in Palo Alto, California, Informatica was founded in late 1993 by Gaurav Dhillon,
the current Chief Executive Officer, and Diaz Nesamoney, the current President.
Originally they were strictly an ETL (Extract/Transform/Load) vendor who competed
with Platinum Technology, Prism Solutions, Carleton, and Evolutionary Technologies
(ETI). With his background at Unisys and Sterling Software, Dhillon understood
that applications were being moved from mainframes to the then-new client/server
technologies. Many companies were struggling to move the required data between
the two environments. Informatica was formed to solve this problem. The first
product to be introduced to market was Open Bridge in early 1994. Open Bridge
was a set of tools that allowed the movement of data from mainframes into the
client/server environment using a graphical front-end.
receiving $2.5 million in venture funding, PowerMart was released in May of
1996 as the next generation of Open Bridge, with the addition of features designed
to allow customers to design the data movements without assistance from the
vendor's consultants. PowerMart has components to allow the designing, building,
populating, and managing of scalable data marts, and the server can be run on
Windows NT or UNIX. One interesting feature of PowerMart is what Informatica
refers to as "Scalable Pipeline Processing", which combines disk staging with
in-memory server-side caching to fully leverage system resources. It also has
an FTP integration feature which helps set up FTP parameters for flat file transfers
(flat files need to be on the same machine as the ETL server as parsing files
across an NFS mount is inefficient). In addition, the product contains an open
metadata repository, a pipelined, multi-threaded server engine, and a scheduling
was followed by PowerCenter in 1998. PowerCenter is an enterprise data integration
hub that integrates and unifies diverse data sources, including ERP systems.
In addition to PowerMart for data movement, the product contains a Global Repository
which, via Informatica's MX2 API, allows business intelligence products to share
metadata with PowerCenter. (See TEC News Analysis article: "Informatica
Conforms to Metadata Standard" November 10th, 1999). The product also contains
pre-built integration with Trillium's data cleansing product and is capable
of extracting from SAP and targeting BW (SAP's Business Information Warehouse),
and can also extract from PeopleSoft and populate their EPM warehouse.
and Business Components were released in 1999. PowerConnect is a family of packaged
software products that allow customers to extract both data and metadata from
ERP and legacy applications. The information is then delivered to PowerCenter.
PowerConnect modules currently exist for SAP, PeopleSoft, and DB2. Business
Components are pre-defined templates that shield application developers from
the complex process of consolidating data. They provide business-level views
of complex systems such as SAP R/3 to simplify the creation of analytic systems.
were also released in 1999. They allow the Entity Relationship Diagrams (ERD's)
for schemas to be imported into Informatica's metadata repository, which allows
model changes to be synchronized with the ETL tool. PowerPlugs currently exist
for Oracle Designer/2000, Platinum/CA Erwin, and Sybase PowerDesigner.
next product to be released will be PowerCenter.e 1.6 in the first quarter of
2000. Informatica states that the product "extends business focus to provide
the first cross-channel data integration platform for eBusiness."
January 5, 2000 Informatica released PowerPlug 3.0. Based on their MX2 metadata
initiative, and jointly developed with the Yaletown Technology Group, the product
is capable of bi-directional metadata exchange, metadata updating and synchronization
from upstream data modeling tools, and metadata comparison, among other features.
development of these products has allowed Informatica to provide a complete
solution for the population of data marts and also data warehouses. (There is
a common misconception about data marts being smaller than data warehouses,
but this is untrue, it is the fact that they are subject-focused that makes
them data marts.) Before purchasing the Informatica product, the United States
Postal Service proved that PowerMart could scale to three terabytes.
Informatica was originally a privately held concern, but completed a public
offering in late April 1999, issuing 2.75 million common shares at $16 and raising
$44 million .
Strategy and Trajectory
strategy has always been to assume that data warehouse projects are too large,
expensive, and prone to failure to be an effective answer to management's need
for business knowledge. As expressed by Gaurav Dhillon, "We looked at the existing
model and realized that the idea of building a centralized warehouse that everyone
can use is not a very good one. It takes too long and the tools are complicated.
The result is that it never happens." Dhillon and Nesamoney decided that the
data mart would be faster and cheaper. (Data marts can either be subsets of
a data warehouse or standalone, and typically represent information for one
group or function). PowerMart was originally developed to allow the extraction
and transformation of data into data marts. PowerCenter is an extension of the
same principle, and now provides for a global metadata repository, multiple
engines, and centralized monitoring of processes. With the acquisition of Influence
Software for business analytics and the release of PowerCenter.e, they are moving
strongly into the eBusiness market, including enterprise resource planning and
extended supply chain access. With the addition of support for MQSeries and
XML in the PowerCenter.e version, much of the work can be done in real time.
We believe that Informatica is moving in the correct direction. Informatica
has also become profitable in recent quarters in spite of its high investment
in research & development (25% of total revenues for fiscal 1998).
has a lead over most vendors in core database extraction technology. In addition,
their MX2 metadata initiative gives them a wide-ranging competency with metadata
exchange among other vendors. They currently have interfaces to products from
tool vendors such as Business Objects, Cognos, and Seagate Software (See TEC
News Analysis article: "Informatica
Conforms to Metadata Standard" November 10th, 1999 .) In addition, the acquisition
of Platinum Technology by Computer Associates has thrown the DecisionBase product
off of its release schedule (by a minimum of six months), which should allow
Informatica to gain further ground against this competitor. (For further information
on the DecisionBase product see "Computer
Associates Splashes Into the Data Warehousing Market with Platinum Technology
Acquisition" September 23rd, 1999 .) With both of its major competitors
in a state of flux, Informatica has an opportunity to gain ground. Moving into
the hot field of eBusiness should provide many sales opportunities.
only been public since April of 1999, Informatica needs to keep increasing their
market share and visibility. Their current competitors are well-funded (Platinum
Technology is now part of Computer Associates, the third largest software vendor,
Ardent Software is being acquired by Informix, another large, well-established
software firm.) Informatica must act fast in both the research & development
and sales & marketing areas to capitalize on the current state of flux in their
market. In addition, competing vendors are adding other capabilities to their
products, such as improved mainframe data access, business intelligence capabilities,
and data cleansing. Informatica will have to work to keep up with these trends.
(For more information on these developments, see "More
Data is Going to the Cleaners" December 1st, 1999). Informatica's push into
business analytics through their acquisition of Influence Software should give
them a chance to compete with other product "suites". Informatica has ambitious
plans for eBusiness, and will have to work hard to realize them in a completely
predict that Informatica will improve their market share to at least 20 percent
in the next year (60% probability). Informatica's stability compared to the
other two major ETL vendors should give them a strong opportunity to excel in
increased market share, customer awareness, and product functionality. We believe
that there is a 30% probability that Informatica will merge with another, larger
software vendor to extend their reach. We also predict that Informatica will
attempt to increase their services revenues from consulting (60% probability).
should keep an eye on Ardent Software, since they are merging with Informix.
Informix's world-wide sales and support facilities will provide Ardent with
a greatly expanded sales and services channel. (See TEC News Analysis article:
to Acquire Ardent Software-Another Vendor's Attempt at End-to-End Data Warehousing"
December 14th, 1999) Ardent has kept their eye on Informatica and considers
it its strongest competition. Informix and Ardent are also working on eBusiness
and will continue to be its strongest competition. Informatica may have a lead
in core database extraction capabilities, but needs to think about the next
level of competition, especially in the area of unstructured data (data which
is not stored in relational tables). Informatica's MX2 metadata initiative,
as well as its contract with the U.S. Postal Service, gives it a powerful chance
to keep up with Ardent.
Informatica should be included on any short list of Extract/Transform/Load tools
and eBusiness solutions with Ardent Software's DataStage (Ardent is currently
being acquired by Informix), Sagent Technology's Data Mart Solution, BMC's ChangeDataMove,
ETI's Extract, the SAS Institute's Warehouse Administrator, as well as others
depending on whether ETL is the only goal. It should be noted that there are
few pure-play ETL vendors left, most have entered into the Business Intelligence
space and eBusiness markets, and/or are attempting to provide End-To-End Data
Warehousing solutions. This fact will make it more difficult to compare vendors
against each other due to differing product mixes.
(File Transfer Protocol) A protocol used to transfer files over a TCP/IP
network (Internet, UNIX, mainframe, etc.). It includes functions to log
onto the network, list directories and copy files. It can also convert
between the ASCII and EBCDIC character sets and supports both character
and binary transfer modes.
(Network File System) The file sharing protocol in a UNIX network. This
de facto UNIX standard, which is widely known as a "distributed file system,"
was developed by Sun. It can be used to make files appear to be on the
local machine when they are in fact somewhere else on the network.
Data that describes other data. Table and column names are examples
of metadata. Metadata for a relational database system is stored in a
The definition of an entire database. It defines the structure and
the type of contents that each data element within the structure can contain.
Schemas are often developed with modeling tools such as Platinum/CA Erwin
and then the SQL Data Definition Language (DDL) is generated from the
tool and applied to the database to instantiate the schema (create the
physical database from the logical model).
(Entity Relationship Diagram) A database model that describes the attributes
of entities and the relationships among them. The typical entities in
an ERD are tables and columns.
Messaging middleware from IBM that allows programs to communicate with
each other across all IBM platforms, Windows, VMS and a variety of UNIX
platforms. Introduced in 1994, it provides a common interface (API) that
programs are written to.
Extensible Markup Language) A document format for the Web that is more
flexible than HTML. Its use is becoming common is eBusiness.