Last month Tableau and Qliktech both declared that Traditional BI is too slow (I am saying this for many years) for development and their new Data Visualization (DV software) is going to replace it. Quote from Tableau’s CEO: Christian Chabot: “Traditional BI software is obsolete and dying and this is very direct challenge and threat to BI vendors: your (BI that is) time is over and now it is time for Tableau.” Similar quote from Anthony Deighton, Qliktech’s CTO & Senior VP, Products: “More and more customers are looking at QlikView not just to supplement traditional BI, but to replace it“.
One of my clients – large corporation (obviously cannot say the name of it due NDA) asked me to advise of what to choose between Traditional BI tools with long Development Cycle (like Cognos, Business Objects or Microstrategy), modern BI tools (like JavaScript and D3 toolkit) which is attempt to modernize traditional BI but still having sizable development time and leading Data Visualization tools with minimal development time (like Tableau, Qlikview or Spotfire).
Since main criterias for client were
-
minimize IT personnel involved and increase its productivity;
-
minimize the off-shoring and outsourcing as it limits interactions with end users;
-
increase end users’s involvement, feedback and action discovery.
So I advised to client to take some typical Visual Report project from the most productive Traditional BI Platform (Microstrategy), use its prepared Data and clone it with D3 and Tableau (using experts for both). Results in form of Development time in hours) I put below; all three projects include the same time (16 hours) for Data Preparation & ETL, the same time for Deployment (2 hours) and the same number (8) of Repeated Development Cycles (due 8 consecutive feedback from End Users):
It is clear that Traditional BI requires too much time, that D3 tools just trying to prolongate old/dead BI traditions by modernizing and beautifying BI approach, so my client choose Tableau as a replacement for Microstrategy, Cognos, SAS and Business Objects and better option then D3 (which require smart developers and too much development). This movement to leading Data Visualization platforms is going on right now in most of corporate America, despite IT inertia and existing skillset. Basically it is the application of the simple known principle that “Faster is better then Shorter“, known in science as Fermat’s Principle of least time.
This changes made me wonder (again) if Gartner’s recent marketshare estimate and trends for Dead Horse sales (old traditional BI) will stay for long. Gartner estimates the size of BI market as $13B which is drastically different from TBR estimate ($30B).
TBR predicts that it will keep growing at least until 2018 with yearly rate 4% and BI Software Market to Exceed $40 Billion by 2018 (They estimate BI Market as $30B in 2012 and include more wider category of Business Analytics Software as opposed to strictly BI tools). I added estimates for Microstrategy, Qliktech, Tableau and Spotfire to Gartner’s MarketShare estimates for 2012 here:
However, when Forrester asked people what BI Tools they used, it’s survey results were very different from Gartner’s estimate of “market share:
“Traditional BI is like a pencil with a brick attached to it” said Chris Stolte at recent TCC13 conference and Qliktech said very similar in its recent announcement of Qlikview.Next. I expect TIBCO will say similar about upcoming new release of Spotfire (next week at TUCON 2013 conference in Las Vegas?)
These bold predictions by leading Data Visualization vendors are just simple application of Fermat’s Principle of Least Time: this principle stated that the path taken between two points by a ray of light (or development path in our context) is the path that can be traversed in the least time.
Fermat’s principle can be easily applied to “PATH” estimates to multiple situations like in video below, where path from initial position of the Life Guard on beach to the Swimmer in Distress (Path through Sand, Shoreline and Water) explained:
Even Ants following the Fermat’s Principle (as described in article at Public Library of Science here: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0059739 ) so my interpretation of this Law of Nature (“Faster is better then Shorter“) that traditional BI is a dying horse and I advise everybody to obey the Laws of Nature.
If you like to watch another video about Fermat’s principle of Least Time and related Snell’s law, you can watch this:
Google+
February 15, 2014 at 2:38 am
Great article, I work in the traditional BI space and I agree with most of the points you make. The one factor I feel you are not factoring in your analysis are the end users. The challenge will be how to bring about the change to the thousands of existing users within a corporate BI setting. It is the same reason Internet Explorer 6 is still around. The total cost of making the switch to the newer tools is prohibitive.
The approach we have taken is to allow power users to install the new age visualisation tools and pull their data from the existing traditional BI tools we use.
February 15, 2014 at 11:02 pm
Hi: I actually recently helped to migrate of 4000+ end users from dead BI (BO, Microstrategy etc.) to Tableau for company with 300000+ employees and the total cost of migration was less then maintaining the obsolete BI infrastructre
February 25, 2014 at 3:06 pm
[…] of all, as I said on this blog many times that BI is dead and it replaced by Data Visualization and Visual Analytics. That was finally acknowledged by Gartner itself, by placing Tableau, QLIK and Spotfire in […]
April 26, 2014 at 5:06 am
Hi Andrei,
I am working with SAP BW & BusinssObjects for past 4 years and I completely agree with you, modern Data Visualizing softwares are way ahead of way old relics like BO, BeX, Microstrategy and OBIEE.
But data visualization is only one side of the story(user experience), i think in coming data we will have better software to replace existing Data Warehouses like SAP BW or TeraData with likes of Hadoops( Cloudera, Hortonworks etc) which will have real time capabilities as well.
I recently got a chance to work on QlikView and i was amazed by the speed, ease of devlopoment and user experience. It is almost impossible for BO or Microstrategy(at current pace) to match the quality and features of QlikView.
I have also started testing Tableau, its quite amazing and i just love playing around with it to explore data visualization possibilities.
Regarding the BO to Tableau migration, I would like to know the challenges you faced? Specially with a lack of Semantic Layer(Universe) in Tabluea.
Regards
SJ