When Bad Data Happens to Good Organizations

When Bad Data Happens to Good Organizations

By Chris Lundberg

While working with some of our first clients and focusing on some of the analytics and reporting components (now known as the Business Intelligence Squad), a recurring theme started to appear:

Bad Data.

It’s a common problem, so if it happens to your organization, don’t be alarmed. Bad data happens to good organizations for many reasons that I won’t go into here, but the implications can be huge.

Bad data makes it impossible to get good reports. We have seen all kinds of bad data, but the worst kind is data that has not been properly coded at the outset. You can always fix a bad email address or re-request corrupt files, but if the data was never there in the first place, there’s nothing you can do to fix it. For example, if a payment or donation was made a few years back and no tracking information was stored with it, it’s extremely difficult to figure out which messaging source to credit it to. And we see this everywhere! Even clients who were religious about using tracking codes and devised fantastic coding schemes saw decays over time in the quality of those codes. People change jobs, knowledge is lost, people forget.

There’s no getting around this: People are generally sloppy about keeping track of thousands of codes.

Not to mention that every system uses tracking codes slightly differently. Some call them Source Codes, Tags, Tracking Codes, Flags, Custom Fields or Market Codes. Many different terms for what fundamentally is just a related set of actions. So, if you wanted to be good at tracking information, you’d have to have a deep knowledge and understanding of how all these systems code and group their data. That’s hard and getting harder as we use more and more technologies.

We thought there must be a better way. Why are we coming up with custom coding schemes for all these different systems? This is a perfect job to automate. There’s even a whole swath of companies who deal with tracking these codes and tags, but they’re not tied into the workflow of most organizations.

Bad data was one of two things that made us realize that marketers needed help getting data INTO their production systems; not just getting data OUT of them. The process of producing content and correctly setting up tracking codes at the onset is crucial to getting good reports at the end.

If we take all that human sloppiness out of data going into communication systems, the reports we can get out can actually be accurate. So we think you should never even have to see a tracking code ever again because, yep, there’s a Bot for that!