Should Regulatory Bodies Be More Prescriptive Regarding Match Rules?

If I am a bank and I lend £100k to person A, then there is a risk that I might not get that money back.

If I am a bank and I lend £100k to person B, then there is a risk that I might not get that money back.

If person A and person B turn out to be the same person then I have lent £200k to that one person and there is an even bigger risk that I might not get that money back. This is not rocket science. It is basic risk management.

The ability for an organisation to identify where multiple relationships actually relate to the same entity (person or business) is a key requirement, necessary to achieve compliance. FSA faster payout, Solvency, Basel all require organisations to have a better view of their exposure to risk and achieving a Single Customer View is an important component which contributes to a more accurate view of that risk.

But achieving a Single Customer View (SCV) is difficult. Actually, achieving a SCV is easy. Achieving an accurate, reliable and trusted SCV is difficult.

This is because the relationship between available data, data quality, business rules and matching algorithms is complex. I don’t intend to expand on this now, suffice to say that it all comes down to achieving an acceptable balance between false positives and false negatives. There is further information of this on my website.

Now there is a great deal of information on compliance published by the FSA, CEIOPS/EIOPA et al but I have yet to see any clear and unambiguous guidance on how to actually achieve the required SCV. What is an acceptable match? Is it OK not to match “John Smith” and “Jhon Smith” because the data is demonstrably different? What about “Fred Bloggs” and “F Bloggs”? Should these be allowed to match or not? What about “Quentin Smyth, 12 Acacia Ave..” and “Quentin Smyth, 21 Acacia Ave..”….I could go on but I think you get the idea.

These questions are what I would expect any organisation striving to achieve a SCV in support of a CRM initiative to be wrestling with. And in that context the answers should be driven entirely by the organisations goals and appetite for risk. But if an external body is specifying that “you must comply with my rules”, should not those rules be clear and unambiguous?

Or is it sufficient to fall back on “best endeavours”?



This entry was posted in Data Quality, Single Customer View. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *