Guiding Web-Based Analytics with Data Publishing
(Stephen E. Arnold Information World Review columnist March, 2002)
"With a company's data doubling every four to six months, analyzing the information is often the difference
between making a profit or losing money," says Andy Coutts, president and chief executive officer of
Databeacon, Inc. (Ottawa, Ontario)and former Software AG executive. "Managers need a way to get data,analyze it,
and make better decisions. They don't have time to figure out software or fiddle with complicated equations."
Databeacon, a seven-year-old firm, is one of the leaders in data publishing. According to Nathan Rudyk,
vice president of marketing, at Databeacon, "Data publishing, is taking data from somewhere in the organization,
personalizing it with a series of reports for an individual or community of individuals, and providing a
point-and-click Web-based tool to allow further analysis of that data so people can answer their own questions."
Unlike traditional data mining and statistical software, Databeacon makes slicing and dicing data dead simple
within Internet Explorer or Netscape Navigator. Databeacon's Java-based approach makes packaging filtering,
maths (British term?), graphing, and extracting particular facts an easy exercise for most business professionals.
No first (more Brit terminology?) in mathematics required, thank you.
Databeacon has a Web site that shows off its data publishing technology at
http://www.storydata.com. The site aggregates information from a wide
range of sources to be shaped, analyzed, charted, graphed, and subjected to mathematical functions. The site
provides anyone who agonized through a course in statistics and database analysis with a sudden insight, "This is
analytics done properly. " Databeacon has banished cryptic commands and inscrutable wording for both its novice and
advanced interface. Pick a category. Click. Compare it to a time series. Click. Graph the results.
Click. Combine data from two disparate sources. Click click.
Databeacon has smashed the walls of complexity that prevent many managers from looking carefully at the
gigabytes of data swirling around them. A Ziff-Davis study of information in organizations reported that
refining and analyzing information was one of the top challenges in knowledge management at organizations. Gartner,
Forrester, the Giga Group, and other consultancies predict that data mining will reach $4 billion by 2004.
The Abeerdeen Group (Boston, Massachusetts) estimates that the market for Web-based analytics will reach $700 million
in 24 months, up from almost negligible sales today.
"Web analytics is where the data analysts are heading," says Raghu Rao, president of InfoZen, Inc.,
a Germantown, Maryland information services firm. "The real impact of the Dot Com era is yet to be felt.
Users want to access data, perform whatever analysis is appropriate, and conduct their business in
a persistent, pervasive information environment. It is only logical that advanced data mining and
analysis functions would find their way to the browser."
"Databeacon's browser-based end-user Java application, the Databeacon Insight viewer, takes most
of the mystery and virtually all of the complexity out of multidimensional data analysis," says Mr. Rudyk.
"Unlike the tools from Cognos, IBM, SPSS, and other established analytics providers, we bring the power
of data analysis in just one click, with no requirement to install end-user software. Updating data is
nothing more than connecting to the network and retrieving the most current block of information."
"To simplify a complex, technical specialty such as OLAP is a formidable challenge," says Mr. Rudyk.
"Databeacon preserves the rigors of OLAP, including the construction of complex "cubes" of information.
We provide backend tools for the information technology and database administrators, a server component
to construct and compress data blocks or cubes, and manage the publishing and dissemination processes.
Most people see the Java-enabled analysis tool in the browser. That's just the tip of the iceberg.
Nine-tenths of our innovation makes customized, real-time data publishing something any organization
can afford to do."
The content of a block or cube of data usually comes from one or more databases, legacy mainframe
applications on a variety of Unix systems, and proprietary software scattered at different locations.
The "cubes" are usually multidimensional so that anyone looking at the data can view different
relationships, flipping between views as insights emerge from various mathematical processes.
"We can handle just about any databased situation," says Mr. Coutts. Most cubes contain information
about different aspects of a business or a particular subject such as Internet visits monitored
by NetRatings or mobile telephone usage at Telephia, Inc., or financial transactions that come
from Actuate, Inc.'s reporting software."
Multidimensional data manipulation is sorting, flipping, slicing and dicing information about
different aspects of information, say, manufacturing output, sales in the same time period, and
other factors. The idea is that relationships can be discovered by looking at information first
one way and then another.
Because Databeacon pushes the cube as well as its Insight viewer to the user's personal computer,
server demand and bandwidth requirements are held in check. However, users want to update their cubes.
Databeacon provides technology that allows a user to obtain a refreshed cube as soon as it is available.
With much of the updating and notifying process automated, information technologists shave hours,
even days of work off a routine OLAP task.
For analysts who have the expertise to look at even larger, more complex datasets, Databeacon
supports cube-to-cube analysis within a browser.
Pfizer Italiana's manufacturing and distribution plant in Latina, Italy, a suburb of Rome.
The firm's information technology staff found itself preparing reports about the production
of about 10 million shipments each month. Business analysts requested frequent updates in
order to monitor and reconcile orders with production of the pharmaceuticals.
The Pfizer Italiana information technology department found itself repeating certain data
extraction and normalization processes. Each business analyst required a different segment
of manufacturing, marketing, and financial data. The data were then saved as Excel files and
provided to colleagues for further manipulation and study.
The Pfizer Italiana information technology department found itself repeating certain
data extraction and normalization processes. Each business analyst required a different
segment of manufacturing, marketing, and financial data. The data were then saved as
Excel files and provided to colleagues for further manipulation and study.
Databeacon provided its server-based suite of software to allow Pfizer Italiana's
staff to shift from traditional, hard coded extraction and cube building mechanisms
to Databeacon's point-and-click approach. Once a data cube was extracted and
processed by the Databeacon system, the cube was provided to the Pfizer Italiana
manager for analysis in the browser-based Insight viewer interface. When a Pfizer
Italiana manager is out of the office, the system allows access or update manufacturing
and sales marketing data as needed.
Pfizer Italiana uses Databeacon's Java software to extract data from a proprietary
ERP (Enterprise Resource Planning) system, assorted AS400-based applications and
the company's RS6000 Oracle database. Databeacon provides suite of tools for data
extraction, preparation, and analysis," says Mr. Rudyk. "We knew that one of the
best ways to exploit Java and Web technology was to make the complicated steps
associated with advanced data analysis easier."
According to Pfizer Italiana, the Databeacon solution reduces the amount of time
required to assemble customized datasets. More importantly, business analysts and
other professionals at Pfizer Italiana are able to access data more quickly and
more easily than before. According to Luca Minchella, custom data modules can be
assembled and provided to a colleague the same day a new request is made. For
routine datasets, the process is largely automated. Productivity in the information
technology department has risen sharply. [Note: from article and Databeacon newsletter]
Databeacon's software is the result of about seven years of development.
Within the last 12 months, the company has seen explosive growth in the licensing
of its technology from traditional consumers of data mining and analysis technology
to content aggregators and publishers engaged in content aggregation and market research.
"When we looked at our install base at the start of 2001, we had a half dozen data and
information publishers using our technology among a direct customer base of 60
(Databeacon has another 500 indirect customers via its growing list of Independent
Software Vendor (ISV) partners such as Actuate). By the end of the first quarter 2002,
we have doubled the number of information publisher customers and we have many more
deals with these types of customers in our pipeline," says Mr. Rudyk, an affable
entrepreneur with several high-technology successes on his cv.
Databeacon's "new way" of doing analytics includes providing a fully documented API
(application programming interface) so that its customers, including names such as
Pfizer, Actuate, Baxter Healthcare, Nielsen//NetRatings, and Telephia, can implement
an automated, customized data publishing system. The Databeacon PageEnhancer API
allows the Insight viewer to be embedded in a portal page or be delivered as a
customized report to a subscriber as a for-fee data service.
The key points of difference between traditional OLAP systems and Databeacon's
boils down to three: First, Databeacon uses the browser environment in order to
make access to functions easy, provide refreshed or new data sets to analysts or
subscribers, and integrate seamlessly with the user's preferred computing environment.
Second, for the information technology staff, Databeacon provides a solution that
permits automation of basic data extraction and cube building functions on a low-cost
machine. The point-and-click approach reduces the time required to perform necessary
routine functions and provide real-time access to data.
Third, the Databeacon solution provides a way to make OLAP and advanced analytics
fit the different work patterns of analysts and their fast-changing requirements
for a particular set of multidimensional data that can be updated in real-time.
Databeacon tools are accessible in any device that supports a browser interface
and a network connection. "Databeacon's server backend allows administrators to
automate most extraction, cube processing, and distributing functions" says Mr. Coutts.
"If there is a special function a client needs, our PageEnhancer API [application
programming interface] allows a programmer to plug in that functionality."
How will Databeacon compete with such giants as Oracle and IBM, two companies
focusing more resources on business intelligence?
According to Mr. Coutts, Databeacon has a trick or two embedded in its technology.
"We are able to allow any user to navigate through an information space consisting
of several cubes of related data by following intelligent links from one cube to
another. For example, a television ratings application syndicated to subscribers
at various programming companies might provide a cube that summarizes worldwide data.
The system can generate and provide to the research firm's customers additional cubes
that provide detailed information for various territories. Any subscriber to the service
can start with a broad outlook and then jump to any of the territorial cubes for detailed
analysis of behaviors in a specific region or city. Our cube-to-cube selects are extremely
powerful and all the analysis takes place in the user's browser without any additional
server or network load."
Databeacon's mature API allows a licensee of the Databeacon tools to turn on or turn off
specific features; for example, a particular view of a data set. The PageEnhancer API can
be used to integrate the data analysis tool into a portal or Intranet environment. The
architecture of the Databeacon system makes it possible for a commercial software company
to embed Databeacon's analytic functions into other products. Just as Verity's search
engine can be placed inside Adobe Acrobat, the Databeacon data analysis tool can reside
within third-party applications, such as Actuate's proprietary reporting software, and be
integrated into Web services crafted from tools from Microsoft, IBM, and BEA Systems, among others.
Databeacon integrates tightly with other applications such as Microsoft Office. In addition,
the Databeacon approach extends the type of functionality that most business professionals
are familiar with from their work with Word or Excel. Databeacon offers enhanced functions
including a complete range of graphs, dynamic data slicing, one-click drill down and trend
analysis, the ability to bookmark specific points in a data set, and a cube navigation map.
"For most business people, the tools in Excel or Word are sometimes hard to use with
multidimensional data," says Mr. Rudyk. "We have worked to make our features and functions
a matter of pointing-and-clicking to get access to extremely powerful data analysis functionality.
In fact, we offer our data publishing customers a totally customizable interface if, for example,
a junior manager requires a limited set of functions. For the analyst professional, the entire
range of Databeacon functionality can be activated."
Databeacon occupies new ground quite a distance from the expert-only data manipulation and analysis
tools and built-in graphing and analysis functions in PowerPoint.
Databeacon has competition. Similar tools are becoming available from such companies as Spotfire, Inc.
(Boston, Massachusetts, Göteborg, Sweden, and Tokyo, Japan), Business Objects, Inc. (Maidenhead, Berkshire),
and such companies as Oracle Corp., IBM, and SPSS, Inc. among others.
Visual Mining (Rockville, Maryland) http://http://www.visualmining.com
Product allows a system professional to build chart views. Menu choices are required to create a chart view.
Visual Mining is a server based application. The Java client pulls data, renders it, and allows the user some
interactivity. The approach creates a load on the data server. Mouse over on graphic objects-for example, a bar
in a bar chart-and the data appear in a pop up. Databeacon offers a similar function. Prices range from about
£500 to £5,400 for Visual Mining's software.
Through the last nine months, Databeacon says that it has watched as its information publisher customer base double.
The company is now focusing on providing its tools to market research firms, data publishing companies in the health
and e-government sectors, and financial institutions who make high-value research available to clients worldwide.
"The combination of our Application Programming Interface and Databeacon's core functionality allows a data
publisher to produce more quickly and economically Web-accessible, browser-based analysis and reporting," says Mr. Rudyk.
"Actuate (San Francisco, California) has integrated our technology into its range of reporting products in order to
generate new revenue from new analytical products. With Databeacon what we've done is built a data publishing,
analysis and reporting dissemination system that breaks the logjam often associated with OLAP and business intelligence
in large organizations."
AC Nielsen's Nielsen/NetRatings (a unit of VNU) is using Databeacon technology to enhance its analytic services
to its Global WebWatch clients. Adds Mr. Rudyk, "Our technology opens new avenues for analytics products. A client
adds a server (as a Java product Databeacon runs on most industry-standard platforms), loads our software,
and begins streamlining the extraction, cube building, and delivery processes in a matter of days.
With server-based licensing versus license fees tied to a set number of users, we offer a fixed cost to our clients.
We know we put cash on our customer's bottom-line in a month or two."
Can Databeacon survive the inevitable competitive pressure that comes when others discover an underserved niche?
For publishers and aggregators of numeric and factual data, Databeacon offers a way to slash the cost of new products
and put new revenues on the balance sheet. With demand for Web-based analytics starting to grow, the company may be in
the right market, at the right time, with the right technology.
1. The Ziff Davis data are located at
2. For a more about OLAP, explore the links at
1565 Carling Avenue, Suite 300
Canada, K1Z 8R1
Telephone: (613) 729-4480
Toll-free (Canada and United State only): (888) 921-8360
Fax: (613) 729- 6711
Stephen E. Arnold
Arnold Information Technologies
Postal Box 320
Harrod's Creek, Kentucky 40027
Voice: 502 228 1966
E mail: firstname.lastname@example.org
Web site: http://www.arnoldit.com
[ Top ] [ AIT Home ] [ Beargrass ] [ Site Map ]