Power BI vs. Qlik

After Tableau and SiSense, Keeping ’em Honest continues with Qlik. Now, after they sold the company and took a hit in the latest Gartner’s quadrant, I didn’t expect much of a completion from Qlik but every now and then I run into a customer considering Qlik and willing to share Qlik’s feedback on how Qlik outshines Power BI in every possible way (can’t share these documents due to NDA although they are interesting and entertaining stories). But to be fair to Power BI, I’ll enumerate next a few reasons that customers have shared of why they prefer Power BI over Qlik and from my experience in in helping several customers transition to Power BI from Qlik. I’ll focus on Qlik Sense, which is the Qlik’s latest tool that competes head to head with Power BI.

Chasing the perfect tool (hint: there is none)? Instead, focus on architecture and data quality. Tools come and go, and they tend to leapfrog each other. Data quality and good foundation stays on.

Data Acquisition

  1. Connectors – the data journey starts with connecting to your data. According to Qlik’s documentation, it offers a very small set of connectors. By contrast, Power BI supports over 70 connectors and the list growing every month.
  2. Content packs – In Power BI, content packs delivered by Microsoft partners allow business users to get pre-packaged reports and dashboards from popular cloud services, such as Dynamics CRM, Salesforce, Google Analytics, Marketo, and many more, directly in Power BI Service (no need to use Power BI Desktop). No such a feature exists in Qlik.
  3. Complexity – One prevailing theme I hear from Qlik users is complexity where the tool requires the end user to be a developer to create scripts and import data. In fact, I know of a large organization in Atlanta which has hired a consulting company to create and maintain Qlik reports. So much about the self-service BI story. By contrast, Power BI Service and Power BI Desktop include user-friendly wizards.
  4. Data transformation – In Power BI Desktop, a data analyst can use the Query Editor to define repeatable steps to transform and cleanse the data before it’s imported in the data model. The Query Editor (Power Query in Excel) resonates well with business users. No such a tool exists in the Qlik world

Data Model

  1. Associative data model – Qlik continues to tout the “associative data model” (learn more in this video and its related videos on YouTube). Besides the self-service vs organizational BI propaganda, nothing new here that Power BI can’t do. As far as the extolled virtues of the “associative data model”, I couldn’t agree more with what Donald Farmer (a former Vice President at Qlik) had to say a while back. “Talking of engines, some have been misled to believe that QlikView’s supposed “associative analysis” represents some significant engine smarts. I have even heard analysts very misleadingly say that QlikView has “association rules” – implying some kind of data mining, such as Microsoft implements in its Data Mining server and Excel Add-ins. QlikView add to the confusion by talking about an associative “architectural model.” However, despite the hype, as Curt Monash points out (or rather, painfully extracted from QlikTech themselves through a long thread of comments) it is not so: “The associative aspect is really more meaningful in describing the end user experience, in that you see visually what is associated and is not associated with any particular selection or drilldown.” As Curt says, “Thank you for admitting that clearly!!! It wastes a fair amount of analysts’ time when your company pretends otherwise.” So, the associative model is nothing more that the relationship auto-discovery in Power BI Desktop.
  2. Data model capabilities — I said this many times and I’ll say it again – no other tool on the market comes even close to Power BI as far as its modeling capabilities which allow business users to create self-service data models that are on a par with models created by BI pros.
  3. Calculations – In Power BI, we use DAX, a language that is far more powerful than Qlik expressions.

Platform

  1. Ecosystem — No matter how good it is, a self-service visualization tool addresses only a small subset of data analytics needs. By contrast, Power BI is a part of the Microsoft Data Platform that allows you to implement versatile solutions and use Power BI as a presentation layer. Want to implement a real-time dashboard from data streams? Azure Stream Analytics and IoT integrates Power BI. What to show reports on the desktop from natural questions? Cortana lets you do it by typing questions or voice. Want to implement smart reports with predicted results? Power BI can integrate with Azure Machine Learning? Want to publish SSRS and Excel reports alongside interactive reports? Power BI supports this. Expect this strength to increase as Cortana Analytics Suite and prepackaged solutions evolve.
  2. Scalability and continuum – Despite Qlik’s scalability white papers that claim to scale to billions of rows, a desktop tool can get you up to a few million rows. Even if its technology can scale to much bigger data volumes, working with gigabyte files is no fun. Not to mention that you’d probably wouldn’t want millions for rows imported and exported all the time. At some point, you have to consider moving your models to a dedicated server. In MS BI, you can transition from Power BI Desktop to Tabular on the server. Same technology, same interfaces. In fact, I tell students that if you know Power BI Desktop, you already know 80% of Tabular. What’s Qlik’s continuum story?
  3. On your terms – Power BI lets you deploy your report to the cloud Power BI Service and on premises to the Power BI Report Server. Qlik is predominantly on-premises solution with a fledgling cloud offering. Besides, Qlik Sense Cloud is more expensive at $25/month vs $10/mo in Power BI (if you decide to go cloud).

Analytics

  1. Natural language – No equivalent to Power BI Q&A which allows business users to ask natural questions. Qlik has a basic story telling via the Narrative Data Science and so does Power BI.
  2. Advanced analytics – No integration with R or built-in machine learning capabilities, such as Quick Insights in Power BI.
  3. Real-time dashboards – Qlik doesn’t have the API to allow developers to push data to a dashboard that updates in real time.
  4. Geospatial reporting – Qlik’s geospatial reporting is not as robust as Power BI. For example, Power BI offers at least four mapping visualizations, including ShapeMap (supports custom shapes) and ArcGIS

Openness

  1. Reporting tools – Reporting tools come and go. If make a decision to go with Qlik, can you use another reporting tool to connect to it? By contrast, in Power BI you can analyze your data with any DAX or MDX capable tool, such as Excel.
  2. Custom visuals – Qlik claims that it can integrate with D3.js to allow developers to create custom visuals. In reality, Qlik extensions basically spit out HTML. By contrast, Power BI custom visuals adhere to predefine interfaces. What this means for a developer is that it’s much easier to implement and test custom visuals in Power BI.

Other

  1. Cost – I’ll let you do your own math here but the number one reason customers gravitate to Power BI is cost.
  2. Speed – Power BI gets new features every month. How often is Qlik updated?

pbivsqlik

  • RJ

    Hi Teo

    I’ve used both. I like Power BI but for me Sense is far and away the better solution still. I would say though give Power BI another 2 years and they might have by then completely closed the gap. Based on their rate of improvement over the last year or so.

    From a user viewpoint (not setting up) though Sense is in a different league to Power BI. I have shown a demo of Power BI vs Sense to one client and they told me to get Power BI out of here. In their view Sense was so much better.

    Re loading and cleansing etc. Sense now have a non script option (Data manager). It fine for beginners to load data using a fun bubbles option for joining tables. I find this option easier to learn than Power BI. But Power BI still does more than the Sense non script option so as things get more complex Sense script is needed (using data load editor).

    Re expressions. More powerful. I don’t agree? I prefer Sense for power and simplicity. DAX is too complex. Sense is very powerful but easier to learn.

    Re costs. The Cloud option is fairly similar price wise now. It depends on what is required.

    • Prologika

      Hi RJ. This is interesting but my experience is different and Gartner seems to me on my side :-). Would you mind sharing why SiSense is a better solution and why the client told you to “get Power BI out of here”?

      • RJ

        Hi. I’ve done a report mainly comparing Power BI to Qlik Sense. Re my clients comment. He felt the drill down was far and away superior for Sense after he saw how both worked. But this is covered in this report. I did this report and then posted on the Qlik site as I prefer Sense. By some distance both as a developer (Qlik script is great but it takes a while to learn) but definitely as a user.

        Re Gartner. I stand by what I have posted in the attached report. For me Sense is close to outstanding. Power BI is quite good but improving quickly. But the main weakness is still filtering / slicing data / drilldown. I don’t know why Power BI just doesn’t copy Sense.

        https://community.qlik.com/docs/DOC-17717

        • Prologika

          You’ve done a great job in that document and keeping it up to date but you need to revisit the Power BI limitations you mentioned. About the Power BI Desktop file size, you should be happy with the latest Power BI Desktop update which has a much better compression. Since you indicate that the main weakness is still around filtering/slicing/drilldown, which features were a showstopper for your client?

          Also, you gave Power BI lower marks for Data Load join cleanse transform, Measures. Could you share why?

          This statement “Many-to-many join. Power BI can’t do this” is incorrect. Power BI has supported for a long time and its data modelling capabilities are second to none. Finally, the in-house server was released as Power BI Premium Report Server.

          • RJ

            Thanks for the corrections and comments. I will update my report and upload soon. And I will answer your other questions tomorrow.

          • RJ

            “which features were a showstopper for your client?”
            I didn’t ask him. For me Sense is still so clearly in front in this area for the reasons laid out in my report. But it was some time back and Power BI has improved a lot. But the Power BI team know they need to do a lot more still. And will do.

            But just linking the various filters as Sense does and having a one click button to show all filters made in either charts, the side filter or slicers would be a big step in the right direction. And in this filters made list being able to change filters or delete as Sense does. This feature while no where near as good as Sense was also available in Qlik View. So I’m surprised it’s not available yet in Power BI. But I suspect Power BI was developed as a punt without enough thought regarding the basics and now its successful they are going back to get some of the basics right.

          • RJ

            Re measures. I will add a separate section to my report when I have time. But I thought Sense was way in front of Power BI. More so than the overall ratings I gave. For 5 main reasons that I can think of.
            1. The table expressions were similar. Because I thought DAX was so awful compared to Sense (hard to learn and use and far from intuitive) I mainly did more complex expressions in the tables with Power BI (rather than just in the charts using DAX). Whereas in Sense the measures are mostly easy and intuitive and also powerful. And when it gets complex using set analysis I just reverted to ‘if’ statements etc. Or did an expression in the table.
            2 Sense also has master measures in addition to standard chart measures. Power BI doesn’t have these. I set up almost all all my expressions using this feature now.
            3 The table expression and chart expressions in Sense are similar except for the addition of the powerful set analysis feature. DAX is so different. It’s another learning curve.
            4 Sense has features that DAX is missing. Like AGGR. I try to avoid aggr as it’s complex. But I have found I can’t at times. And once the AGGR concept is understood its not too difficult from then on. Another is Dual. I’m surprised PBI is missing this. It can be done in a different way but its a bit complex.
            5 Multi filters in a measure. eg 10 customers cf to the balance. This is so simple using set analysis
            sum(SalesValue£). That’s it. Its also possible to use ‘123*’ to include all number starting with 123. Or ‘-‘ (minus) to exclude customers etc.

            Compared to PBI
            SalesKeyCust = CALCULATE(sum (SalesDetail[SalesValue£]), Customer[CustNum] = 123 || Customer[CustNum] = 456 || Customer[CustNum] = 895 etc)

            Here’s another viewpoint. I thought Mike did a good job on this. although I found learning simple (eg above) set analysis easy enough. Agree it does get complex for the more difficult calculations but I revert to ‘if’ when in a hurry.
            https://nform.biz/creating-measures-dax-qlik-script/

          • RJ

            And to finish. “Data Load join cleanse transform”. When I first looked at PBI I thought their non script option was much easier for a beginner than Sense. But now that Sense has their non script option (Data Manager) even for a beginner my view is Sense is easier. By some distance. So a new user can start and then if interested (or necessary) learn script as required. It’s even possible (June 2017 release) now to combine the data manager with the written script.

            Re an experienced user. I much prefer script than the PBI edit queries approach. And my view is more can be done by a non programmer like myself than with Power BI. For example
            >> I always try to just have one calendar. So one month and Year etc that works for all dates. As I’ve seen users use the wrong date. This can be done using canonical dates (or a date Island)
            https://community.qlik.com/blogs/qlikviewdesignblog/2014/02/17/canonical-date
            Its an option that I use a lot. But others prefer different ways of doing this
            >> I set up a BoM for one client. I found this as a non programmer quite simple thanks to this excellent blog post.
            https://community.qlik.com/blogs/qlikviewdesignblog/2015/08/31/bill-of-materials

            I looked at PBI to do this and gave it a miss. Maybe a more experienced PBI could do this though.

            But overall I have been amazed what I have done in Qlik apart from the above. And I’m aware that this might bias my judgment though so haven’t written that much on this in my report. I would need to spend many more hours on PBI to fairly compare this. But of the time I have spend I felt Qlik is superior.

          • Prologika

            Thanks for the input. Role-playing dimensions can be handled either by importing a date table multiple times or force the join using USERELATIONSHIP DAX function. About scripting and complexity, why should a self-service tool require scripting? And I think the Query Editor deserves much higher marks.

            Anyway, let’s see how the tools will evolve over time. To me, the Power BI biggest advantage is its ecosystem as no tool is a kingdom on its own. To this end, you might be interested to take a look at where MS will bring Power BI in next couple of months (https://www.youtube.com/watch?v=zaVGt-IbuhU).