Skip to main content
Announcements
Qlik Connect 2024! Seize endless possibilities! LEARN MORE
cancel
Showing results for 
Search instead for 
Did you mean: 
brett0123
Contributor
Contributor

Qlik and the Associative Model?

Hi all, recently I've been reading up on the latest versions of QlikView and Qlik Sense, and while I do appreciate a lot of the other features of competitor products, such as Tableau and PowerBI, it seems like this Associative Model has a number of interesting advantages.

In the past, it seem like Qlik wanted to offer a "guided" experience to BI, versus a more "self-service" experience by other competitors, and of course, this has been an area of development and debate for some time.

However, recently I found that SAP Lumira (formerly SAP Visual Intelligence), had adopted this Associative approach, and after looking into it a bit, started thinking that it might really be a best practice. Or, to try to be more precise, for large, complex, enterprise data sets, it's very hard to be fully "self-service", since you'd need certain guidance to have some knowledge/confidence that you're doing the right thing, looking at the right place, or having it the right way.

Anyone else feel that way? Or, do anyone feel that a standard relational model has advantages vs. this approach favored by Qlik & SAP?

Labels (1)
1 Solution

Accepted Solutions
Qlikstu
Partner - Contributor III
Partner - Contributor III

Hey Brett

This is a great question and I'm surprised it hasn't been answered yet, so having experience with "standard" data models and the Qlik Associative model I'm happy to give my opinion. Not everyone may agree but here's my 5 cents. 

Qlik's associative model, as you rightly mentioned, offers some unique advantages when it comes to self service. It allows users to explore data in a non-linear fashion and uncovers insights that may remain hidden in a hierarchical, standard relational model.  As you know, Qlik's model does not limit users to pre-aggregated data and pre-determined questions, which can empower users to ask unexpected questions and explore data from various angles. This dynamic approach can be particularly beneficial in dealing with large, complex enterprise datasets. I've always said that Qlik answers questions you never thought to ask - which is extremely powerful. 

However, in the interests of giving a balanced view, you could argue that from an end user point of view the associative experience could, at least initially, require a steeper learning curve for those end users accustomed to the tradional relational model. Without proper guidance or knowledge, end users may misinterpret data, overlook important nuances, or miss out on critical insights. This means the balance between guided analytics and self-service is often crucial, and this balance may vary depending on the organization's context, the users' data literacy levels, and the complexity of the data. 

Taking this to it's natural conclusion, standard relational models will provide a nice foundation as it's a traditional, well understood model and is a well established and understood way of organising data both from a self service and guided aspect. Qlik also does this if you want, but without the Associative element "baked in" other tools simply won't offer the same levels of flexibilty and interactivity that Qlik does. 

I always say best practice isn't a set of rules to blindly follow. So much of how you approach things (and which tools you use) depends on the specific requirements, data maturity, and strategic goals of the business. 

Hope this helps. 

Stu

 

 

View solution in original post

1 Reply
Qlikstu
Partner - Contributor III
Partner - Contributor III

Hey Brett

This is a great question and I'm surprised it hasn't been answered yet, so having experience with "standard" data models and the Qlik Associative model I'm happy to give my opinion. Not everyone may agree but here's my 5 cents. 

Qlik's associative model, as you rightly mentioned, offers some unique advantages when it comes to self service. It allows users to explore data in a non-linear fashion and uncovers insights that may remain hidden in a hierarchical, standard relational model.  As you know, Qlik's model does not limit users to pre-aggregated data and pre-determined questions, which can empower users to ask unexpected questions and explore data from various angles. This dynamic approach can be particularly beneficial in dealing with large, complex enterprise datasets. I've always said that Qlik answers questions you never thought to ask - which is extremely powerful. 

However, in the interests of giving a balanced view, you could argue that from an end user point of view the associative experience could, at least initially, require a steeper learning curve for those end users accustomed to the tradional relational model. Without proper guidance or knowledge, end users may misinterpret data, overlook important nuances, or miss out on critical insights. This means the balance between guided analytics and self-service is often crucial, and this balance may vary depending on the organization's context, the users' data literacy levels, and the complexity of the data. 

Taking this to it's natural conclusion, standard relational models will provide a nice foundation as it's a traditional, well understood model and is a well established and understood way of organising data both from a self service and guided aspect. Qlik also does this if you want, but without the Associative element "baked in" other tools simply won't offer the same levels of flexibilty and interactivity that Qlik does. 

I always say best practice isn't a set of rules to blindly follow. So much of how you approach things (and which tools you use) depends on the specific requirements, data maturity, and strategic goals of the business. 

Hope this helps. 

Stu