Qonnections 2012 – The Insights, Rumor and Assumptions & What They Mean for the Future
Qonnections 2012 came to a close in South Beach Miami on Wednesday bringing to an end what seemed to be a break from the normal Qonnections affair; there seemed to be little concrete announced but there were many pointers to the future direction of Qlikview which made a welcome change.
Firstly it should be said that I wasn’t present at Qonnections so everything here is made up from Press Releases, Twitter Feeds and a little sprinkling of inside information…and of course a fair amount of conjecture.
So; let’s get some of the more routine stuff out of the way first.
It was good to hear from Lars Bjork that revenue continues to grow at a good rate (42%), this is something I wasn’t surprised about; the number of new green field implementations I hear about in the City of London alone is enough to indicate a rapidly growing company. There will of course come a day when everyone who wants Qlikview already has it and growth may slow but I don’t see that for a good few years.
The winner of the Mobile App of the Year was also announced and I’m pleased to say that Victa won with their SnowView App as I tipped them to do in my review of the Top 20 entries (https://qvdesign.wordpress.com/2012/03/09/qliktech-global-partner-mobile-app-contest-the-runners-riders-1-thoroughbred-a-few-also-rans-and-lots-of-old-nags/), well deserved winners indeed. It seemed on the day that the feelings being expressed on Twitter for the other entrants weren’t great but hopefully this will lead to higher standards next year; it is after all early days in the Qlikview Mobile space.
One of the major concrete announcements was the imminent launch of Qlik-Market; a place for the sale of Apps, Extensions and Connectors which is due for release in the summer – to Partners only initially I understand with general users in the autumn. This seems like a good step forward in spreading useful developments although I am a little concerned as to the reliability of the content; developers are required to support their content and I know from experience this isn’t always easy; what works perfectly on your system will fall at the first hurdle on someone else’s. It will certainly be interesting to see what’s made available; the current ‘beta’ offerings are all things that can be bought from the vendors already (QV Source, GeoQlik etc), I really want to see App’s and Extensions that aren’t otherwise available making it onto the Market (Extensions could have massive potential as many Partners and developers simply don’t have the skills or time to create them). It will also be interesting to see what sort of price the content attracts as quite frankly some of the things currently on offer seem vastly over-priced; in some cases costing much more than Qlikview Server itself. If possible I’ll certainly look to add some content; perhaps some dashboard design templates and themes and also a few of my more complex home grown utility apps – but it won’t be to the detriment of posting things for free to QVDesign.
Interestingly in true ‘MacRumors’ style the Qlikview Market would appear to have been outed a month or so ago with an update to one of the Devloper’s LinkedIn Profiles: http://www.linkedin.com/pub/nick-webster/12/b20/902
A Beta Version of the QlikMarket can be accessed here: http://market.qlikview.com/
Another slightly smaller definite announcement was that in a few weeks Qliktech will be launching their own: ‘Qlikview Design Blog’. This should only be a good thing as Qliktech seem to have picked up their aesthetic practices of late so I’m reasonably confident that the majority of the content should be of use. Erica Driver tells me that it will also have a technical angle which is also very welcome. I’m assuming that this will be presented via the ‘Blogs’ section of the QlikCommunity – http://community.qlikview.com/blogs
Also regarding the Qlikview.com site; the Partner Portal is getting a re-design to launch shortly.
Qliktech also announced that they will be stepping up their Social Media activity and engagement which is good to hear but to be honest any company not saying the same in the current Social world would be crazy. Personally I’d love some more engagement from Qliktech, I don’t expect the red carpet or the keys to Qlikview Labs but they’re missing a valuable opportunity by not engaging with the wider blogging / social media community – it’s not just developers and existing users who read this blog. To be fair one long time Twitter contributor did get some Qlikview cufflinks at the Partner only Community meet up at Qonnections so I guess that counts as engagement!
There were a number of other Partner centered announcements relating to selling and certification that I won’t go into as only Partners can access them and they’ll be made aware by Qliktech in the not too distant future I’m sure.
On to the fun stuff: The Future.
One man seemed to be at the centre of Qonnections; Donald Farmer, Qliktechs ‘Product Advocate’ (whatever that means?), he presented a key note on day one and presented the White Paper on the final day which outlined Qlikview’s 5 future themes which in my opinion constitute the biggest shake up to Qlikview in it’s history.
The keynote as far as I know covered the usual loose and general BI topics that Donald has ably presented on numerous occasions, he always speaks sense and whilst I don’t agree with him entirely on several areas he’s certainly been a good addition to the company.
Whilst the keynote was surely of interest to those in the room it was the White Paper and the 5 themes for the future of Qlikview that were of real interest.
Firstly the White Paper can be downloaded here: http://www.qlikview.com/us/~/media/Files/resource-library/global-us/register/whitepapers/WP-The-Vision-for-QlikView-Next-EN.ashx
There’s a lot here and there’s also a lot of things that can be inferred from this paper so stick with me; this might take awhile.
– The next version of Qlikview will seemingly be collectively called ‘Qlikview.next’ – from what I’ve seen so far the ‘.’ and lower case ‘n’ are intentional – very ‘new iPad’. I’m still unsure as to what to think of this name; I agree that soldiering on until ‘Qlikview 28’ isn’t a good idea (nor is ‘Qlikview 13’) so a name change was predicted however the obvious question begs; what’s the version after this going to be called? – ‘Qlikview.One-After-Last’? I know it’s boring but my preference would probably have been to switch to date naming like SQL 2012 or MS Office 98; it maintains the ‘versionality’ without suggesting ‘we have to re-release our buggy software every 2 minutes’. I just feel that in a few years time when people are asked what version they’re using it’s going to sound stupid saying ‘Qlikview.next’ when it’s out of date; ‘Next’ means ‘after’, ‘forward in time’ and as far as I’m aware hasn’t been used in the wider world to name something that will certainly be obsolete down the line; ‘Pepsi Next’ is a stand alone product; it will either sell forever or be with-drawn, there’ll be no ‘Pepsi Next 2’. Perhaps it would be better to follow say Intel’s lead and use generic ‘non-time’ names like SandyBridge? But; as they say; ‘what’s in a name?’, it shouldn’t make too much difference and I’m sure the people at Qliktech marketing know what they’re doing. It could of course be the case that ‘Qlikview.next’ is nothing more than simply the name of the concepts and changes we’re seeing and won’t actually be the name of any of the final products that are delivered under it’s codename (the White Paper does elude to this being the case).
– The release cycle is being extended. There’s been no official announcement when Qlikview.next will be released but I’m hearing that it should be Q1/Q2 of next year which will mean a good 50% longer between releases than v10 to v11. This is a good thing. I know of so many customers (especially larger Enterprises) who were just about to upgrade to Version 10 once it had reached a reasonable SR only to hear about Version 11 giving them no confidence in what they’re about to laboriously install. It also increases the likely-hood that we’re going to see significant changes to Qlikview in the next release – that becomes a certainty when we get to the White Paper. This is undoubtedly called for; I’ve been saying for years that Qlikview needs an overhaul especially on the Front-end in order to maintain it’s position in the market; releasing Mekko Charts, and Server tweaks just won’t cut it anymore.
– Qliktech are working on caching for the iPad. This is to fit in with one of the 5 Themes from the White Paper ‘Mobility with Agility’, in that Mobile BI is often (and always with Qlikview) useless without a WiFi or strong 3G connection. This can only be a good development and I’m really interested to see how this is implemented given that there’s no way a 2GB Qlikview model is going to sit on an iPad; it’ll either have to be partial caching or limited to smaller ‘apps’. Of course the great thing about Qlikview’s mobile solution at the moment is that it’s the server that performs the work so you aren’t restricted by the iPads relative lack of memory and power.
The 5 Themes for the Future of Qlikview.
I won’t go into detail for all the themes as you can read the paper for yourself, I’ll outline my thoughts about each one and what I think they mean for the future. We need to keep in mind that these are ‘themes’ and not a case of ‘Qlikview.next will deliver these’ – it’s all well and good making grand statements but unless they’re delivered in the product they’re worthless.
1. Gorgeous & Genius.
For me this is the most important theme and I’m really pleased it’s number one of the 5. I’m going to say it here and now (and I really hope I’m not let down); I predict Qlikview.next will get a complete (and much needed) visual and Front-end overhaul. I don’t think this will just be a change to the client’s ‘skin’ or new menu options and an odd chart or two. I think it will be an almost complete re-working of the entire product with the way developers and users (I think the line between the two will be blurred Tableau-style) interact with Qlikview and their data totally changing. This (despite protestation from Qliktech) will be at least partly inspired by Tableau’s successes; I like Tableau but it’s nowhere near as good as Qlikview; there isn’t anything I’ve seen in Tableau that Qlikview.next couldn’t equal or surpass whereas Tableau (and others) will never get Associative Logic and therefore fall short. If my prediction does come to pass then: fantastic, Qlikview will really have it all and in my view will be un-touchable* in the ‘New’ BI world we find ourselves in.
* The only chink in that may be ‘Big Data’ – which is vastly over-blown in my view; in most cases the larger the data gets the lower the returns are from analyzing it (that of course isn’t a rule; use some common sense when interpreting it!). Most of the buzz around ‘Big Data’ is simply because it sounds good; who wants to talk about ‘data’ when there’s ‘BIG DATA’ – in much the same way we’ll be seeing lots more ‘Data Scientists’ over the coming years.
Another reason for my expectations for a comprehensive overhaul is one little phrase that Donald mentioned during his Keynote: ‘Suggestive UI’ – ie: the UI points devs/users towards making the right chart or colour choices for their data in a similar way to Tableau (although as Donald tweeted to me “Tableau does it the boring way – giving the user guided choices. Nice but suggestive is way cooler”). Simply put I don’t see any way how such a process could be added into the current methods of creating Qlikview objects and therefore a complete interaction overhaul is the only logical path.
It was interesting to see ‘Scandinavian Design’ as an inspiration mentioned and a visual provided to illustrate it…who’s most obvious component was designed by a Dutchman; the black lights are Marcel Wanders’ ‘Sky Garden’ for Italian firm Flos – that’s nit picking in the extreme I know but I bring it up to illustrate a point: if ‘Scandinavian Design’ is implemented with the slightest amount of sloppy-ness it won’t work. Don’t forget; that wobbly coffee table you bought from Ikea thinking it was a bargain was meant to be ‘Scandinavian Design’.
Authors admission: I’m a huge fan of Scandinavian design, especially when it comes to furniture, I’m looking at a vintage Yngve Ekstrom ‘Windsor’ chair as I write this and the first Mid-Century piece I ever bought was an Arne Jacobsen ‘Grand Prix’ chair so I know the potential for the principals ‘Scandinavian Design’ stands for. I could write for pages as to the merits and pitfalls of re-working Qlikview in this way but I feel that’s for another post; hopefully nearer the time of release when a beautiful, intuitive and reliable Qlikview.next has been revealed.
In everyway possible; fingers crossed. Please don’t mess this up Qliktech.
2. Compulsive Collaboration
I don’t doubt that collaboration has its place in BI delivery and creation but I think it’s importance is often over blown; it sounds great on paper but I’ve rarely seen it work in practice. I hope Qlikview.next can deliver something that changes that but to be honest; I’d be surprised, I just don’t see how reliable, meaningful collaborative analysis…sorry Data Discovery can be achieved with a disparate set of users and often disparate information requirements.
3. Mobility with Agility
I think Qlikview already scores quite highly on this front; the ‘develop once, deploy many times’ mantra works well, I’m a fan of the automatic touch layer added by the server to existing documents. I see this as perhaps only being slightly adjusted, iPad caching has already been announced as being in development but I think we’ll also see further enhancements to the AJAX Client to further unify the way that Qlikview is consumed and maybe just maybe we’ll see a fully online version of Qlikview (development as well as consumption).
Much like Collaboration though I see Mobility as being a ‘good on paper, not always good in practice’ thing – I don’t doubt there are apps that are much better ‘mobile’ or due to the situation can only ever be ‘mobile’ but on the whole it falls into that all too common BI trap of ‘wow!, that looks good therefore it must be good’ when in fact it offers nothing new; people need to dig down and get at the true value. Before people jump to Mobile’s defense; I know there are times when it’s brilliant; the doctor on a ward, the construction worker up a skyscraper, it’s just I don’t see the benefit when it’s the desk based manager who starts using Mobile and can actually do less than they can via the Qlikview client. I know that’s very un-cool but certainly up to this point it’s the truth founded in experience, I hope Qlikview.next proves me wrong.
4. Enabling the NewEnterprise
This is one of those ’says a lot but means nothing’ sections; you could take everything it says back 5yrs and claim that Qlikview has already achieved it. The notion to make it easy for IT departments to deliver self-service BI to users is a good one and one that I think Qlikview already does well but any improvements will be welcomed.
The only potential sticking point for me is the term ‘Self-Service BI’. In relation to Qlikview this could mean one or both of two things. Firstly a developer creates a document and end users (sorry Donald) access the document to find answers to their own questions from within the data, this is one area in which Qlikview excels. I’ve seen stacks of static Business Objects report or Excel workbooks replaced by simple Qlikview dashboards just because they allow users to find their own answers instead of telling them what answers they’re looking for. The second meaning (and I think we’re heading in this direction) is for developers to present a fairly loose document and then end users are set free to create their own reports and chart objects; we’ve seen this advance through the last few releases of Qlikview. Personally I don’t like it (I know you can just not implement it so I’m not overly concerned), I’ve rarely come across a data set that is in any remote state where I’d consider letting a user loose on it; business data is mostly far too complex to allow someone to knock up a chart based on what they think they know. By way of an example; a client recently gave a copy of a rather complex financial dashboard to the associated Business Analyst as he wanted to create some additional charts, he had a reasonable knowledge of the system and a basic understanding of Qlikview yet within minutes he’d created charts that were wrong as he’d made incorrect assumptions about the data. The biggest danger of all is that situations like this give the user an answer it just unfortunately is the wrong one and that’s worse than no answer at all.
5. The Premier Platform
This is another continuation of developments we’ve seen in the past and is entirely logical: who wouldn’t want to make Qlikview more deliverable and accessible.
I’m going to make another totally unfounded prediction; Qlikview.next will include QVSource functionality. This section talks a lot about ‘Data Connectivity’ and this is exactly what the guys at Industrial Codebox are allowing with QVSource (www.QVSource.com). Also there’s some history here as well; Industrial Codebox developed the initial version of the iPad touch interface and I believe the Qlikview Workbench before Qliktech bought the IP and bundled it with Qlikview.
API’s are also mentioned heavily and this could be another area of great change, for instance truly delivering Qlikview as ‘Software as a Service’ which could have great potential; there are a myriad of apps that I could create and offer up but I can’t afford the licensing and server hardware so a change to that would be greatly welcomed.
So, whilst nothing major was definitively announced there was a massive amount that was at least suggested through the White Paper and other indicators and most of it was good if not great.
I have to say I’m left with an overall good impression for the first time, Qliktech really seem to have transitioned from being a Startup to now being a strong, growing, reliable & professional company. Not only does the vision and scope seem much wider reaching and quite frankly better than in the past it also seems from the other areas of Qonnections such as the seminar headings for example that Qliktech is starting to do everything right in moving forward in a confident and assured manner. If everything suggested at Qonnections comes off then Qlikview will be the product that we all know and love but it will be so much more as well and do wonders for achieving Qliktech’s aim of touching a billion people’s live’s.
Next years Qonnections will be held in the Bahamasfrom the 7th April at the Atlantis Resort, I’d love to attend but seeing as I’m no longer working for a Partner it may not be an option unless Qliktech feel generous & open and offer an invite.
I’m sure some of you – especially those in attendance – don’t fully agree with my thoughts above so feel free to comment if I’ve missed anything or you see things differently.
***UPDATE: Since writing this post I’ve managed to view a smuggled video of Donalds Qlikview.next presentation and whilst it’s no less ‘conceptual’ than the White Paper it really does re-enforce the view that we’re going to see major creation/interaction/consumption changes with the next version(s) of Qlikview.***
(Unfortunately Donald has requested that the video not be shared in the public domain as it’s meant for Partners eyes only and whilst I’m all for freedom of information etc I can’t go against an idividual’s express wish.)
As always; all the best,