QlikView’s associative model is a fantastic design that enabled the Green/White/Grey feature. This feature greatly helped us highlight the relationship between different kinds of data.
To further explain the Green/White/Grey feature, let’s take a sales report of an electronics shop as an example. In this report, it may have dimensions (Product Category, Salesman) and expressions (Sales Qty, Sales Amount).
Green means the values of dimension that you have selected. Under the Product Category dimension, it will have different values e.g. Smartphone, Laptop, Desktop. If we want to see the sales performance of Smartphone, we can click on Smartphone, then it will become green. We can also select multiple values e.g. Smartphone and Laptop in the dimension.
White – shows the values of dimension which are related to the another dimension values that you have selected.
Grey - shows the values of dimension which are NOT related to the another dimension values that you have selected. By following the previous example, we have selected Smartphone in Product Category dimension. In the Salesman dimension, Salesman e.g. Jeremy or Janice who have have sold Smartphone will be shown in white color. Another Salesman, e.g. Roy, who didn’t sell Smartphone will be shown as Grey. This is very useful for us to identify relationship between dimensions
Its “Search Object” function is also very useful for us to search for the information we needed.
“Section Access” is another great feature as we can identify user authority on accessing data in an easy way.
Another feature "Set Analysis" which enabled us to configure different behavior for each dashboard. So that some dashboards (e.g. YTD) can be freezed and not affected by some of the filter selections.
My experience using QlikView was in a F&B chain store with outlets around the world. In such a F&B chain store, there are a lot of dishes and they are changing frequently with a high transaction volume.
In such an environment, a search function is very helpful for us when we want to find out the sales performance and trend of a specific dish.
Qlik’s associative model greatly helped us identify customer purchase behavior and we can define combos according to this information.
By comparing with the latest data discovery tools like Tableau, Qlik Sense and Sisense, the layout and colors are not eye catching enough. It looks like a very powerful excel with a lot of powerful features for analysis. But it may not be good for presentation.
Qlik is able to import data directly from a transactional database and create joins between the tables in there. However, I encountered a performance issue with high volumes of data.
Therefore, I suggest designing a data warehouse and use it instead as Qlik’s data source. The performance will be much better.
Actually, I can’t remember any issue that I was required to check with customer service or technical support. For nearly all of my questions, I can get answers in the Qlik community.
I have previous experience with IBM Cognos. It was a very comprehensive solution that can be used for pixel-perfect reports, email schedules and also data analysis. However, there are many more development and implementation steps than with Qlik.
Also, Cognos requires a steep learning curve even for a user with a technical background.
As we wanted an easy and user-friendly self-service BI solution, QlikView is more suitable from this perspective.
On the server side, it is quite straightforward and we didn’t encounter any problem.
On the client side, the user can access it by browser or a desktop application. For the desktop installation, it can be downloaded from internet. And the configuration to connect the server is very easy; even end-users are able to do it. We prepared just one slide and end-users are able to follow it.
We implement it through a collaboration with the vendor. The vendor team is responsible for installation and configuration. And the in-house team is responsible for development.
My advice is to use a data warehouse as the data source, which can greatly reduce Qlik implementation time, as creating too many joins in Qlik is not a best practice.
My advice is to start small with the named license scheme as the price is much cheaper and it will be much easier to gain management approval.
People who have interest on QlikView should also take a look on Qlik’s another product Qlik Sense. They have similar features (e.g. Section Access, Search function, Associative model).
QlikView relies more on the IT side. IT builds the base and a majority of end-users make use of it. Some power users can build dashboards, share with others and create formulas in there.
Qlik Sense is a self-service BI tool. It is easier to use and stronger in visualization. But some advance formulas might not be available.