although very powerful, these products are extremely complex and require people with a very advanced skill set to use them. The talent pool of data scientists with the requisite skill set is very small, which means that customers often struggle to use them fully. This dearth of data science talent has resulted in a major push to make these tools usable by less skilled staff – perhaps someone with a degree in computer science rather than a PhD in computational and data sciences.
Mindshare in the predictive analytics space used to be owned by SAS and SPSS (which was acquired by IBM in 2009). Other well-known platforms are KXEN (now part of SAP), Tibco Analytics and StatSoft, which was acquired by Dell in March 2014.
Lattice Engines, founded in 2006, focuses on predictive sales and marketing analytics and very recently announced a partnership with the new Salesforce Wave platform.
The company that has grabbed all the mindshare of late though is Revolution Analytics and their open source software, R. This product has become so central to the new BI/Big Data landscape that more than two million people now use R, and it is the highest paid IT skill, and the most used data science language after SQL.2
Why all this frenetic interest in R? The answer is that R is a free analytics engine that can be easily embedded in big data platforms. Instead of bringing the data to the analytics platform, it's the other way around. The predictive engine is embedded in the data store or other BI technology component. It's very easy to build on, with over 5,000 packages to extend functionality, and it has very strong data visualization capabilities built-in.
Corporate Performance Management (CPM) is a discipline closely related to Business Intelligence. BI is focused on gathering and processing disparate data and presenting it in an easy to digest form like a report, visualization or a dashboard. However, just reporting and displaying data is not linked to an organization's strategy. It does not include any mechanisms for planning, controlling or managing towards business objectives or KPIs. CPM is all about leveraging the data provided by BI to guide the organization towards its objectives. The KPIs and scorecards that are the end point of BI systems are the starting point of CPM, which links those metrics to the strategic goals of the organization. This is a process-oriented set of tasks that involves financial activities like budgeting, planning and forecasting. Other key capabilities of CPM products are profitability modeling, financial consolidation, and statutory and financial reporting.
These domains are highly complementary and there are already a number of hybrid tools with combined BI and CPM functionality, such as:
- Board International
- Host Analytics
Pure-play CPM tools really fall into a separate category and are not covered in this guide. The major enterprise software vendors such as Oracle, SAP, SAS and IBM all sell CPM suites, many of which, notably IBM's TM1, are integrated into their existing BI offerings.
Adaptive Insights and Anaplan are two powerhouse pure-play multi-tenant SaaS CPM vendors, but Adaptive has now broadened its scope by adding business intelligence capabilities with the acquisition of BI vendor MyDials in 2012. MyDials is a cloud-based data visualization tool, and this acquisition allows Adaptive to provide self-service data discovery and visualization tools to their customers, who are usually finance executives and analysts. The company changed its name from Adaptive Planning to Adaptive Insights to reflect this broader range of offerings, and their BI tool has been re-named Adaptive Discovery.
Adaptive Planning is highly rated on TrustRadius with 19 reviews and a high “Likelihood to Recommend” rating of 9.4.
Anaplan is a true cloud-based CPM platform for large enterprises and competes directly with IBM, SAP and other enterprise CPM platforms.
The topic of Big Data has produced much ink and discussion over the last couple of years, but it's not all talk and hype. The big data explosion has resulted in some impressive new technology from emerging companies, and also from the established BI vendors determined not to be eclipsed. It's still in its early days, but many major organizations like UPS, Morgan Stanley and Amazon have invested heavily in this new technology and have achieved excellent results already.
We now routinely collect huge volumes of data as a direct result of Internet and information technologies that have emerged over the last few years. The problem that Big Data technology vendors are trying to solve is how to actually use this data to improve business outcomes. Terabytes of digital information are collected from actual physical devices like sensors and machines, along with human-sourced communications like text, images and videos. Most existing BI systems cannot easily comprehend this kind of data, as they have been designed to make sense of highly structured data organized in tables and stored in a data warehouse. That leaves a vast quantity of potentially very useful data out in the cold. This is the driver behind the rapid ascension of the Hadoop and noSQL data stores like MongoDB and Cassandra, and the constellation of products that have developed around them.
Hadoop is a very unusual kind of open-source data store from Apache. The whole idea of Hadoop is that data is spread across many commodity, inexpensive servers, although there are several commercial distributions of Hadoop from Cloudera and Hortonworks who wrap services around the technology.
Unlike a traditional database, Hadoop can handle huge volumes of both structured and unstructured data including log files, streaming data, images, audio and video files. All of this data can be put into the Hadoop cluster and accessed, modified and processed in place, eliminating the need to duplicate and structure data in a traditional warehouse.
Once this huge volume of structured and unstructured data has been stored, how do you extract any value from it? Since Hadoop is not a structured database, structured query languages like SQL do not work. But Hadoop has its own data processing and query framework called MapReduce. Developers can use MapReduce to write programs that can retrieve whatever data is needed. However, MapReduce has several constraints affecting performance and a newer product like Apache Spark provides an alternative distributed computing framework, which is significantly more efficient. Similarly, products like Hive and Cloudera Impala provide a SQL-like query language, which is much easier for data analysts to learn and use.
Big data has moved far beyond the theoretical, and is now a reality within many major corporations. For example, UPS spends more than $1 billion a year gathering massive volumes of data from its truck fleet to discover the best delivery routes. Morgan Stanley no longer does portfolio analysis on traditional SQL databases, but instead uses Hadoop to analyze investments on a larger scale and with better results. Amazon uses a very large number of Hadoop clusters to run their business, including supporting their affiliate network, risk management efforts and website updates.
Today many of the products listed in this guide are scrambling to achieve interoperability with the Hadoop environment. An entire set of big data tools has emerged to simplify access to data stored in Hadoop or make it more SQL-friendly. Datameer and Alteryx are two products reviewed on TrustRadius that fit into this category.
Datameer is a product that provides user-friendly, Excel-like analytics on top of Hadoop and masks some of the complexity of the MapReduce paradigm from users. The value proposition of this product is that you don't have to be a data scientist to use it, as it overcomes Hadoop's complexity by providing a GUI interface with around 200 pre-built functions for analytics and data visualization. This product has mixed reviews on TrustRadius with a “Likelihood to Recommend” score of 7.1.
Alteryx is designed to make it easy to blend data from Hadoop and noSQL big data stores with more traditional, structured data in SQL-based data warehouses and spreadsheets. It is a self-service product designed for business users but—unusually—focuses on back-end data blending and data modeling; it looks much like an ETL layer except that it does not require any programming. Alteryx uses a drag-and-drop metaphor allowing business users to blend data from multiple sources before staging it for visualization in another tool like Tableau or QlikView. In fact, Alteryx has partnerships with both companies. The platform is actually built on top of the R statistical programming language and offers powerful predictive capability as an additional component. Alteryx is rated very highly on TrustRadius, with 14 reviews and “Likelihood to Recommend” score of 9.1.