DV Competitors
Traditional BI Vendors – they are behind in Data Visualization (DV) but trying hard to catch-up with DV Leaders:
- SAP – they think that buying BI (e.g. Business Objects) and DV vendors can make them DV leaders.
- IBM – bought 24 companies during 2006-10 for $14B. IBM does not understand that Java is the wrong bet.
- SAS – mindshare leader, made the same Java mistake as IBM and Oracle.
- Oracle – losing their BI share despite good company-wide financial results; mishandles many purchases, like IRI, Hyperion, Sun etc.
- Microstrategy – pure BI player with good reputation but far behind in DV market.
- Microsoft – has the best BI stack (SQL Server, SSAS, SSRS, BIDS, PowerPivot, Excel 2010 etc.) but has no DV product due corporate stupidity.
DV Leaders
- Qlikview – fast growing, with best DV functionality, especially visual drilldown, multitude of clients (including major smartphone platforms) but has limited reach for large enterprises. Qlikview has an excellent in-memory (64-bit) columnar database and so called AQL (“Associative Query Language”).
- Spotfire – lost(in 2007-10) its DV focus due TIBCO’s product diversity, but still has the best web client, analytics (SPlus, IronPython) and API; still the best option for large enterprises. Spotfire has excellent in-memory (64-bit) data engine and unlimited ability to scale-up to disk.
- Tableau Software – the fastest growing DV Vendor (123% YoY as of November 2010), easiest for end users with minimal need for IT, has best access to OLAP cubes and best Pivot Control among DV vendors. Recent Tableau 6.0 release has a very capable and fast in-memory (64-bit) data engine.
Other DV Vendors
- Visokio – Omniscope is one of the most advanced DV product, but implemented in Java (also without in-memory columnar DB), which seriously limited its ability to handle large datasets.
- Pagos – its SpreadsheetWEB and upcoming Data Visualizer make it one of the most interesting BI and DV vendor, but it needs breakthrough in marketing and sales
- Panorama – MDX inventor, has Microsoft and Google as OEM partners, best integration with OLAP
- Information Builders – One of the oldest BI vendors, has too much inertia from existing customers. Information Builders OEM partners with Advizor Solutions and delivering Data Visualization to its customers with Advizor product. David Raab has a very high opinion about Advizor.
- Actuate – One of the few successful Open Source BI Vendors
- Panopticon – very good DV vendor but caged in Java jail the same way as Visokio/Omniscope
- more competitors – Vizubi, BonaVista Systems, Pentaho, Dundas, etc. list can be expanded…
Very interesting content. it is very useful to have both a global and accurate picture of the DV market. i am not a software developer so it would be nice if you could explain (in a future post) why java is the wrong bet and compare the java option vs other options. thanks, patrice
Hi Patrice:
thanks for your kind words. Many people asking me about Java and I usually trying to delay the answer, because almost all discussions of this topic will eventually look like religious war where all infidels have to die.
2+ reasons preventing Java-based DV tools to be competitive:
1. Owners of Java destroyed Java’s future with own hands: Sun got $1.6B from Microsoft and Oracle recently sues Google over use of Java. In 1st case Java disappears from Visual Studio and in 2nd case Google can as well remove Java from Android. Guess what: most developers developing either for Microsoft or for Google environment. If you will add that Apple (today it had $290B market cap) does not endorsing Java for Apple’s environments (can you say Objective C?), than now we have 3 “leaders” of the software world are rejecting Java. I am just following the money here…
2. Many vendors tried to use Java for Data Visualization, best of them is Visokio. Their wonderful product – Omniscope has more features then most DV competitors, but due limitations of Java they cannot handle large datasets and this making their appeal very limited. Java-based DV tools from other vendors are behind Visokio…
Hi! Does Panopticon have the same problems as Visokio? Problems with large datasets? I’ve been trying to find pros and cons of their product and continue to search through your excellent blog for any relevant comments. If you happen to have a link to something, I’d love it!
Best,
Rob
Rob:
Thanks for the excellent question and I think it deserved a detailed answer, because I am getting similar questions all the time and it means that answer on it has a very public value. Here you go:
I like Panopticon, but I cannot recommend it because it’s UI and Data engine cannot compete with Qlikview, Tableau and Spotfire. I do not have enough time (busy with work, blog, family etc.) to cover everybody in huge and fast growing Data Visualization (DV) market and frankly most of small and not competitive players will eventually disappear or will be irrelevant.
This is a list of DV vendors who will be around for a while:
1. Qlikview (IPO made them even more relevant, but they have some growth and management problems – I hope they can solve it). Qlikview also has known problem with upcoming “big data” wave: their datastore has to be in RAM and it does not take advantage of cheap huge hard disks and SSD as virtual memory – so as of now Qlikview is not very scalable in terms of size of datasets (Spotfire can scale to disk and latest version of Tableau has some of that scalability too). Both Qlikview and Spotfire completely and mistakenly ignore Microsoft backend (SSAS and new Tabular mode from SQL Server 2011 and PowerPivot as the best in-memory columnar DB engine)
2. Spotfire has an excellent feature set and a very good corporate parent (TIBCO), but their sales and partnership programs needs to be improved ASAP. Also I wish they will improve UI (they as Tableau have Wrong User Interface [compare with Qlikview UI] for Drill-down Functionality and No MDI support in Dashboards, all charts share the same window and paint area); Spotfire backend is too messy, for example they still use Tomcat (and Java, which slowing them down) as Application Server; I told them that many times but they (I guess somebody TIBCO headquarters) think they can live for a while with these problems. I completely disagree and I think it will bite them big in a long run,
3. Tableau is growing faster then anybody but I still have some old problems with Tableau (it is possible that somebody will buy them soon, like Teradata or Microsoft):
- No MDI support in Dashboards, all charts share the same window and paint area;
- Wrong User Interface (compare with Qlikview UI) for Drill-down Functionality;
- Tableau’s approach to Partners is from stone ages;
- Tableau is 2 generations behind Spotfire in terms of API, Modeling and Analytics
4. Other “me too” vendors have and will have impact on DV market:
- Microsoft has the best backend (SQL Server 2011, PowerPivot) for DV applications;
- Visual Insight from Microstrategy (9.2) is promising;
- SAS, IBM, SAP and Oracle trying hard to be DV-relevant, spending huge money on it, but so far all of them failed to impress me.
5. I already mentioned on this blog a while ago that Google has a lot of excellent DV components: Google Analytics, Google Maps, Google Earth, Motion (inherited from Hans Rosling), Timeline (recreation of some well-designed financial Charts), Sparkline and Scatter charts, Google Public Data Explorer, Google Fusion Tables etc. Unfortunately Google keeps ignoring DV market (by not producing DV tool [it can be part of Google Apps or separate product]) for similar reason to Microsoft reasoning: they think it does not fit their business model. I happened to completely disagree with both Google and Microsoft on this stupid reasoning, but they will not listen me anyway…
In terms of Visokio: I love Visokio, but as I said before Visokio has a few problems:
1) Omniscope from Visokio is written in Java and therefore it never will be competitive in terms of large datasets.
2) Visokio is extremely small company, driven by their large and loyal customers and therefore has limited capacity to execute. For example it took year and half to build the next version (2.6) of Omniscope, but hopefully it will be finally released by the end of September of 2011.
Thank you, Andrei!