Qlik Sense – The 6 Best New Features
By: Scott Duthie, Senior Consultant at Pomerol Partners US
The November release of Qlik Sense has just come out and included is awesome new functionality worth discovering. With Qlik on a 12-week release cycle, it can be challenging to keep up with what’s new, so the top new functions are summarized below.
Much of the new functionality is aimed at bringing back features from QlikView, improving the experience for self-service users and enabling rapid front-end development.
The great news for developers is that many of the pain points have been fixed, especially within the front-end report development UI. Think show/hide conditionality on charts, chart containers, ‘Always One Selected Value’ in a field, buttons with actions, variable input boxes, sliders and more…
1. Dashboard Extension Bundle
This bundle of great extensions comes as part of the software install. Simply tick the box to enable these when running a software update and they will appear under ‘Custom Objects’ in the front-end UI. These Qlik supported extensions remove the need for many of the open source extensions we have been using. This removes the risk of open source code and enables a lot more creativity in dashboard design.
○ Date picker object: Just like booking a flight – select a custom range date range or a default range (Rolling 12 month, Year to Date etc.)
○ Button object to perform navigation and selection functions.
○ Container object to hold multiple charts.
○ Show/Hide container to hold multiple visualizations and reveal based on conditionality.
○ Variable input object: Text input, slider, buttons, dropdown menu of values etc. Combining the variable set buttons with a conditional object in a container is very powerful and will allow us to create dashboards in a similar fashion to the ‘layered’ QV methodology.
2. Insight Advisor: Qlik’s Built-in AI Tool
Users can now help ‘train’ the Insight Advisor by giving feedback to the program when it gets thing wrong, offering guidance on which fields are dimensions/measures etc. This AI component has seemed a bit immature over the last few releases, but the power of this is now becoming very clear, especially once ‘trained up’ in a large self-service environment. Whether you choose to use the suggested chart types and insights or not, you can still use the auto created objects as a start and greatly accelerate the pace of report development.
3. Alternate States
This is awesome. You can now set up alternate states of the same dimension within the master items library. This enables really slick comparative analysis without the need for complex set analysis and variables. A typical example would be comparing sales from a range of months, to sales from another range of months. This removes the need for the ‘Prior Calendar’ island methodology we would traditionally deploy. This is a huge advancement for self-service, as set analysis has always been a complex topic for non-developers. Speaking of set analysis….there’s also a new ‘wizard’ to auto-create the syntax for set analysis in the expression editor.
4. ‘Always One Selected Value’
Another super simple but incredibly powerful function that we all missed from QlikView. This forces one value to be selected from a field at all times. Perfect where you need the user to select a language, currency, FX conversion type etc.
5. App Opening Default Selections
Another really powerful function – Set up a default selection state in an app and save it as a bookmark to be deployed when the app is opened. Great for opening a dashboard on ‘Year to Date’.
6. The Mapping Tool Continues to Improve
We’re seeing more and more of the advanced GeoAnalytics functionality available within the standard mapping tool.
Mini charts (pie charts, bar charts) can be plotted on geographic points. For example you can show ‘Sales by Product’ for each store on a map, or ‘Incoming and Outgoing Goods’ at a warehouse location.
Other features from the last few releases: A to B line layer mapping (great for visualising networks or distribution routes), heat map layer for high-density geo locations, drill down through layers (Eg. State > County > Zip > Store Location), auto lookups for coordinates and shape files for locations and areas.