architecture of the system
Live Query Engine interprets abstract queries generated by VizQL into language that is understood by popular database systems, such as SQL and MDX syntax. Thus, increased data accessibility and usability of databases through a uniform user interface that interacts with a diverse range of databases, formats, and sizes. Live Query Engine allows users to query databases without having to first import the data, the query is instead interpreted and run by the database with only the results rendered. This technology provides for data consistency and avoids data movement while still being scalable, secure, and flexible. Further, it allows Tableau connects to open-source Hadoop databases, proprietary MapReduce technologies, and cloud data warehouses like Amazon Redshift and Google BigQuery. Column stores, databases designed to process unstructured data, and web applications such as Salesforce and Google Analytics are also able to be connected (Form 10-K, 2016).
Tableau’s In-Memory Data Engine further supports user’s ability to analyze large amounts of data independent of database systems. Much of the today’s data is not stored in databases or stored in databases that are too slow for interactive analysis, hence, the need for analysis outside of the database. The In-Memory Data Engine uses column-based storage and compressed representations of data while leveraging RAM-based indices to provide users with fast calculations without the complications, costs, and delays of a database system (Form 10-K, 2016).
Tableau has developed their own visual query language (VizQL) that translates drag-and-drop actions into data queries and then expresses the information visually. Queries and visualizations used to be separate tasks and the queries often required scripts, chart wizards, or dialogue boxes. The VizQL technology increases speed and flexibility, provides a creative and engaging experience, and brings a significant improvement in the ability to gain insights from data (Form 10-K, 2016).
The VizQL, Live Query Engine and In-Memory Data Engine work harmoniously to form Tableau’s Hybrid Data Architecture, allowing users to fully exploit flexibility and power without any programming or scripting. Flexibility to access and analyze data from a range of sources while optimizing speed and performance for each source is the core of the hybrid strategy. Customers are able to integrate live data with in-memory data on a single visualization or dashboard through these amazing technologies. The In-Memory Data Engine could be used to import a data sample from a large database in order to ask a question from a visualization. This visualization can then be queried against the entire database using the Live Query Engine to answer another question or find a new pattern/trend (Form 10-K, 2016). The breadth of usability is truly magnificent and is what makes Tableau stand out from its competitors.
Tableau suite can interact over 40 data sources (Form 10-K, 2016), on premise or in the cloud (Tableau, 2017) to including those from the top five database vendors.
Tableau is very short-sited in overall scope; however, it is extremely advanced within its limited scope. The software itself does not offer any ETL technologies nor does it consolidate, clean, transform, or reduce data. Tableau imports already cleaned, consolidated, transformed, and reduced data to ask a question and create a visualization from it. This visualization can then be used for another query; however, the platform focuses more on the visualization of the data rather than the actual mining of it. It prides itself on such seamless integration with other BI tools in order to supplement fully without having to provide all of the other functions.
data mining methodologies
Tableau does offer data mining through classification, clustering, and association rules within the drag-and-drop interface.
coupling with database or data warehouse systems
Tableau’s hybrid architecture allows the software to run outside of the database. See ‘architecture of the system’ above for a more in-depth explanation.
Tableau suite can be dialed into the perfect combination of user flexibility and control. Existing security protocols can be seamlessly integrated to provide central governance of metadata and security rules. User and group level authentication options are available as well as pass-through data connection permissions and row-level filtering (Tableau, 2017).
Tableau has incredibly strong live visual analytics that allow users unrestricted data exploration capabilities. The drag-and-drop interface provides the ability to use reference lines, forecasts, and statistical summaries to tell a visual story through trend analysis, regressions, and correlations. This method of storytelling appeals to the psychological aspect of learning and calling for action. Users are able to capture emotion and logic, taking the viewers on a journey through the data. Viewers are more likely to digest and retain the dat. Further, viewers are more likely to identify with the data, which drives change. Static slides and boring presentations are no longer relevant or captivating (Tableau, 2017).
Tableau has strengthened its portfolio with a new, free application, Vizable, that turns data into interactive graphs that can be shared from an iPad and explored on the go without the need for a server or any cloud-based services. The technology queries data, aggregates, and generates a visualization on the tablet within seconds. The exciting interface uses hand gestures such as dragging, swiping, and pinching to receive instant feedback.
graphical user interfaces.
Insights can be embedded into workflows for employees, customers, partners, and suppliers to provide analytics anywhere needed. Interactive dashboards can be embedded into existing business portals including applications like Salesforce, SharePoint, and Jive. Users are able to switch between extracts and live connections to data with just one click, or schedule automatic extractions. Team members can securely access published dashboards from any mobile device or external browser (Tableau, 2017).
Can you propose one improvement to such a system and outline how to realize it?
It appears the industry-wide recognized weakness of the Tableau system is its inability to load data for preparation before use. In the beginning of my research, I thought this should be improved upon and could be realistically strategized. However, after digging deeper, this seems to be a characteristic that Tableau prides themselves on and leverages to provide flexibility and efficiency to their customers. Tableau has developed a hybrid architecture to fully emphasize the advantages of this approach.
Tableau focuses on visualization of data, while others feel this may limit them, they are interested on developing new technologies and features for visualization rather than expanding the functionality. They are pioneers in their field of expertise, they know what they are good at and they are sticking with it. There is nothing wrong with this approach, it is just viewed as lacking by many who try to be ‘do-all’ technologies.
While I can appreciate Tableau’s approach, they should remain guarded as others are quickly implementing new technologies to match their level of visualization capability, they may need to consider expanding their portfolio.
Read the company annual report (or 10K) and give an overview of the company, their competitors, their customers, their products and overall strategy. You do not need to include any financial analysis; however, you are strongly encouraged to evaluate the performance of Tableau and its overall direction financially for your own benefit.
Tableau version 9.0 is currently available in 8 languages with over 39,000 customers in over 150 countries. This statement alone is a testament to the mission of the company, help people see and understand their data. Distribution strategy is designed to capitalize on the ease of use, low up-front investment, and collaborate facets of the software usually evolving from a free trial to different departments and potentially to an enterprise level. Total revenues have increased to $653.6 million from $412.6 million in 2014. Tableau is committed to constantly innovating and advancing, they spent $204.1 million in R&D for 2015.
Tableau cites its primary competitors into three categories; large technology companies (IBM, Oracle, Micorsoft), business analytics software companies (Qlik, MicroStrategy, Spotfire), and SaaS-based products or cloud-based analytics providers. Tableau expects competition to increase and realizes many of their competitors outweigh in resources and history, further understanding this could lead to a loss in market share or price cuts. Beyond relentless development, Tableau further recognizes their weaknesses and other uncontrollable factors that could impair or diminish success. It is understood that there is a fine balance of development and retaining revenue for success that must be juggled going forward. Tableau currently has 16 issued U.S. patents and 35 pending patent applications (Form 10-K, 2016).
competitive analysis of Tableau and two of their competitors
These two products are both well-known in the BI software market but they are distinct in the markets they target. Tableau is a leader visualization tool with its drag and drop modern interface. Users of all levels can create meaningful dashboards and reports. Cognos obviously uses visualizations, however, providing a complete, enterprise level BI platform is their focus. I think this is the true discrimination between the two. Cognos is excellent for multidimensional and relational data sources that can be used by experts to improve strategy and monitor performance. However, this complexity the product is valued for also makes it difficult for all levels of users to access the insight they need. So, while this is Tableau’s strength, Tableau is fragile in terms of integrating data from different sources in preparation for analysis. Data preparation would instead be a strength of Cognos (Scavicchio, 2016).
Spotfire BI solutions parallels Tableau’s goal of allowing users to quickly visualize data from various sources, however, their approach is unique. Spotfire requires a more advanced user to make predictions with data whereas Tableau allows less advanced users to drill down into data without statistical analysis. Spotfire can be troublesome when attempting to customize visualizations and drilling down to specific data details. Spotfire is recommended for companies looking to improve sales, marketing, and customer experience (Spotfire, 2016).
Like Tableau, QlikView emphasizes data visualization and analytics with easy to use GUI and the ability to integrate data from a plethora of data sources. However, QlikView also encompasses other BI tools like QlikView Expressor (a metadata intelligence solution) and NPrinting (report generation, scheduling, and distribution). Users say the interface is clean, easy to understand, and easily integrates with Excel. Therefore, this solution is effective at an enterprise level where different features can be utilized in different departments. Other features may be stronger than Tableau’s parallel such as good third-party integration, advanced data filtering options, and data manipulation. QlikView can be difficult to learn and operate because of its many facets and intricacies, but Tableau also comes with a learning curve. Data management and mapping can require IT assistance with QlikView, visually appealing reports can be difficult to create, and hardware can be extremely costly. As previously noted, Tableau does not offer ETL capabilities which is a huge shortcoming. QlikView is also able to integrate more data sources than Tableau (Foley, 2015).
The conversation is not one of which is best, rather that of what the end goal is. This ties into last week’s conversation surrounding Data Scientists versus Data Analysts. If you are going to have data analysts and end users gathering what they need from the data available, Tableau is an excellent option. However, if you have an enterprise with a data scientist, you likely will choose a solution that allows an expert to use statistics for predictive and prescriptive analytics. Drilling down into the data available will no longer suffice, more information of the data will be needed to find new data sets for users to query.
Foley, A. (2015). QlikView Vs. Tableau: Software Showdown. ClearPoint Strategy. Retreived from https://www.clearpointstrategy.com/qlikview-vs-tableau/
Form 10-K. (2016). Tableau Software, Inc. United States Securities and Exchange Commission. Retrieved from http://d1lge852tjjqow.cloudfront.net/CIK-0001303652/893d1eb0-642d-4226-b2ff-853d712155e6.pdf
Scavicchio, J. (2016). Tableau vs. IBM Cognos: Compare Key Features and Functionality. BetterBuys. Retreived from https://www.betterbuys.com/bi/tableau-vs-ibm-cognos-differences/
Scavicchio, J. (2016). Tableau vs. Spotfire: Price and Feature Comparison. BetterBuys. Retreived from https://www.betterbuys.com/bi/tableau-vs-spotfire/
Tableau. (2017). Business Intelligence and Analytics. Retrieved from https://www.tableau.com