WikiData

We've been speaking with a lot of firms in the industry over the past year about the trade off between data quality, coverage, timeliness and cost. A challenge faced by data managers, particularly in the business entity space, is that a central data repository can feed many different groups with distinctly different requirements. Credit risk, for example, will have an extremely low tolerance for latency or errors but may not need massive volumes of entities - just those to which their firm has exposure to. Marketing on the other hand may place a higher priority on a huge database of potential corporate customers to which they can target for campaigns. In this case, quality, while important, does not have the same value as it does for risk.

So the question is how does one balance these requirements while leveraging a central data repository? Six sigma level data quality across an entire CRM database would cost far too much. Incomplete data population for marketing initiatives significantly compromises the effectiveness of a campaign.

What about all of us sharing, for free, very basic data about entities? Say, legal name, country & region(where necessary) of incorporation and perhaps a few other bits as agreed by the community. We can agree a mechanism that ensures contributor identities are not disclosed unless they choose otherwise. All contributors have the capability to check, update and comment on data, very much like a Wikipedia model.

This data then serves as a platform for a financial institution's internal team and/or a third party to perform additional verification/certification. Over time, the shared and free data asset becomes more comprehensive and reliable. Firms like mine will continue to provide verification services however we will need to continue to enhance our service/content offerings in order to increase our value proposition to our market over time. And most importantly, a free and open foundation of basic data will enable anyone on the internet to identify errors. Is there a chance that some contributors may corrupt the data by accident or intentionally? Absolutely. However similar models like Wikipedia are now proving that such abuse is rapidly identified and corrected by the well intentioned community members.

Financial institutions have consistently asked for a free data utility to help address business entity data quality and identification issues. Which of you is ready to proactively participate in such an initiative? It won't happen without your involvement.

I look forward to your comments, criticisms and ideas for improvement.

Ken