|
Data
is the current buzz. Driven by the need to provide front
office personnel with quality data and by regulatory,
budget and/or integration pressures, every firm needs
to address the challenge of data management. Most firms
are struggling with similar questions: How do you get
access to clean data?Should you buy, build or outsource? How
do you manage the data? Who owns the data? Who
should oversee the efforts? Should data management
be centralized or decentralized? What kind of organization
should be in place?What system best supports data management?
This
article is a case study contributed by a former Director
of Investment Systems at a large investment management
firm. What follows are reflections of the experience
with data management and data warehouse initiatives.
The
experience
During a
strategy session our management concluded that to gain
a competitive advantage our firm needed to invest in data
management. In the late 1990s we embarked on an ambitious
effort to build a proprietary data repository. With little
experience in either data warehousing or the base technology,
we delivered only a small slice of what was originally
envisioned.
Leadership
shifted and our philosophy switched to “buy over build”,
so we went on to purchase the investment industry’s premier
solution. We bought into the story that the vendor’s
centralized data repository would provide the business
answers and the technology solutions with which our team
had been struggling. We believed that this best-of-breed
approach would integrate vast amounts of data into a
wide array of applications and accelerate project deliverables.
What we bought was overkill for the firm’s requirements.
Since that time vendor solutions have matured, and there
is a broader array of offerings. However, during the
course of this effort we learned many lessons that still
apply today:
Don’t
underestimate the effort
We learned that the implementation of a data warehouse
is a multi-year and very expensive effort. We misunderstood:
-
The
enormity of the initiative
-
The
extensive knowledge and skill required to execute
the project plan
-
The
business disinterest in the level of detail required
to get results
Don’t
oversell the initiative
Business users and senior management expected data management nirvana. After
many person-years of effort and vast resources expended, the project was perceived
as over-promised and under-delivered. Without a tangible and immediate benefit,
business sponsorship weakened.
Create
partnerships
Identifying the right business sponsor is a key to success. Currently compliance
and operations have significant influence, given the environment of regulatory
pressures, and are very good candidates for data initiative sponsorship. They
have a capacity for detail, a strong appreciation for the end result and funding
for such initiatives.
Prioritize
the data
Spend the up-front time necessary to document requirements and priorities.
What we learned is that success with data management projects happens when
a firm starts with a single area that will provide visible returns for business
users. A winning strategy is to tackle only what the firm has the capacity
to handle, keeping it simple and taking incremental steps. This approach to
data planning is an opportunity to build consensus with all the parties involved IT,
Ops and the business.
Conclusion
Technology does not solve data problems, and the deployment of a data warehouse/data
hub platform is a complex and multi-year effort that often does not solve
the immediate needs of the business. The most successful data projects
occur when they are the backbone of more tangible business initiatives,
such as the consolidation of order management systems, enhanced client
reporting or streamlining operations. Assembling a team of data experts
across business and IT functions helps to build business buy-in and the
executive sponsorship necessary to keep projects moving forward. Success
occurs when business benefits are visible and incremental.
|
