Vendor
Genesis
Ardent
Software, formerly known as VMARK Software, was founded in the mid-1980's
to develop and market extended relational database systems. The product
originally ran on UNIX operating systems, and soon the systems were also
licensed for use as an embedded part of business applications.
In
February 1998, Ardent merged with Unidata Inc., a marketer and developer
of relational databases, object databases and software tools. In addition,
Ardent entered the data warehousing market in 1996 with the initial release
of DataStage, an extract/transform/load tool. The vendor has also acquired
a number of other software vendors in recent years, and has attempted
to integrate the purchased technologies into their existing DataStage
product. The tool now offers metadata management and data quality management,
which according to the vendor will "simplify integration of multiple data
sources and business intelligence tools".
The
Ardent product is a suite that contains:
- Extract/Transform/Load capabilities in the form of DataStage 3.6.
This product has made Ardent one of the top three ETL vendors.
- Metadata management technology acquired with the purchase of Dovetail
Software. This is a proprietary meta model, Ardent promises an XML-compliant
repository in a future release.
- Data quality technologies acquired with the purchase of Prism Solutions.
Prism had previously purchased QDB Solutions, which created QDB/Analyze,
a tool for complex data cleansing.
- Improved mainframe functionality, also acquired with the purchase
of Prism Solutions. Prism's mainframe based ETL tool provides Ardent
with mainframe job scheduling and improved access to legacy data sources.
The
most recent release is known as DataStage XE and includes all the features
of DataStage 3.6 and the addition of support for unstructured data (i.e.,
XML). This was announced as the "first component of Ardent's newly announced
Enterprise Information Infrastructure (EII) initiative by addressing the
structured data domain."
According to Mikael Wipperfeld, Vice President of data warehouse marketing
at Ardent, "Customers are looking for a complete set of information infrastructure
capabilities for their business intelligence and analytical applications
which will enable them to adapt their data-to-information processes as
their environments change."
In
the second phase of this effort, Ardent intends to deliver Web-enabled
access to the business and technical metadata. Access to unstructured
datasources such as spreadsheets, presentations, and text files will also
be added. In the third phase, Ardent plans to deliver the foundation for
an enterprise portal by utilizing an integration layer that unifies related
content and enables personalization.
The
latest development was the announcement on December 1, 1999 that Ardent
Software was to be acquired by Informix, a major relational and object
database vendor. The transaction was completed on March 1, 2000. Under
terms of the transaction, Ardent stockholders will receive 3.5 shares
of Informix common stock for each share of Ardent stock. The ramifications
of this development are still uncertain, as Ardent was highly successful
in 1999, and Informix is still rebounding from near-death (their stock
had plummeted to as low as $6 per share within the last 52 weeks, and
sales of their core database products were poor). How successful the marriage
of the two companies will be remains to be seen.
Another
complicating factor is the lawsuit which IBM has filed against Informix
for patent infringement. Many of the six patents referred to in the suit
are old, one dating back to 1981, and Informix's position is that the
suit is frivolous. However, defending against the suit will take cycles
away from Informix executives who could better spend their time on Ardent
software and personnel.
Vendor
Strategy and Trajectory
Ardent's
stated strategy before the announcement of the Informix acquisition was
in the area of an "enterprise application architecture". As explained
above, this was supposed to build a framework that would allow access
to structured (i.e., relational database tables), and unstructured (i.e.,
XML or text files), in a seamless manner to allow data access and integration.
Ardent was well positioned to succeed in this direction given its integration
of the acquired data cleansing, mainframe data access, and metadata management
tools.
After the announcement of the Informix acquisition, Ardent added a strategic
move towards "web analytics". This makes some sense given Informix's analytic
application strengths provided by its acquisition of Cloudscape, which
brought a 100% pure-Java database management system into the mix, as well
as web analytics combined in the i.Informix web-enabling product. The
key will be a successful integration of the products, clear product focus,
and the ability to prevent employee attrition.
Ardent
commented that in addition, the acquisition by Informix provides it with
access to Informix's worldwide sales and support staffs and increased
global reach. Ardent also gets to leverage Informix's installed base,
which is known to be extremely loyal.
Vendor
Strengths
Ardent
has been consistently successful in the extract/transform/load market,
especially since users moved away from scripted data movement tools to
a graphical user interface. Their market penetration has been consistently
higher from period to period, and revenues have increased steadily. The
DataStage product has been superior to many of its competitors due to
advanced data cleansing capabilities and superior mainframe data access.
The addition of a metadata management layer was key to staying ahead of
the competition.
Ardent
has also been very financially successful, and Informix has returned to
consistent profitability over the last 8 quarters.
Vendor
Challenges
Obviously,
the most difficult challenge for Ardent Software now is the successful
transition to Informix. This will be discussed further in vendor predictions
and recommendations. In addition, some other ETL vendors, particularly
Informatica, are considered to be superior in core data movement technology.
Another
challenge that would have been faced by Ardent before the Informix acquisition
(and may still be), is that some of its competition are very large software
companies with deep research & development pockets. Examples of vendors
in this category are IBM, Microsoft, Oracle, and Computer Associates (who
may not choose to spend R&D dollars on ETL products).
An
additional challenge has been Ardent's lack of name recognition and market
awareness. We believe this will be alleviated to some degree by the Informix
merger, since Informix is extremely well known throughout the data warehouse
industry. Informix must be very careful to integrate the corporate cultures
in very short order to keep attrition to a minimum. The clash between
east coast-based Ardent's culture and west coast-based Informix could
be damaging.
Other
challenges involve customers developing the perception that they must
have an Informix database in order to use Ardent data movement software,
or that DataStage will work most efficiently with Informix databases.
This must be quickly addressed with a marketing campaign.
One
challenge that will be difficult to address is getting cooperation from
other vendors in the database space who consider Informix a competitor.
Most of the more successful ETL vendors have taken great pains to be impartial
(or at least appear to be) by not also owning a database product. Companies
such as Microsoft and Oracle might be less willing to provide quality
assurance and API assistance to a direct competitor.
Vendor
Predictions
We
predict that the completion of the Informix acquisition, with the attendant
staff attrition, adjustment to a new corporate culture, and other inevitabilities
of a merger, will cause Ardent's release schedules to suffer by as much
as 6 months (80% probability). Schedules will suffer further due to attempts
to integrate functions from DataStage and Informix's products into each
other (80% probability). There will also be a loss of some key personnel
in the areas of sales, marketing, product management, and development
(50% probability).
TEC
also predicts that other, smaller ETL and infrastructure vendors such
as Informatica and Sagent will be forced to merge with larger software
vendors (such as Sybase) in order to be able to compete in this market
with behemoths such as Computer Associates, Informix/Ardent, Oracle, and
Microsoft.
Vendor
Recommendations
Ardent
and Informix should do everything in their power to ensure a smooth transition
for all the affected employees. Informix would also be well advised to
allow the current Ardent management to continue to control of their particular
product strategies and marketing. Changes in announced dates for beta
release schedules and general availability of software should be kept
to an absolute minimum. One example of how not to do this is the Computer
Associates/Platinum Technology merger. This merger caused the release
date for DecisionBase (a competitor of DataStage) to be knocked nine months
off schedule, a virtual lifetime in the software industry. DecisionBase
may never recover its momentum.
User
Recommendations
Customers evaluating data movement and data infrastructure technologies
should definitely include DataStage on a list of possibilities. However,
close attention should be paid to how successful the merger is progressing.
The first sign of trouble may be a change in a previously announced release
date. If this occurs, the vendor (at this point Informix), should be closely
questioned on the reasoning. The customer can then decide whether the
explanation is reasonable and may work in their favor (i.e., the addition
of new features the customer needs), or that the vendor is trying to cover
for the fact that the marriage is not going smoothly. Informix is moving
to a strong web/e-business-centric solution, so DataStage in its future
releases will be most appropriate to customers evaluating these types
of solutions. Other products in this space that should be evaluated are
Informatica PowerMart, Sagent DataLoad, IBM's Visual Warehouse, and Computer
Associates DecisionBase (when version 2.0 becomes available). Customers
looking for specific point-to-point ETL solutions, particularly to extract
data from ERP systems, can also consider Acta and Hummingbird.