QlikView vs Pentaho

[#1]

Hi, 

Can you explain difference and details of Qlikview vs Pentaho? 

Steven

Regards

[salto]

Hi,

Here my two cents:

• Pentaho is a combination of different Open Source products -  QlikView has everything integrated.
• Pentaho works with cubes- not too flexible when it comes to add new data sources. QlikView works with AQL (Associative Query Language). QlikView allows to integrate new data sources easy and quickly.
• Pentaho has limitations when it comes to add more dimensions, because of OLAP limitations.
• QlikView licensing can be more expensive. Both solutions have their free personal editions but the professional edition is not free not in Qlik, not in Pentaho.
• Usually, the implementation timing is lower in Qlik.
• The Gartner 2012 about Business Intelligence: QlikView leader, Pentaho in the opposite.
• The learning curve is lower in Qlik..
• QlikView loads the data in memory and does not query the server once the application is open.

Hope this helps.

[#3]

Hi!

You were very kind and helpful.

One last thing. I ask you another favor.

You could liken the two suites under these parameters?

1. facility (about its use)

2. speed

3. cost (licence + cost of minimum electronic requirements: server, memory etc)

4. applicability (to Mac, windows ecc)

5. mobile (capacity to function on tablet, smartphone..)

6. documentation available (about its features,functioning...)

7. after-sales support to customers

8 graphics (dashbords...)

thank you very very much

best regards

[salto]

Hello Stefano,

I am sorry but I do not have the knowledge to make a honest comparison in those features between both platforms. I do not have the needed hands-on experience in Pentaho in the same conditions for QlikView to make a honest valuation, so I prefer not to do it.

Thanks for your understanding.

[Michael_Tarallo · Accepted solution]

Hello Stefano,

I worked for Pentaho for a little over 4 years. I was there from 2007 to 2011.  I will provide some high-level, objective and impartial bullets. Please note that some of the names / ownership / positioning etc. might have changed since I left but software-wise things are still pretty much the same.

First I'll describe the difference and uniqueness of QlikView - I won't get into too much detail about the architectures and development etc. but will provide links out for you to follow so you can see it in action.

QlikView

Simply stated and in all fairness - you cannot really compare the two products 1:1 or have them share the same category. QlikView provides a "Business Discovery" platform (Qlik has been credited by Gartner for coining the term) - what Gartner refers to as Data Discovery. Though they may be seen together in the BI Gartner Magic Quadrant, they fundamentally work differently. In fact it has been stated by Gartner that: "By 2015, the majority of BI vendors will make data discovery their prime BI platform offering, shifting BI emphasis from reporting-centric to analysis-centric."

QlikView's unique capability is in the associative engine that relates data in-memory and allows for a more natural way of exploring data and making decisions; providing unmatched performance and the ability to ask that next unprecedented question. It works the way the human mind works, associatively. Able to visually associate raw data and derive information from its "surroundings".

During analysis and interaction with QlikView, it does not pass SQL to an underlying RDBMS or use a traditional BI semantic layer - therefore making it very performant and non-dependent on the source data system. However, QlikView can perform direct analysis on the source if needed with our Direct Discovery feature. Direct Discovery is usually for specific use cases - used with Big Data (SAP Hana, Teradata, Google BigQuery) or when real-time data information is needed. It offers a hybrid data model combining both in-memory and direct data access. You can learn more about that here: http://www.qlik.com/us/explore/products/big-data

Initially, extracting and combing data from single and multiple sources is for simply loading the QlikView in-memory data model. Which is then used with QlikView interface(s) / tools to create dashboards and explore data.  The power is in the QlikView engine, the development process, its associative data model, its performance factor and its interface(s) that users can easily understand.

With QlikView, data is automatically associated in-memory and calculated and aggregated on the fly as a user interacts with the visual objects and filters. The power of "green, white and gray" - in our selection scheme - allows uses to explore ALL relevant data freely in one place, without the need to set up predefined paths, parameters and links as you may do with traditional BI tools - or - without the process of creating / managing a semantic layer - so rapid prototyping and deployment can be easily achieved and drilling to detail is automatic.

You can see the data you've selected (green), the data that is relevant to other selections (white - linked data) as well as the data that is not associated (gray - non-linked data) - all with in one view.

Figure 1: Power of green, white and gray

This selection scheme and interaction is both in the QlikView desktop client or in the web based interface when the application is published to the web (AccessPoint).

Its overall purpose allows one to answer that very important next question, the one that was not anticipated - hence the term Business Discovery. For example if I was a sales manager and I wanted to see sales, products and location in a given category - after simply selecting the category - not only would I see the current sales figures in the chart, for that selected category (green) - but I would also see the values for other dimensions (such as products and locations) related to the sales in that area (white) as well as those not related (gray). So if I discovered that widgets x,y,z were highlighted in gray - that would prompt me to ask the next question such as "Why are widget x,y,z not selling?"- even though I initially set out to find out the sales figures in a particular category. All this occurs in the model automatically, there isn't any predefined setup.

Your analysis path occurs on the SAME QlikView screen (Dashboard) and you CAN START ANYWHERE in the data.

Figure 2: A simple sample QlikView Selection Scheme  with multiple list boxes

For QlikView developers, QlikView Desktop uses an editing interface with scripting syntax that is easily inserted and supplemented by push button dialogs, wizards, intelli-sense and auto-completion. The script is initially used, during development, to access, extract, transform and load data into the QlikView in-memory data model. Beginners can still get started quickly without writing code or a data warehouse. If you have a data warehouse / store / mart etc. and it has everything you need and the data is perfect (which is not always the case) then simply loading that data to the QlikView application is very easy. There are also wizards for Excel, Inline data, HTML and web based data if needed. The scripting interface is very powerful and allows one to perform transforms and other data storage and enrichment functions. This is no different than any mature BI tool that needs to use SQL, functions or an API to perform specific tasks on data. In fact the script allows you to handle imperfect data and apply "other things" that are needed for analysis - including the use of variables and extending the product's capabilities. And to be honest - it is not difficult to learn or use either.

Figure 3: Wizard controls for connections, functions, variables and data

Figure 4: Wizard for data selection and transform

I have heard some customers say:

"What I can do in QlikView in 15 minutes - I can do in 5 minutes in another tool - BUT what I can do in QlikView - I CAN'T EVEN DO in the OTHER tool." QlikView is also seen as a powerful application development framework. It all depends on the solution and business use case where the use of extensive scripting is needed.

You can learn more about QlikView and its architecture here:  http://www.qlik.com/us//~/media/Files/resource-library/global-us/direct/datasheets/DS-Technical-Brie...

Watch some Quick Start Videos here:  http://www.youtube.com/watch?v=aKQMhtQnaz0&list=PLW1uf5CQ_gSoRmiPsKg0o7DXIzm3CTOxA

Qlik - YouTube - other videos

Try the demo site here: Qlik Demos: See QlikView in Action | Demo.Qlik.Com

There is also an extensive library of plug-ins, connectors and solutions on our market place QlikMarket

Pentaho

Pentaho is a commercial open source BI vendor. It offers a traditional Open Source BI software stack that can be compared to those vendors in the traditional proprietary BI space.  It is 100% Java so will work on most platforms that support Java. Pentaho products and components (community projects - either owned, managed or sponsored by Pentaho) are professionally managed by Pentaho team members. The software stack consists of a combination and variety of different desktop and web based tools.

These projects / products include:

• Pentaho Data Integration - (ETL)  Kettle Project - one of the most powerful ETL tools I have used. The community edition of PDI - was the most full featured, well thought-out product I have had the pleasure of working with. There are also interfaces in the same tool to create metadata for the analysis and reporting tools and to view your data. (model editor)
• Pentaho Reporting - (Pixel perfect, structured reports) formerly JFreeReport
• Pentaho Platform - original project platform - execution layer / web user interfaces / Dashboard (Flash based) Ad-hoc web based reporting / administration / metadata / interactive reporting / scheduling and distribution
• Pentaho Analysis - uses the Mondrian Project OLAP engine - and analyzer front end
• Pentaho Data Mining - Weka Project
• Community Development Framework (CDF) - APIs, utilities for building and embedding dashboards and interactions into your own applications.

Some of these have been re-branded under a new name: Pentaho Business Analytics.

There is an open source Community version without any "Enterprise" related features. "Enterprise" versions of the software would most likely include support, extensive Q&A, bug fixes as well as security, code and license compliance testing; for a subscription based fee. "Enterprise" features may vary - I am not too sure what is and is not considered "Enterprise" - I know that the visual interface that works on the Mondrian OLAP engine might be one of them and that they have added a bunch of visualizations to enhance that part of the product. The products and features will vary depending on what you need. Again, they range from ETL to Data Mining and have done some significant work in the Big Data space and have been or should be recognized for their accomplishments in this area.

• Pentaho primarily works under the covers by sending SQL or possibly a native dialect (depending on the source) to the underlying database or source using an appropriate connector at run time during analysis. It also be able to insert and pass parameters into the queries that will run on the database or source. Some pieces use a caching mechanism to speed up queries and can have the cache primed to improve performance.
  
  
• The entire Pentaho Platform has the ability to access a tremendous variety of both structured and unstructured data through SQL, its own metadata layers, APIs, its own ETL products, NoSQL, Big Data, web etc. from its Reporting and Analysis products.
  
  
• Pentaho Reporting can use access a variety of data both SQL, Metadata and API driven. Reporting Metadata requires the creation of a metadata model with another tool. - Metadata Editor - or a simple metadata model can be created in the Pentaho Data Integration model editor - this will create both metadata layers for the reporting and analysis products.You can create web based ad hoc and desktop based pixel perfect reports that can be published and distributed.
  
  
• Pentaho Analysis (Analyzer)- requires a separate OLAP schema created with another tool. - Schema Workbench - or a simple schema can be created in the Pentaho Data Integration model editor - this will create both Metadata layers for the reporting and analysis products.
  
  
• Pentaho Dashboard - can use previously created analyzer reports / charts / maps, interactive reports, pentaho reports, and create inline flash charts and tables with an ad hoc query wizard, all with parameter and filter support dialogs to provide a holistic interactive view of data.
  
  
• Pentaho Data Mining - is a separate product - I do not know much about it nor worked with it that much while I was there.

You can download it here: Pentaho Community

You can get a trial of the Enterprise Version here: Download Pentaho Business Analytics & Data Integration

• Pentaho: Demo Big Data Analytics & Integration | Test Drive Pentaho

I hope this is helpful and gives you a better understanding of the products.

With all due respect please perform your due diligence and research and follow the appropriate evaluation processes in order to conclude what product is right for you. It will all depend on the business need and the type of solution you are after. I believe I described both products fairly and to the best of my knowledge. If there is anything that is not correct I will make sure to edit where appropriate.

Kind Regards,

Michael Tarallo

Senior Technical Product Marketing Manager

QlikView and QlikView Expressor

@mtarallo - Twitter