Panopticon


Datawatch published today its 2013 (ending 9/30/13) yearly and quarterly results and its YoY growth is impressive 16% (2013-over-2012), which is better then TIBCO (less then 13%) ! See Earnings Call Transcript here: http://seekingalpha.com/article/1849661-datawatch-corporations-ceo-discusses-q4-2013-results-earnings-call-transcript?part=single and webcast available here: http://www.investorcalendar.com/IC/CEPage.asp?ID=171788

Since Datawatch bought recently well-known swedish Data Visualization vendor Panopticon (which had 112% YoY in 2012!) for $31M in stock, Panopticon’s sales for a first time added to Datawatch sales (at least $1.5M revenue per quarter), total Datawatch quarterly revenue (as expected) grew to almost $9M per quarter and to $30.3M per fiscal 2013 (ending 9/30/13).

You can compare “moving” Datawatch YoY index for last 6 quarters vs 2 Top DV Performers (Tableau above 70% YoY, Qlikview above 20%), vs similar (YoY-wise) DV Vendor (Spotfire about 12%) and finally vs 2 Traditional BI Vendors (Microstrategy and Actuate). The thickness of lines reflects Vendor’s ttm (the revenue for Trailing Twelve Months) – click on Image to Enlarge:

Year-over-Year Growth for Trailing Twelve Month (YoY4ttm)

Year-over-Year Growth for Trailing Twelve Month (YoY4ttm)

Datawatch founded in 1985(!), public (traded on NASDAQ as DWCH) since 1992; it has 44000+ customers (including 99 of Fortune 100) and 500000+ end users. Datawatch management team is experienced in BI space and includes veterans from IBM, Applix, Cognos etc. In last 3 years (since 10/1/10) DWCH shares increased in value more then 10 times:

ReturnIn3Years

2nd V for BigData: Data Variety.

The first version of main Datawatch software, called Monarch Professional was released in 1991 and developed by Math Strategies. Overtime Datawatch added a lot of features to this ETL software, including the support for the broadest variety of data types and data sources simultaneously—including traditional structured relational databases, semi-structured sources like reports, PDF files, EDI streams, print spools and documents stored in files systems or enterprise content management systems, with a  new mix of unstructured data such as machine data and social media stored in Big Data solutions or streaming directly from a host of real-time applications.

Datawatch Desktop does ETL from all above Data Sources and then extracts those data into Variety of Standard Formats: Excel spreadsheets, Access Databases, PDF reports, into Panopticon Workbooks etc. Simple example of how Monarch 11 does it you can see here:

or more professional and free video training you can find here: http://www.datawatch.com/information-optimization/item/196-guided-tour-monarch

The latest release of Monarch Professional is in version 12 and it has the new name as Datawatch Modeler; it also integrated and bundled together with Panopticon Desktop Designer under new name Datawatch Desktop and that bundle is available for $1895. As a result Datawatch created for itself an excellent up-sell opportunity: current customers on maintenance can trade-up to Datawatch Desktop for $366 (it also includes first year maintenance) – this is 5 times cheaper than Tableau Desktop professional. My understanding that maintenance of Datawatch Desktop is 22% per year of its price but you may get a better deal.

Monarch11

Datawatch Modeler v.12 has new Core engine with 16 External Lookups (was 9 in version 11), 512 Columns In Table (was 254), 100 Multi-Column Regions (was 40), Optimized for modelling large inputs Data Preview (work with first 100 records), has new PDF Engine, 10GB Internal Database size (was 2GB), Utilized 4 Cores for DB operations (was 2).

1st V for Big Data: Data Volume.

Math Strategies developed for Datawatch another tool – Monarch DataPump (recently renamed as Datawatch Automator or Datawatch Server – Automation Edition, currently in version 12). On 3/30/12 Datawatch acquired intellectual property for its underlying Monarch Report Analytics platform from Raymond Huger, d/b/a Math Strategies (Greensboro, NC).

Datawatch developed other editions of Datawatch Server:

  • Formerly Enterprise Server has new name now as Datawatch Server – Content Edition, version 12. Datawatch Server supports all Monarch functionality on server-side, integrates with web server(s) and related infrastructure, manages all users, their credentials, access rights, roles, privileges, user groups, manages and aggregates all content, data, data extracts etc.

  • Datawatch Server – Automation Edition (Data Pump) – automatically collects and refreshes all content, data and data extracts, both on-demand and on-schedule, manages all schedules etc.

  • Datawatch Server – Complete Edition includes Formerly Panopticon Server (manages all Data Visualizations and its users, converts Visualizations to web applications so they can be accessed through web browsers and HTML5 clients), Datawatch Enterprise Server and Data Pump.

V3

Theoretically Datawatch Server (with help from Datawatch Automator) can support up to 524 Petabytes (1015 bytes) of Data which I consider a very Big Data for 2013.

3rd V for Big Data: High Velocity

Datawatch/Panopticon in-memory data engine supports data visualization for real-time business dashboards and it has low-latency display of analytics that are based on streaming data as it arrives. This enables Datawatch to handle the demanding continuous-intelligence applications, where quick responses are required. This is a big differentiator. An in-memory, OLAP-based StreamCube is associated with each graphical display object. The system processes new data as it arrives, selects the subset of important data, recalculates the relevant sections of the model and refreshes the associated parts of the display immediately. The parts of the model and the display that are not affected by the new data are not touched. This is faster and more efficient than conventional data visualization tools that operate on batch-loaded snapshots of data, run less frequently, and then recalculate the model and rebuild the display for each iteration.

Somebody I know was able to refresh and REPAINT 25000+ datapoints per second per one Datawatch/Panopticon Chart and this is much faster then any competitor.

Datawatch platform integrated with message-oriented middleware, including ActiveMQ, Qpid, Sonic MQ and Tibco EMS. It has connectors to Complex Event-Processing platforms (CEP), such as kx kdb+tick, OneTick CEP, Oracle CEP, StreamBase Systems’ Event Processing Platform and Sybase Event Stream Processor. Datawatch also has interfaces for retrieving data from time series databases, conventional relational and columnar databases, files, Open Data Protocol (OData) sources and in-memory DBMSs. It can be customized for proprietary data sources (recent example is a Visualization Accelerator for Splunk) and even embedded within other applications. Like other leading data visualization tools, it  supports a wide range of charts. It has a development studio (Desktop Designer) for designing and implementing dashboards, and HTML5-based clients/support for mobile applications.

4th and most desirable V: Data Visualization

Datawatch is trying to get into Data Visualization (DV) field and it has potentials to be a 4th major Vendor here: it has a competitive DV Desktop, a competitive DV Server, an excellent HTML5 Client for it and set of differentiators like ready-to-use 3V triplet of features (see above) for Big Data and real-time DV. Datawatch Designer supports rich set of Graphs, Charts, Plots, Maps, Marks and other types of Visualizations, for example:

  • TIME SERIES Graphs: Candlestick, Horizon, Line, Needle, OHLC, Spread, Stack Area, Stacked / Grouped Needle, Table with Micro Charts, Sub Totals & Grand Totals, Timeseries Combo Charts, Timeseries Scatter Plot.

  • STATIC & TIME SLICE Graphs: Bullet, Heat Map, Heat Matrix, Horizontal/Vertical Bar, Horizontal/Vertical Dot Plot, Multi-Level Pie Chart, Numeric Line, Numeric Needle, Numeric Stacked Needles, Scatter Plot, Shapes / Choropleth, Surface Plot, Surface Plot 3D, Table with Micro Charts, Sub Totals & Grand Totals, Treemap.

  • many Visualization Demos still available here: http://www.panopticon.com/Advanced-Data-Visualization and here: http://www.panopticon.com/demos

3232161725_e648f09137_o

In my humble opinion in order to compete with leading DV vendor like Tableau I think that Datawatch needs a few gradual changes, some of them I listed below:

  • Gradually on as-needed basis add features which other 3 DV Vendors have and Datawatch does not (it needs serious R&D)

  • Create free Datawatch Public (cloud service) to make people to learn and compare it (similar to Tableau Public) and to win mindshare

  • Create Fee-based Datawatch Online cloud service (similar to Tableau Online and Spotfire Cloud services)

  • Add more DV-oriented Partners (similar to Qlikview Partner Program, which has now 1500+ partners)

  • Create fee-based Data Visualization Practice in order to help large clients to implement DV Projects with Datawatch Desktop and Server.

  • Add support for Visual Analytics and Data Science, including integration with R Library (similar to Spotfire’s S-Plus and TERR or at least the integration with R like Tableau 8.1 did today)

  • Add support for Storytelling, similar to what next versions of Tableau and Qlikview will have (soon) and communication abilities (similar to what Spotfire 6 has with TIBBR)

  • I may expand this list later as I see the fit, but Datawatch really has an unique opportunity here and large potential market!

Feedback 11/22/13 from multiple visitors of this blog:

I Quote the email from one of frequent visitors to my blog: “The fastest growing sales are in DV field (e.g. Panopticon revenue was 112% YoY in 2012). For example in 2006, when Qliktech’s Sales were $44M, its YoY was 81%; 4 years later, in 2010, when Tableau had $40M revenue, YoY was 106%, see it here: http://www.prnewswire.com/news-releases/tableau-software-doubles-revenue-with-2010-landmark-year-114913924.html and 4 years later, in 2014 history can repeat itself again if Datawatch will allow to unbundle its DV Products and sell them separately. Instead, currently Datawatch prevents its own salesforce to sell separately own DV products like Panopticon Desktop Designer (you may call it now as Datawatch Visualization Studio) and Panopticon Server (you can call it now as Datawatch Visualization Server). That artificial limitation has to be removed!” visitor said to me over email… All I can say: it is not my call… Additional links: 

Advertisements

In past Vikings discovered America, conquested or colonized parts of England, Russia, Ireland, Scotland, even Southern Italy and Iceland… But in 21st century (as far as this blog is concerned) Sweden became a Motherland of Data Visualization:

4SwedishDVVendorsLogosLet’s start with most famous Data Viking and most known Storyteller in Data Visualization field – prof. Hans Rosling from Karolinska Institutet and chairman of the Gapminder Foundation (in Stockholm). Gapminder’s team invented the popular and useful 6-dimensional Motion Chart and developed Trendalizer which was bought by Google in 2007, see it here: https://developers.google.com/chart/interactive/docs/gallery/motionchart . The recent example of Prof. Rosling Storytelling you can see  here:

In Stockholm you can find another Data Visualization Innovator – Panopticon is a leader in Complex Even Processing and real-time Visual Analytics. Among other innovation here is the example of Panopticon’s invention (by its senior developer Hannes Reijner) of Horizon Chart, see sample here:

HorizonGraph

and short video about it here:

In 2012 Panopticon posted 112% Year-Over-Year revenue growth (comparable with Tableau). In 2013 (the all stock deal closed by the end of September, 2013.) Datawatch bought Panopticon for $31.4M and I assume it will try to move some R&D from Sweden to Chelmsford, MA.

In  Göteborg/Gothenburg you can find R&D office of another DV Leader – Spotfire with 60+ Data Vikings. In 2007 TIBCO bought Spotfire for $195M but even now in 2013 unable to move R&D into USA. So now Spotfire actually has 3+ main offices: TIBCO Corporate Headquarters in California, Spotfire Headquarters in Somerville, MA (estimate is 15% of Spotfire workforce) and main R&D office in Sweden. In addition, lately TIBCO choose the strategy to buy rather then build new features, for example, just in 2013 they added to Spotfire portfolio the following new companies and as result they have even more distributed R&D team now:

  • Extended Results (PushBI) in Redmond, WA
  • MAPORAMA in Paris, France
  • StreamBase Systems, Inc. in Waltham, MA

As a result, despite the fact that Spotfire 6 is the most mature Data Visualization platform on market, people in TIBCO Corporate Headquarters running into risk of do not have enough knowledge of their own major Intellectual Properties.

In southern Sweden – Lund, we can find Swedish Headquarters of the major DV Leader – Qliktech, who occupied almost half of Data Visualization market in terms of sales. At least 140 Data Vikings located in Lund and may be another 200 elsewhere in Sweden. Qliktech’s Data Vikings are major innovators with features like the fastest in-memory Data Engine, most natural Visual Drill-down, Associative Query Language to name a few. This also presents a major problem for Qliktech, because they have Headquarter in Radnor, PA (where only 150+ employees work (estimate), which is less then 10% of Qliktech’s workforce!), Main marketing, sales and support office in Newton, MA (estimate: less then 5% of workforce) and most R&D in Lund (estimate: at least 10% of workforce).

This means that almost 500 technically advanced Data Visualization experts (engineers, developers, architects etc., which is at least 23% of total Qliktech+Spotfire workforce) are still in Sweden. The simple observation of Tableau’s TCC13 conference in September 2013 shows that Tableau’s top managers and officers know their product deeper and more intimately then their counterparts in Qliktech and Spotfire. That is very easy to explain: because 650+ Tableau’s employees (almost 65% of their workforce and most developers, managers and officers) work in the same Main HQ office in Seattle, WA and they obviously talking to each other in-person and often!

My humble advice to Qliktech, Spotfire and Datawatch is simple – gradually relocate as much Data Vikings from Sweden to appropriate headquarters in USA or find and hire local american equivalents of those Swedish geniuses…

As a background for this advice, please consider this information (updated on 11/17/13): statistics of job openings clearly showing that all 3 DV Leaders keep doing (by inertia) what they did in past with only difference that it worked recently for Tableau and does not work for Qliktech and Spotfire. Here are specific examples:

  • Tableau has 176 job openings (much more then Qlikview (only 80) and Spotfire(only 18) combined)!

  • 97 (55%) of Tableau openings are in Seattle, more then half of Tableau’s openings are engineering and technical positions!

  • Qliktech has 17 (21%) positions opened in Lund, only 10 (12%) in Radnor and 4 (5%) in Newton, MA. Only 11 (14%, 9 times less then at Tableau in absolute numbers) Qliktech’s openings are engineering and technical.

  • Spotfire has only 18 openings (1 in Göteborg, 5 in CA, 4 in MA) and only 4 Spotfire’s positions (out of 18, 22% that is) are engineering or technical.

This statistics clearly showing that neither Qliktech no TIBCO see the wrong pattern and huge problem here and that can be a reason for disruption in the future and the gradual  relocation of Data Vikings is only way to prevent the danger… And of course, if you can afford, find and hire equal talents in USA Headquarters then by all means keep geniuses in Sweden without relocation which is a half-similar to what Tableau does (HALF is because Tableau historically does not need to maintain the significant R&D office outside of USA)!

Yesterday I got invited by Qliktech for their semi-annual New England QlikView Boston User Group meeting. It was so many participants, so Qliktech was forced to hold the Keynote (of course the presentation and the demo of Qlikview.Next) and 4 cool presentations by Customers and Partners (Ocean State Job Lot, Analog Devices, Cybex and Attivio) outside of its own office but in the same building  on the 1st floor @Riverside Offices in Newton, MA @Rebecca’s Cafe.

It was plenty of very excited people in a very large room and very promising demo and presentation of Qlikview.Next, which actually will not be generally available until 2014. Entire presentation was done using new and capable HTML5 client, based on functionality Qliktech got when it bought NComVA 6 months ago.

I was alarmed when presenter never mentioned my beloved Qlikview Desktop and I when I asked directly about it, the answer shocked and surprised me. One of the most useful piece of software I ever used will not be part of Qlikview.Next anymore. As part of Qlikview 11.2, it will be supported for 3 years and then it will be out of the picture! I did not believe it and asked one more time during demo and 2 more times after presentation in-person during Networking and Cocktail Hour inside Qliktech offices. While food and drink were excellent, the answer on my question was the same – NO!

LeafsAndNeedlesOnGrass

I have the utmost respect for very smart software developers, architects and product managers of Qlikview, but in this particular case I have to invoke 20+ years of my own advanced and very extensive experience as the Software Architect, Coder and Software Director and nothing in my past can support such a decision. I do not see why Qlikview.Next can not have both (and we as Qlikview users need and love both) Qlikview Desktop Client and Qlikview HTML5 client?

I personally urge Qliktech (and I am sure the majority of 100000+ (according to Qliktech) Qlikview community will agree with me) to keep Qlikview Desktop client as long as Qlikview exist. And not just keep it but 1st,  keep it as the best Data Visualization Desktop Client on market and 2nd, keep it in sync (or better ahead) with HTML5 client.

In case if Qlikview Desktop will disappear from Qlikview.Next, it will be a huge gift to Tableau and Datawatch (Spotfire Cloud Personal will no longer have access to the Spotfire Analyst desktop product and therefor Spotfire Cloud Personal is making a similar (partial) mistake as Qlikview.Next)

.

tableau_cmyk

Tableau recently invested heavily into progress of all variations of Tableau Desktop (Professional, Personal, Public, Online, Free Reader) including (finally) migration to 64-bit and even porting Desktop to MAC, so it will instantly get the huge advantage over Qlikview in desktop, workstation, development, design, debugging, testing, QA  and offline environments.

DATAWATCH CORPORATION LOGO

It will also almost immediately propel the Datawatch as a very attractive contender in Data Visualization market, because Datawatch got (when they bought Panopticon this year) the extremely capable Panopticon Desktop Designer

Panopticon_Data_Visualization_Software_logo_file,_800x155,_Real-Time_Visual_Data_Analysisin addition to its own very relevant line of products.

Again, I hope I misunderstood answer I got 4 times during 4-hour meeting and during follow-up networking/cocktail hour or if understood it correctly, Qliktech will reconsider, but I will respect their decision if they don’t…

So I have to disagree with Cindi Howson (as usual): even if “QlikTech Aims To Disrupt BI, Again“, it actually will disrupt itself first, unless it will listen me begging them to keep Qlikview Desktop alive, well and ahead of competition.

SunsetOnCapeCod102413

You can find in Ted Cuzzillo’s article here: http://datadoodle.com/2013/10/09/next-for-qlik/ the actual quote from Qliktech’s CEO Lars Björk: ““We can disrupt the industry again”. My problem with this quote that Qliktech considers itself as the insider and reinventor of the dead and slow BI industry, while Tableau with its new motto “DATA to the people” is actually trying to be out of this grave and be inside own/new/fast growing Data Visualization space/field/market, see also blogpost from Tony Cosentino, VP of Ventana Research, here: http://tonycosentino.ventanaresearch.com/2013/09/21/tableau-continues-its-visual-analytics-revolution/#!

You can see below interview with Time Beyers, who has own doubts about Qlikview.Next from investor’s point of view:

Basically, Qlikview.Next is late for 2 years, it will not have Qlikview Desktop (big mistake), it still does not promise any Qlikview Cloud services similar to Tableau Online and Tableau Public and it still does not have server-less distribution of visualizations because it does not have free Qlikview Desktop Viewer/Readers similar to free Tableau Reader. So far it looks to me that QLIK may have a trouble in the future…