Bill Hewitt is busy these days.
As the CEO of data management startup Kalido, he’s on a circuit tour that
took him to Florida for the Gartner Symposium ITxpo show earlier this month.
Two weeks later, he keynoted a breakfast discussion about enterprise
information management for the Kalido User Group Conference.
Kalido has come out of stealth after three years with more of a bang than a
whimper, offering profitability management,
performance management, enterprise data management and master data
management software to customers struggling to gain more insight into the
data that keeps the business spinning on its axis.
Hewitt, who cut his teeth at Novell, PeopleSoft and IBM, recently took time
to chat with internetnews.com about the information management
market, its players and its future.
Q: Kalido isn’t your typical startup. Where did it come from?
Kalido was formed as a project within Shell Oil 10 years ago. The project
was started because Shell was having difficulty determining at the end of
the year, which products they thought they sold and which products they
actually sold.
Fundamentally, they had a naming problem. They could call the
same product in five different regions five different things. So they built the
project as a master reference data project to rationalize their global
product catalog.
Over the years, they found that they could use it for any
number of different things.
Five years after they built it, the team that was running the project and
had deployed it across Shell convinced Shell to create the team as a
subsidiary so they could sell the technology commercially.
In 2003, the
company was taken private by Atlas Ventures and Benchmark Partners. They
actually bought the company from Shell and put more money into it. In 2004,
Matrix Partners joined, as well. Since 2003, the company’s grown to 250
implementations for 47 customers.
Q: What problems does Kalido’s software solve?
At Gartner [IT symposium in Orlando] we announced a new set of packaged
solutions around the concept of active information management. With active
information management, we now bridge the gap between IT and the business
organization when it comes to delivering information.
We work with the
business side of the house to determine how they want to look at their
business, and then we create that in a business model that drives into the
data and delivers a view based on that business model.
If a business wants
to change their view, combine three units into two, or expand product lines from four to six, all they have to do is change the business model, not the underlying data structure, and an automatic new view is generated.
Q: How do you deploy this in a customer environment?
For example, Owens Corning struggled to find out why they were losing margin
in one of their product lines. They asked us if we solve this problem.
In a five-week period, we developed the business model I mentioned based on the user requirements. Over a second five-week period, we worked hand-in-hand with the IT and business folks to refine the business model until the user said, “That’s the view we want.”
At this point, it automatically generated the view. What they learned almost
immediately is that their two fastest growing product lines were unprofitable
and would never reach profitability.
They shut down those two lines and saved
tens of millions of dollars instantly. They now also use that information
every day to change the incentive compensation for their channels based on
the margin performance of the product.
So they use the information to drive
behavior versus at the end of the quarter going “Oh, how’d that happen?”
Q: IBM has just created an Information Server. How does what you’re offering compare to that? Or are we talking about apples and oranges?
Information Server does a great job of integrating lots of different
products that IBM has bought into one integrated system to pull data out of
operational systems, and then basically present for use in a business.
We would sit right on top of the IBM Information Server. What a company would
do traditionally is they’d take that integrated view and that delivers that
information to a staging area.
That staging area would then get loaded into
a data warehouse, and from that data warehouse, data marts would be created
to serve different parts of the business.
And then business intelligence
would be applied to allow different users to view the data in the way they
want to see, whether it’s in reports, graphs or charts.
We take everything from the data-warehouse component all the way up to the
BI layer and automate that and use this business model-based approach.
So we eliminate all of the custom programming the typically exists in that
environment and we allow the business to play a role in what information
they see when.
That’s all driven by an underlying subject-independent
data store, which allows us to map any piece of data to any semantic
definition that the business has.
We also time-stamp every single piece of
data so the company gets a bulletproof audit trail. We allow businesses to
basically recreate any view of the business at any time.
Q: Sounds like your product would be complementary to Information
Server.
Highly complementary. They’ve done a good job of getting the information out
of the operational systems. Now we take it and get it more efficiently into
the hands of the business users.
Q: How is your relationship with IBM?
Our partnership was with Ascential, which was acquired by IBM. We have a
product called MDM — master data management.
IBM has a product called Master Data Management from a company they acquired called Trigo. What Trigo does for master data management and what we do for master data management are two different things.
However, what the product actually does and what
the PowerPoint says are often in conflict. So there’s a group of people
within the IBM information management division who believe we are a direct
competitor because of our MDM functionality, and we are working with the
Ascential team and the MDM team at IBM to help them understand the
differences and the synergies.
And we believe there’s a little overlap but
not much and we’re having conversations about how we can work together to
make both offerings better. We’re not doing anything with IBM right now, but
I think that will change over the next several quarters.
Q: Aside from the complicated, cooperative competition relationship with
IBM, what other competitors do you see out there?
On the data-management side, we still see a lot of custom data warehouses
being built. We see more hardware-based data management, like TeraData and
the upcoming players in the data warehouse appliance business.
The
fundamental difference in the traditional data warehouse, even if it’s
hardware-based, is still just a [static] snapshot of the data, where ours is
active because it can change as the business changes. You literally can look
at that information instantaneously.
Q: Where do you see this data-management space going in five to 10 years?
I think there’s going to be significant continued competition for the
middleware stack, which has moved from just application integration into
data integration and information, as well.
The market will drive to an
independence for information management that includes not only the data
management components but master data management, data quality, operational
data management, all of the different pieces of the puzzle.
What you’ll
basically end up with is information-as-a-service much like people are
thinking about delivering IT-as-a-service. You’ll have producers of data,
consumers of data and managers of data.
Each one of those constituents will
have a different set of tools or interfaces to access or manage that data
more effectively.
Q: Sounds like your classic service-oriented architecture.
It is indeed.
