Time to write another philosophical post that will upset some people and will bore some others to death. Hell yeah. But don’t worry, we’ll also have fun stuff, so bear with me. Here’s my story: In the last few months, I’ve been working a lot more with Qlik Sense and, even though I’ve learned to love it, I still have some doubts about its adoption and the future of the Qlik platform in general.

A few days ago, I saw a post by Qlik Luminary Aaron Couron where he asked the world a simple question: QlikView or Qlik Sense? What’s your preference? While I was reading the answers, I remembered many treads I’ve encountered in QlikCommunity and some good discussions we’ve had in our Qlik Dev Group events, but it specially reminded me of the comments I got after tweeting these images not so long ago:

32_01

32_02

Some people favored one over the other, some argued that both should coexist and some others simply went nuts. But regardless of which side you choose, it’s amazing to see the number and variety of opinions in this regard. In a way, I find it interesting how we’re still asking ourselves if this town is big enough for two Qliks. After all, Qlik Sense was released almost three years ago, which in technology standards is a lot of time!

Don’t forget to check Aaron’s full post here. Lots of cool stuff!

Like many of you, I’ve grown quite fond of Qlik Sense as well. It has many advantages like a flexible security schema, more competitive prices and amazing governability in terms of its users, data, apps, metrics and pretty much everything else. However, I don’t think I’m ready to get rid of QlikView just yet.

In the meantime, I also started questioning myself about how the users, developers and even our business model have changed recently. Even though I agree with Stephen Few regarding the soundness of Gartner’s Magic Quadrant (witness the mega-burn here), that’s definitely a message we should not ignore. 

In today’s post, I want to share some of my ideas about this topic. But more importantly, I want to hear what you think. I’d love to hear business users, IT specialists, consultants and Qlik employees from all around the globe. Our markets behave very differently and it would be fun to hear everyone’s opinion.

Side note: People tend to take these things waaaay too seriously, so here’s a little disclaimer. I work with QS around 40% of my time (the rest goes to QV). I like them both and alongside R, they’re my favorite tools for data analysis. If something here sounds a little harsh, remember that I’m JO-KING.

32_03

But anyways, let me start by discussing why I disagree with some of the comments I’ve read in social networks (not that anyone cares, but here we go).

“QlikView and Qlik Sense are different products”

This was one of the most common comments on my Civil War image: “QlikView and Qlik Sense are different products, you simply can’t compare them”. Well, duuuh, of course they’re different products.

32_04

But seriously, I must disagree with the second part of that statement. I CAN compare them. In fact, I do it quite often. I compare them every time a new prospect asks me which is the best tool to cover their information needs or when an old customer asks me if they should increase the number of licenses they have, switch platforms or get both. This type of comparison is natural and most companies will do it in order to choose one of them. (Of course, there are some firms that can easily acquire both, but where I come from that’s not the most common scenario. Damn you first world countries!)

“Both products are there to fulfil two very different purposes “

Well, I understand there are several things that make each platform unique, but “two very different purposes” sounds like if we were comparing apples to oranges. Let’s face it, even though we want to make them look like day and night, they are similar tools and solve similar problems. In fact, there are many cases where both can perfectly fulfill the company’s requirements. From a business perspective, they are amazing means to integrate multiple data sources and present stunning and powerful dashboards that help users make better decisions by letting them navigate freely through the data with the associative model. In conclusion, same-same, but different.

“QlikView is for guided analytics. Qlik Sense is for data discovery”

Yeah, we get it. You learned your marketing mantra, but please control yourself. Once again, I understand that some well-thought features make the self-service data discovery process easier in Sense, but come on! You can’t simply put them in boxes and say they can’t do nothing more. In fact, when we started talking about business discovery back in 2012 (-ish?), we used to talk about QlikView.

32_05

I think that business discovery has more to do with the people and how they interact, transform, visualize and interpret data than with a tool itself. By the way, do you remember this gem? It’s one of my favorite speeches by Donald Farmer:

Using QS doesn’t mean you’re doing data discovery just as using QV doesn’t mean you’re doing guided analytics. Believe me, I’ve seen companies that use Sense as an old reporting tool from the 80s and some others who think QlikView is a visual query plugin for Oracle. Once again, it’s not about the tool, it’s about what you do with it.

32_06

“Don’t expect Qlik Sense to do the same as QlikView”

I have to admit that this one is true: it would be mistake to assume that you can (or should) do exactly the same things in both platforms. You have to change your mindset and create a new style for your dashboards. However, this is not my point.

When I complain about Qlik Sense not being able to do certain thing, people often tell me “don’t expect Qlik Sense to do the same as QlikView”. Well, it’s not that I want them both to do the same things. My problem is that there are many basic features that I think Qlik Sense should offer out-of-the-box that are simply not there. And once again, it’s not because I want every single feature I had in QlikView. It’s just because I think that having a freaking dynamic label for your tables is a pretty reasonable thing to ask for.

32_08

Woah, now I feel better. I really had to get that out of my chest… [Did you notice the fancy section divider? I spent 20 minutes looking for it!].

Now, here’s a collection of random ideas I’ve had regarding the QV vs QS debate (you know, that kind of thoughts that hit you while you’re stuck in the traffic or taking a shower at 5 am before work or while drinking your 12th beer in the pub).

There’s still something missing…

“Qlik Sense is the future”, “Can’t you see the potential?”, “QlikView will disappear soon, Qlik Sense will prevail”, “Get rid of QlikView already!”. Well, I’ve been hearing that for almost 3 years and yet, here we are. Don’t get me wrong, I know that a day may come when QlikView ceases existing and be replaced by a more modern tool…

32_09

After using it for the last months, I still find Qlik Sense a little unstable. I’ve had several issues like bookmarks disappearing, apps not opening, QS desktop randomly closing while I work, extensions going crazy after an upgrade, etc. As I mentioned in the last section, Sense still lacks a certain je ne sais quois to fully steal my heart. And from what I’ve read from many users and analysts, it’s not only my imagination.

However, truth must be told: I’ve seen great improvements in shockingly short periods, not only regarding bug fixes, but also better functionality. You want variables? Boom, there you have it! Pivot tables? Done! Mike Tarallo, Mark Bilotta and the Qlik support team are always ready to help you with any problem you have, so I’m sure all these minor issues will be solved soon.

Split efforts

Most of the parents say that they love their children equally. However, it’s not strange to see that they give one of them a little bit more. I think that’s the case with the Qlik brothers. Having two products means that you have to split your R&D / Support / Marketing teams to serve both of them. And someone is certainly receiving more attention.

I’m not saying this is wrong. Actually, it is OK to pamper the youngest ones in the family. However, while Qlik is working to make Sense more competitive, they have nearly forgotten about QlikView. Think about it. When was the last time a QV release that made you say “wow, that’s cool”?

And not only that, while Qlik is dividing its efforts between these two, other companies like Tableau and Microsoft are releasing very powerful features for their products at great speed. I still think the Qlik platform is better, but it’s important to stay focused and define a concrete vision that will keep us as leaders rather than scattered and sometimes divergent efforts. Whereas SAP and Microsoft can maintain dozens of products, I don’t think we have the size to do that.

The new developers

The average Qlik user may or may not be the same, but there’s a stakeholder that has definitely changed: the developers. Every time I’m assigned as the lead instructor for a QlikView course, I start by telling the participants something like this:

“If you’re a technical guy, you will learn QlikView easily. But if you’re from the business side (Finance, Sales, HR, etc.) you don’t have to worry. As long as you know a little bit of Excel and have a clear idea of what you want to show in your dashboards, you’ll be OK.”

And I really mean that. Everyone can learn how to build applications in QlikView regardless of their background. Some might be a little intimidated by the script, but once they’re used to seeing raw lines of code, they start enjoying the process. And it’s even easier with Qlik Sense. The drag and drop interfaces make it simpler for the non-technical developers. However, there’s another side to it.

In order to get the most out of Qlik Sense, you also need other kind of developers. Those who know their way around JavaScript, HTML, CSS, D3, Leonardo and Enigma. And believe me, these guys are not likely to be hiding in your HR or your Finance departments.

This new skill set is much more robust than the one you require for common Qlik development and not everyone will be eager to learn it. Chances are that you’ll have to hire someone (if your budget allows), or ask the IT department to lend you a hand. Yep, that same IT department you told your boss you didn’t need because Sense is a super-easy self-service platform. Oh, and don’t forget that you will have to spend some time / money in their training too!

Power vs Speed

Simplicity has always been one of the things that define Qlik. That’s one of the reasons why I love these platforms. They’re simple, they’re fast, they’re powerful. But in a way, this has also changed.

With QlikView, you can perform a wide range of things. Some of them are simple, like loading an Excel file and renaming fields to create keys. Some others are complex, like creating a tile grid map based on a pivot table.

32_10

With Qlik Sense this gap has widened. The simple stuff is even simpler. You just drag the file into the app and boom: the data is loaded. Grab a bubble and drop it next to the other: a key has been created. However, the complex stuff is even more difficult. Instead of following a tutorial or copying an object shared via AskQV (where the best Qlik content is), you have to create an extension. This might not be an extremely hard process, but it still requires more time to build, test, package, upload and maintain than a native object. And moreover, not everyone can do it, leaving you with a partial-self-service tool.

32_11

But of course, there’s a benefit behind this. The extensions / mashups / QAP are extremely powerful. Once you get the knowledge you need to create them (or enough money to buy them or hire someone who does), your possibilities are endless. A small step for a man, a giant leap for mankind. And this is only the beginning, as we’re still figuring out the things we can achieve with these tools. I’m sure we’ll see amazing stuff in the next months.

32_12

[Damn, those state-of-the-art diagrams look soooooo good. I made them in Paint / PowerPoint if you were wondering…]

A whole new world

There are multiple ways to overcome this challenge, but as always, the all have pros and cons:

  • Time to learn: JavaScript, HTML, CSS and all those things are not exactly rocket science. There are a lot of free resources online and some well-structured courses (at reasonable prices). Of course, you’ll need some time and practice. The downside is that many users who are interested in building Qlik apps don’t feel comfortable with the “hardcore” stuff.
  • Hire someone: These modern technologies are very popular, so finding someone who’s already literate shouldn’t be that hard. Besides, having someone who has experience building web apps isn’t going to hurt you. This option obviously has an associated cost, so be sure to give the new recruit a lot of things to do (you don’t want her to be just sitting there, doing nothing, after she finishes the custom object you asked for).
  • Qlik Branch: Open source projects developed by generous professionals around the globe. Cool. Yeah, open source is the best. It’s free! However, most of them were developed by people who… well, have jobs and personal lives… so don’t ask for major improvements every week or expect them to fix your issues on a Friday evening.
  • Buy them: I’m sure that we’ll see a lot of companies offering these products in some sort of online store soon. I wonder if Qlik will try to regulate it with some kind of app certification process or just let anyone sell whatever they want. Having multiple players in the market is always positive, but it also means that we’ll have to spend some time evaluating their products, prices and conditions. The schemas are also to be defined: monthly / annual subscriptions, individual packages or extension suites, ad-hoc deliverables, support packages, SLAs, etc. A downside to this model is that there are some companies where registering a new business partner (a.k.a. supplier) is a long and boring process that can take weeks and the lady at the purchasing department will be mad at you if you make her fill all those forms just to buy a $9 USD extension.

32_13

DLC (Downloadable Content)

For those of you who are not familiar with the gaming industry, let me give you some context beforehand. Most of the videogames are self-contained, meaning that you don’t need anything else to understand the story or enjoy the game. On top of that, some publishers offer DLC, which is extra content that is sold separately. Sometimes they have additional levels, new characters or better weapons but it’s generally non-vital stuff they sell for a low price.

However, some companies are known for releasing half-baked games and selling DLC as patches to correct them. Others offer lousy base games and sell all the good stuff as DLC at high prices, so you end up paying more for these “extras” than the actual game.

32_14

Now that we’re all on the same page, take a look at this image:

32_15

You know where this is going, right? I always though Qlik Sense was a self-contained tool (a full videogame), but as time went by, I realized that I didn’t enjoy it that much without those extra modules (web apps, extensions and mashups) that had to be bought / developed separately.

Today, this is not a big problem, but maybe in the future I’ll have to invest more in the DLC than in the game itself. And if that’s the case, should I omit Sense and go directly to the QAP? [Mmmmm… That hamburger looks delicious… Maybe I should cancel that pizza and go buy a hamburger…]

32_08

My dream

I’m no IT guru and I certainly can’t foresee what’s the best road to walk, but as a user, developer and Qlik enthusiast, I’d love to see a unified platform that merges both products into an ultimate data solution. And let me tell you, I’ve seen it in my dreams! It has all the out-of-the-box power and flexibility of QlikView, but with Qlik Sense’s extensibility, modern technologies and governance capabilities. It can be deployed on premise, cloud and hybridly [is that even a word?] without having to change anything at all. It has a flexible, scalable and competitive licensing model [Yes, I’m implying lower prices]. And of course, it offers new, powerful features like native connectivity with R and Python.

What do you think about this? Does the world really need two different Qliks? Which one do you prefer? I’d love to hear what you have to say, so please share your comments in the section below!

Until next time!


Fuente: Qlik Freak
Visita: Double Qlik?