QlikView vs Tableau: What's the Best Choice for a Business Intelligence Suite in 2019?

Some have theorized that this shake-up of an over-populated market was the inevitable result of the success of Microsoft's Power BI suite. Gartner has consistently rated Microsoft's analytics offering as the market leader in its respective Magic Quadrant since the product's official launch in 2015. Therefore, it's possible that the 'independent' players in the BI arena needed to scale up in order to stay viable against such a committed and influential industry player.

Others believe that the consolidation is the result of convergence, as the heightened interest in enterprise-level big data applications coincides with the current breakneck pace of development in machine learning and artificial intelligence—promising more powerful insights to whichever suite can best leverage these technologies in the most approachable, performant and economically viable way.

In any case, since the BI space is changing so much, we thought to take a fresh look at Tableau vs QlikView, the two suites which have been considered 'standard' choices over the last 15-25 years.

In 2014, Qlik finally decided to address many of the product's shortcomings, relative to Tableau's glossy and integrated solution, by launching a parallel product to its venerable desktop client. The cloud-based QlikSense runs over the same data architecture as the desktop client QlikView, but matches Tableau's focus on transparent data discovery, dazzling and innovative visualizations, easy view creation, and full compatibility with mobile devices.

QlikSense was never intended as a successor to QlikView. It was conceived as an equivalent product for companies looking to analyze and discover existing data rather than invent complex and challenging new data-flow structures for reporting and analytics. For those requirements, QlikView remains an able and powerful resource. If anything, QlikView's more accessible tooling could be argued to suit better to today's emerging machine learning and big data developer environments, where GUIs and interconnections are often explicit, rudimentary, or experimental.

However, Qlik's product-base splits Tableau's equivalent capabilities over two products, which is an additional consideration in choosing between the contenders. Nonetheless, since we need to compare the most essential and versatile aspects of each offering, we will consider Qlik's original and more configurable product in relation to Tableau.

Qlik is attempting to transition to platform-agnostic network environments via its more modern (but restricted) QlikSense product. Therefore, it foregoes the extra development effort needed to support an OSX or Linux client. Currently QlikView supports 64-bit versions of Microsoft Windows, including 7, 8.1, 10, Server 2008 R2, Server 2012, Server 2012 R2, Server 2016, and Server 2019.

QlikView's local install requires 4 GB of RAM and 300 MB of disk space.

Though virtualization is not banned for QlikView, it's neither directly supported nor encouraged.

Tableau offers desktop clients for 64-bit Microsoft Windows from version7 upwards, as well as Microsoft Server 2008 R2 or newer. Additionally, there is an OSX client, which requires a minimum of OSX 10.13 running on an iMac or MacBook from 2009 or later.

The company specifies a minimum 2 GB RAM for a Windows install, and 1.5 GB of disk space across both platforms.

Official support is provided for a number of virtual environments, including Parallels, Citrix, Microsoft Hyper-V and Azure, Amazon EC2 and VMware.

QlikView provides a dedicated iOS client that can also work offline if used on an iPad (with certain pre-conditions). For general mobile availability, Qlik has developed an Ajax-driven framework that makes accessing your information via a compatible mobile browser the best current solution. The mobile client is actually just a dedicated browser-wrapper.

After discontinuing its early Android mobile client some years ago, QlikView is currently piloting a new beta for Google's OS. Apparently, it is also a browser wrapper for an HTML5-oriented view of user data, drawn through QlikView's AccessPoint document listing service.

Tableau provides dedicated iOS and Android apps. The apps can access data either via Tableau Server or the cloud-based Tableau Online platform. Layout views can be configured directly for end-user mobile devices.

All the dedicated apps listed above can be controlled and configured via Mobile Device Management (MDM) solutions.

The QlikView interface retains the retro stylings of Windows XP, circa 2001, bucking the modern trend for minimalist user interfaces and context-based GUI practices. Most of its dialogues are modal pop-ups, and its creation processes rather linear.

Whilst lacking the unfortunate and dated stylings of QlikView, Tableau's own interface is similarly menu-dependent.

Tableau's workflow is less cluttered, and more contextual and responsive. Its increased emphasis on drag-and-drop functionality can make for a flatter learning curve in respect to QlikView.

However, we should evaluate BI software less superficially, in the light of 'back-office' functionality intended to support useful and clearly-presented data insights. The 'end users' are frequently those that we are trying to persuade, rather than those who are building the processes that make the data explicable.

QlikView's Associative Selection model, which provides contextual feedback in the interface based on user input, is generally praised as one example of how the software's 'in-memory' technology can speed up interactive performance, compared to Tableau.

However, this touches on differences between the two products which are conceptual in nature: QlikView was conceived as an architectural toolkit for data visualization, and Tableau as a data discovery framework, which necessarily must trade off some flexibility against latency. Indeed, concerns around 'ease of discoverability' are one of the core challenges that Qlik has chosen to address in the QlikSense platform, rather than in its older desktop client.

Therefore, the speed you're enjoying via QlikView's associative filters has to an extent been paid for upfront in design/development time, or at least in load time. In one sense, it's a 'compiled' feature with a measurable cost in terms of implementation and inflexibility.

The other factor fueling QlikView's responsiveness is its increased RAM requirements (relative to Tableau) as the data increases, along with the number of joins and relationships necessary to develop and visualize insights.

In its Windows desktop client, Tableau can natively utilize multidimensional or OLAP data sources, known as cubes, to speed up data exploration. QlikView can access cubes via the SAP OLAP Connector. Cubes are pre-compiled data groupings in which decisions about the available granularity of the information have already been made in order to increase performance. This pre-mapped approach can affect the user's ability to explore the data ad hoc.

There is no apparent prospect of OSX support for cubes in Tableau at the time of writing.

In line with the cryptic standards and practices of enterprise licensing, you'll need to build a detailed and very well-researched use case to understand what the costs are likely to be for either offering, as well as which products you're going to need, and (in some cases) the number of required seats per product.

Qlik's products are already fragmented between QlikView and QlikSense, and between them there are many deeper layers of pricing granularity depending on features required, the number of users, and the type of license. Licensing agreements generally range from 1 to 3 years' commitment.

However, the QlikView Publisher, Information Access Server, and Extranet server all come with their own range of price-points, depending on your needs. The recently implemented Unified License system can help users cherry-pick the strengths of QlikView and QlikSense. You might need professional advice or direct consultation with the company or a reseller.

QlikView has a free personal edition for recent Windows X64 operating systems. However, this can only open documents created by the user's own version of the software, as well as educational files that QlikView provides in its help section. Therefore, it may be of limited use for evaluating QlikView's broader capabilities.

Tableau switched from a perpetual licensing model to a SaaS model in 2017.

Tableau does offer a more leisurely stroll through the suite's functionality in the form of Tableau Public. As the name suggests, any data you run through the system will be publicly available, and data is limited to 15,000,000 rows per workbook.