I had a very good conversation today.
We talked about how data is not only completely silo’d in most organizations, but it’s also in very different formats. So, when it comes to unifying data for the purposes of extracting better insights from it – it’s a slog.
Therein lies the problem. People don’t want to hear that it’s a slog.
Take, for instance, the variety of ways that a phone number can be captured.
In North America, I can think of five formats off the top of my head:
(212) 555 5555
212 555 5555
1 212 555 5555
Is it all in the same field? Sometimes the country code is stored in a different field from the area code. The router code is commonly stored alongside the local portion, but not always.
Not that router numbers are all that useful anymore these days – but the point remains the same – since there are a number of ways to store a piece of information in a database and across systems, if there is no form of data governance, then disparate databases become difficult to merge.
And of web analytics?
What if one side of the organization is using Coremetrics, and the other side is using Omniture?
Each vendor still has it’s own set of definitions on how things are captured. Even if the entire site was double tagged, you’d get two different sets of results. They’d be within a margin of error, but it would be hard to tel which one was more accurate.
So what do you do?
Work to make it less of a slog.
…and set expectations. 🙂