DS0409_REPORT_p13-p14_F3:Layout 1 1/4/09 16:01 Page 14
REVIEW
“
Data governance is hard unless
it is tied to a business outcome.
using outsourcing as a way of avoiding collaborative working.
As data governance joins that list of demands in invitations to
tender, outsourcers and software vendors have been scrambling to
adapt. For some, it has been more a case of waiting for the
marketplace to catch up.
At Kalido, vice-president of strategy and business development
Winston Chen, says that having an audit trail is one of the main pillars
of data governance. “You need to know who did what, when and
why? It is difficult to find a tool in the vendor market to support that,
but all our software does it from the ground up,” he says.
From its original development, the company included automated
reporting on each action that is taken, such as changing the
classification of a customer or the description of a table in a database.
“It is built in, so you can go back and find the history. We think the
human component in data governance is critical,” says Chen.
Most often, breaches of data protocols are the result of human
Since marketing departments often create SCVs outside of the error or ignorance, rather than malice or software error. By having
corporate data infrastructure precisely to avoid the strictures and visibility of everything that gets done to the SCV, the organisation can
controls of IT, they may not be so keen to implement data be more confident that it will remain compliant – and that its staff will
governance. As the value and sensitivities around data management be encouraged to follow procedures and stay within their boundaries.
grow, however, this may no longer be an option. These business rules are another of the pillars of data governance,
“Marketing is less averse to controls now. They are aware of the according to Kalido. Its software is model-driven, which means that an
need to protect data. But a lot of people don’t know what to do for existing set of business rules can be adopted into the solution, or a
the best,” says Gregory. He argues that a good starting point is to new model worked out between IT and business functions. These rules
adopt a practice routinely used by commercial data owners – seeding include how customers are defined, placed in hierarchies and added to
data with special names and addresses that can be used to identify systems. Rules also need to cover what to do with exceptions.
how data is being used and by whom. “It is the simplest way to From these rules, governance processes have to be defined that
monitor use of your data,” he says. spell out what actually gets done with individual records. “The key is
This mechanism will pick up when data usage goes wrong once it to set these up in such a way that the business owns the data and
has been put to work outside of the enterprise. But some of the most does the day-to-day management,” says Chen.
important safeguards are those that protect customer data within the This can be the hardest part of a data governance project. Finding
SCV against improper or unauthorised usage internally. data stewards means putting an individual at the heart of a set of
Many of these potential misuses are not malicious – they are just a processes and with a battery of metrics that require constant upkeep.
case of somebody viewing data, making changes or moving files Chen says that companies which already have a business process
beyond their authorised domain. In the absence of established data orientation, such as managers of order to cash, for example, will find
governance procedures, it is difficult for staff to know exactly what these easier than those without these cross-functional roles.
the boundaries of their legitimate domain might be. Tom Scampion, partner in enterprise services and risk at Deloitte
This is where technology can play an important role. “The Data, makes a similar point. “Data governance is seen as hard unless
approach for us is a software answer delivered in-premise, deployed it is tied to a business outcome. It shouldn’t be seen as something
to the standards of IT, but supported by SmartFocus to the demands abstracted or separate to what the business is set up to do,” he says.
of marketing,” says Chris Underhill, CEO of SmartFocus. To maintain this business orientation, he argues against establishing
The solutions his company is now deploying against SCV are more data governance as a centralised function which can be seen as
sophisticated and complex than would have been the case even five separated from actual business activity. “It can end up being seen as
years ago. “Historically, the need for SCV was to serve a business about controls, processes and rules without benefits to a particular
requirement, like marketing’s need for retention and competition for profit and loss line,” says Scampion.
customers, rather than for governance and complexity,” he says. “If you do align data governance to better outcomes, you get more
In the current marketplace, both of these requirements have been buy-in and acceptance. It does well when it is an enabling function, so
converging, with IT keen to avoid customer data passing outside the it has to come from the business,” he says.
corporate firewall, but also still wanting to service marketing’s need People are a major element in this positioning and acceptance.
for rapid and flexible data access. That is why the data mart may sit Scampion says it is important not to “fly in” people who have no
on a server inside the corporate architecture, accessed by mining, understanding of how the business operates or who see governance
modelling and reporting tools through SOA and Web browsers, then as an end in iself. “It is about managing costs and the challenge is to
fed back to marketers as the view they need. any business which is not seeing data as an asset,” says Scampion.
“As a supplier, it is about us applying the data quality rules and This is the “lightbulb moment” – recognising that the single
knowledge behind the data that marketers use to support their activity,” customer view has an asset value, even if it is not on the balance
says Underhill. To win contracts for SCVS these days, both marketers sheet. That is when data governance becomes a must-do and the
and IT have to believe their requirements are being fulfilled, rather than search starts for how to protect and maintain the customer asset. a71
14 DATA STRATEGY | REPORT | APRIL 2009
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21