Thursday, May 28, 2015

QlikView vs Tableau

Spent a few days playing around with QlikView. Here are my initial thoughts on how it stacks against Tableau.

Cool factor: Tableau charts feel more slick and modern compared to QlikView. The options provided are more focussed, and Tableau even suggests optimal visualisations based on data selection. The ability to share an interactive dashboard with users is common between the two. 
Overall winner: Tableau

Usability: Tableau seems to have done a better homework on Usability and visualisation. QlikView can do as much, but it is not as business user friendly: the sheer number of options in QlikView feel over-whelming to start with for a non-technical user. Consequently, you would spend more on training on QlikView and would need a more technical minded person to be able to build dashboards with it. 
Overall winner: Tableau

Ease of Data Transformation/Extraction/Access: Both need data in database format. However, at the first look it is not immediately obvious whether QlikView can integrate with the same range of data sources as Tableau. Business users would need a data analyst/ technical person to extract, transform, and load data in both, but more so in Qlikview. Tableau provides access to OLAP cubes while QlickView doesn't. 
Overall winner: Tableau

Raw Power: In terms of technical power, QlikView appears more technical and customisable than Tableau for a power user familiar with databases and SQL.
Overall winner: QlikView 

Pricing: Oveall complex pricing plans, but I understand that Tableau may be overall cheaper.
Overall winner: Tableau

Ease of exploration & training: QlikView provides an unlimited personal use edition and Tableau provides free public desktop application. There is good support for training.However, QlikView appears to have a more well defined learning path than Tableau. For example, I could find a well-defined certification track for QlikView, but not for Tableau (as on 28 May 2015).
Overall winnerQlikView

Summary: If you have a team who knows how to build visualisations and is reasonably strong technically, go for QlikView for the sheer power of the tool. You will have to spend some time training for it. If you are in a hurry and want to get to a good looking solution fast, Tableau is the answer. Personally, I am inclinded to overall favour Tableau of the two for its ease of use for the business user. 

The above review was done as a first time business user based on initial impressions by trialing the products over 14 days.

Friday, May 22, 2015

14 Days of Tableau

I recently trialed Tableau, a data visualisations tool used for Analytics. My initial impression as a novice user (based on loading data from an excel and doing a few online trainings) is that it is a fantastic tool for following scenarios:
  • We have database acceess, or want to analyse data coming directly off a known client system (SAP, Salesforce)
  • The data available is structured inline with database tables
  • We need to monitor KPIs over a period of time
  • To extract regular insights from predictable management information
  • It is a dedicated tool with sole purpose of getting insights out of data, and is therefore more powerful than excel for charting/ data visualisation
  • The ability to connect with and combine varied data sources makes it a class apart for large data sets that would simply kill excel
  • The ability to build graphical, interactive storyboards that can be shared online take analysis and engagement to a new level
  • Very strong for the four scenarios identified above, especially for building dashboards that refresh based on an underlying, changing data
  • Needs data in a database format. This may need extra effort and planning. E.g. Converting a simple monthly Management Information for balance sheet to a database friendly format can be quiet a task. This would especially hurt if the model changes often, or the number of rows is very large.
  • Missing data points and incomplete excel data can be tricky to deal with
  • Without training, too many options for data cutting can be confusing
  • May not always be worth the effort as there may not be enough time or resources for it (vs doing a simpler excel based analysis)

After this, I intend to trial Qlikview (another BI tool) and compare. Will keep posting my thoughts.

Wednesday, May 20, 2015

Making Sense of Data

With the ever improving ability of companies to track and store data, it is only natural that business intelligence (BI) tools have gained increasing importance in C-suites. In fact, the market is glutted with umpteen BI tools, vying for this extremely lucrative space. Needless to say, it has been an equally lucrative opportunity for consulting companies as well.  

However, the businesses need to be very careful. Data analysis can often be the right way to reach a wrong conclusion. To truly unlock value from mountains of data, they would need multi-disciplinary teams that understand the importance of the task and are willing to work together. The same is true for consulting firms that want to cash-in on the desire of businesses to use their data more effectively.

From business perspective, there are four key challenges on the road to using data effectively:
  1. Capturing and organising data
  2. Making the data useful and accessible for the business user
  3. Ensuring that data analysis is statistically significant
  4. Making the whole process timely and cost effective 
Based on these, logically it would help if the team has the following profiles:
  • business owner to fund the project & articulate the answers sought from data
  • business analyst to bridge the gap between techology and business
  • trainer to teach business users about how to go about accessing/ manipulating the data
  • data analyst to help business owners to analyse data and decide if the data interpretation is actually significant statistically. Many modern BI tools can help business owner to manipulate data by himself, but he may still need support to extract, transform, and load data into the system. Given the technical nature of things, help may be needed to troubleshoot as well
  • business consultant with industry specific expertise to help business owners to refine the questions they seek answers to, and create processes to ensure data integrity
  • project manager to coordinate the team, budget, and execution

Needless to say, one person can play multiple roles.