By John Randles CEO PolarLake

There is no end to the amount of literature in the world of IT, Business Consulting and the Harvard Business Reviews of this world on the age old debate in how to align business and IT. But how relevant are these works to the Reference Data world? When it comes to Reference Data we see it as a very peculiar case. This is because business requirements are so distributed across the organization and so many stakeholders that are peers of each other. And the famous executive sponsor tends not to have the patience for refereeing disputes in minutia. A typical example would have various departments being feed from a centralized source, such as Equities department, Fixed Income department etc. who all have different SLAs for a centralized Reference Data service, different data types, formatting requirements, cut off points etc. Then there is the age old debate of who should be responsible for filtering out data that I don’t need, formatting, creating the correct cardinalities etc. Should this be done locally or centrally? Add to this the fact that the downstream systems owners do not want to (or cannot afford to) change today’s processes you can see real Business IT alignment problem far beyond the classic business requirements round tripping case studies from Management Consultants on a single business application implementation.

Beyond the alignment of requirements there is an obligation on the Centralized Reference Data Service to provide an excellent service otherwise no matter what the dictates from on high the downstream systems will source their data directly from the data vendor and do their own cleansing and enrichment on the data themselves resulting in more duplication of effort. The best of breed operators of the centralized data services understand that requests for new feeds, asset classes, format changes, delivery methods etc. are an opportunity to over impress, over deliver and end up with a very happy customer. The way to genuine centralization is to have the downstream data consumers want to use the central service as well as the usual mandates. Otherwise people downstream are very creative at building exception cases for not using the centralized service.

However imagine a scenario where a centralized Reference Data Service has a single standard format, with all data going to all of the downstream systems daily (remember all the Equities systems flooded with Fixed Income data at the height of the crisis!). Then one day each downstream system could subscribe to a centralized service for what data they wanted, determine the format of the delivered data, using their own data cardinalities, specify usage and security policies and with a specific delivery schedule and method. All specified in their tool of choice – excel. 95% of the data mappings that need to get executed are automatically generated. The central group work with the consumer to configure the remaining 5%. Each line of business is responsible for controlling their data usage, data mappings, delivery schedules etc. The centralized group then executed these policies and is responsible for control. Did I deliver what I committed to deliver in the time allowed and if I didn’t what went wrong. Help me fix it!

What this allows is the highly parallel, fast alignment of Business and IT specific to the case of Reference Data. Requirements are owned and controlled by the business resulting in minimum IT round tripping. When looking downstream there are an ever increasing number of consumers with varying requirements. There is renewed focus on the Business and IT alignment as the focus on risk and more vigorous regulation take hold as everyone’s main Data concerns. Being able to handle this in a highly parallel way gives downstream consumers what they want certainty, confidence and control. The centralized service gets full traceability & control while meeting the consumers SLAs and beating expectations, because they are directly executing on the consumers requirements.

We at PolarLake see so many of the problems around Reference Data being Business IT alignment. It is no longer valid to say it’s not a technology problem it’s a governance problem which was the fashion for so long. Neither is a valid to say if you just implement this new mousetrap and all will be well. We believe all stakeholders Business, IT, Operation, Executive, Governance etc. all need to play a role and need to be facilitated to play this role as well as possible.