20110126

Poll about Data Visualization tools

On New Year Eve I started on LinkedIn the Poll "What tool is better for Data Visualization?" and 1340 people voted there (which is unusually high return for LinkedIn polls, most of them getting less then 1000 votes), in average one vote per hour during 8 weeks, which is statistically significant as a reflection of the fact that the Data Visualization market has 3 clear leaders (probably at least a generation ahead of all other competitors: Spotfire, Tableau and Qlikview. Spotfire is a top vote getter: as of 2/27/11, 1pm EST: Spotfire got 450 votes (34%), Tableau 308 (23%), Qlikview 305 (23% ; Qlikview result improved during last 3 weeks of this poll), PowerPivot 146 (11%, more votes then all "Other" DV Tools) and all Others DV tools got just 131 votes (10%). Poll got 88 comments (more then 6% of voters commented on poll!) , will be open for more unique voters until 2/27/11, 7pm and its results consistent during last 5 weeks, so statistically it represents the user preferences of the LinkedIn population:



URL is http://linkd.in/f5SRw9 but you need to login to LinkedIn.com to vote. Also see some demographic info (in somewhat ugly visualization by ... LinkedIn) about poll voters below:



Interesting that Tableau voters are younger then for other DV tools and more then 82% voters in poll are men. Summary of some comments:

  • - poll's question is too generic - because an answer partially depends on what you are trying to visualize;

  • - poll is limited by LinkedIn restrictions, which allows no more than 5 possible/optional answers on Poll's question;

  • - poll's results may correlate with number of Qlikview/Tableau/Spotfire groups (and the size of their membership) on LinkedIn and also ability of employees of vendors of respective tools to vote in favor of the tool, produced by their company (I don't see this happened). LinkedIn has 85 groups, related to Qlikview (with almost 5000 members), 34 groups related to Tableau (with 2000+ members total) and 7 groups related to Spotfire (with about 400 members total).

  • Randall Hand posted interesting comments about my poll here:    http://www.vizworld.com/2011/01/tool-data-visualization/#more-19190 . I disagreed with some of Randall's assessments that "Gartner is probably right" (in my opinion Gartner is usually wrong when it is talking about BI, I posted on this blog about it and Randall agreed with me) and that "IBM & Microsoft rule ... markets". In fact IBM is very far behind (of Qlikview, Spotfire and Tableau) and Microsoft, while has excellent technologies (like PowerPivot and SSAS) are behind too, because Microsoft made a strategic mistake and does not have a visualization product, only technologies for it.

  • Spotfire fans from Facebook had some "advise" from here: http://www.facebook.com/TIBCOSpotfire (post said "TIBCO Spotfire LinkedIn users: Spotfire needs your votes! Weigh in on this poll and make us the Data Visualization tool of choice..." (nothing I can do to prevent people doing that, sorry). I think that the poll is statistically significant anyway and voters from Facebook may be added just a couple of dozens of votes for ... their favorite tool.

  • Among Other Data Visualization tools, mentioned in 88 comments so far were JMP, R, Panopticon, Omniscope (from Visokio), BO/SAP Explorer and Excelsius, IBM Cognos, SpreadsheetWEB, IBM's Elixir Enterprise Edition, iCharts, UC4 Insight, Birst, Digdash, Constellation Roamer, BIme, Bissantz DeltaMaster, RA.Pid, Corda Technologies, Advizor, LogiXml,TeleView etc.



Permalink: http://apandre.wordpress.com/dvpoll/

20110116

Big Data Analytics: Signal-to-Noise ratio even lower then in BI?

"Big Data Analytics" (BDA) is going to be a new buzzword for 2011. The same and new companies (and in some cases even the same people) who tried for 20+ years to use the term BI in order to sell their underused software now trying to use the new term BDA in hope to increase their sales and relevancy. Suddenly one of main reasons why BI tools are underused is a rapidly growing size of data.



Now new generation of existing tools (Teradata, Exadata, Netezza, Greenplum, PDW  etc.) and of course "new" tools (can you say VoltDB, Aster Data (Teradata now!), nPario "Platform". Hadoop, MapReduce, Cassandra, R, HANA, Paradigm4, MPP appliances etc. which are all cool and hot at the same time) and companies will enable users to collect, store, access and manipulate much larger datasets (petabytes).



For users, the level of noise will be now much bigger than before (and SNR - Signal-to-Noise ratio will be lower), because BDA is solving a HUGE (massive amounts of data are everywhere, from genome to RFID to application and network logfiles  to health data etc.) backend problem, while users interact with front-end and concern about trends, outliers, clusters, patterns, drilldowns and other visually intensive data phenomenas. However, SNR can be increased if  BDA technologies will be used together and as supporting tools to the signal-producing tools which are ... Data Visualization tools.



Example of that can be a recent partnership between Tableau Software and Aster Data (Teradata bought Aster Data in March 2011!). I know for sure that EMC trying to partner Greenplum with most viable Data Visualizers, Microsoft will integrate its PDW with PowerPivot and Excel and I can assume of how to integrate Spotfire with BDA. Integration of Qlikview with BDA can be more difficult, since Qlikview currently can manipulate only data in own memory. In any case, I see DV tools as the main attraction and selling point for end-users and I hope BDA vendors can/will understand this simple truth and behave accordingly.



Permalink: http://apandre.wordpress.com/2011/01/16/bigdata/

20110109

Donald Farmer moved from Microsoft to Qliktech

I never saw before when one man moved from one company to another, then 46+ people will almost immediately comment on it. But this is what happened during last few days, when Donald Farmer, the Principal Program Manager for Microsoft BI Platform for 10 years, left Microsoft for Qliktech. Less than one year ago, Donald compared Qlikview and PowerPivot and while he was respectful to Qlikview, his comparison favored PowerPivot and Microsoft BI stack. I can think/guess about multiple reasons why (and I quote him: "I look forward to telling you more about this role and what promises to be a thrilling new direction for me with the most exciting company I have seen in years") he did it, for example:

  • Microsoft does not have a DV Product (and one can guess that Donald wants to be the "face" of the product),

  • Qliktech had a successful IPO and secondary offering (money talks, especially when 700-strong company has $2B market capitalization and growing),

  • lack of confidence in Microsoft BI Vision (one can guess that Donald has a different "vision"),

  • SharePoint is a virus (SharePoint created a billion dollar industry, which one can consider wasted),

  • Qlikview making a DV Developer much more productive (a cool 30 to 50 times more productive) than Microsoft's toolset (Microsoft even did not migrate the BIDS 2008 to Visual Studio 2010!),

  • and many others (Donald said that for him it is mostly user empowerment and user inspiration by Qlikview - sounds like he was underinspired with Microsoft BI stack so is it just a move from Microsoft rather then move  to Qliktech? - I guess I need a better explanation),


but Donald did explain it in his next blog post: "QlikView stands out for me, because it not only enables and empowers users; QlikView users are also inspired. This is, in a way, beyond our control. BI vendors and analysts cannot prescribe inspiration". I have to be honest - and I repeat it again - I wish a better explanation... For  example, one my friend made a "ridiculous guess" that Microsoft sent Donald inside Qliktech to figure out if it does make sense to buy Qliktech and when (I think it is too late for that, but at least it is an interesting thought: good/evil  buyer/VC/investor will do a "due diligence" first, preferably internal and "technical due diligence" too) to buy it and who should stay and who should go.

I actually know other people recently moved to Qliktech (e.g. from Spotfire), but I have a question for Donald about his new title: "QlikView Product Advocate". According to http://dictionary.reference.com/ the Advocate is a person who defends, supports and promotes a cause. I will argue that Qlikview does not need any of that (no need to defend it for sure, Qlikview has plenty of Supporters and Promoters); instead Qlikview needs a strong strategist and visionary

(and Donald is the best at it) who can lead and convince Qliktech to add new functionality in order to stay ahead of competition with at least Tableau, Spotfire and Microsoft included. One of many examples will be an ability to read ... Microsoft's SSAS multidimensional cubes, like Tableau 6.0 and Omniscope 2.6 have now.

Almost unrelated - I updated this page:  http://apandre.wordpress.com/market/competitors/qliktech/

Permalink: http://apandre.wordpress.com/2011/01/09/farmer_goes_2_qlikview/