Why Analytics Projects Fail – #9: Bad Data

In my experience, most of the time analytics projects fail its generally traceable back to a purely human problem. However, sometimes you see things fall apart because of technology, the misuse of technology and/or just bad technology. This is the case when projects fail because of bad data.

There are a lot of ways bad data can happen.

One common way you end up with bad data, is the data was not captured correctly. Perhaps the data was manually input with lots of error. Or maybe your data is not consistently collected so it has gaps. Knowing what exactly goes into capturing your data and being able to understand how it is collected is extremely important.


Another cause of bad data is that you are not getting all the data or you are getting data that has been altered. A lot of times when data passes from the collection point to you, it might be being truncated, or blended, filtered or converted. Lots of databases are structured for optimal data storage, not usage. A lot of database admins who don’t really know the data will add data flow shortcuts. Or maybe the fall under the datakeepers category and partition or cut out some of the data you need.

Bad data also comes in the form of old and out of date data. When you are making decision on data that just not recent enough, it can lead to a lot of problems. Keeping data fresh is something some companies just don’t value. If that’s the case, you will likely see your analytics initiatives come up with analysis that points you in the wrong direction.

In all three of these examples, one solution I suggest to mitigate the chance you have bad data is to build a data map. Learn about every point in a data flow that touches your data. Talk to the ones in charge of each touch point to make sure your data is not being impacted in any way that can result in bad data. Even if you cannot fix the problem, understanding it can help you set more realistic expectations of what your analytics project can achieve.

I have found using Visio to build data flow visuals is the best way to explore, document, and report how the data being used in my projects is being impacted by the environment it lives in. Knowing Visio is a valuable skill for an analyst. If you don’t use it, I promise you that once you do you’ll be sending me a thank you.


Analytics Culture – The key to using analytics in a business is like a secret sauce that fuels Data-Driven Decison-Making. It is a unique combination of analytics talent, technology and technique that are brought together to enrich and empower an organization.

A successful analytics culture is not easy to create, but DMAIPH can show you how. Contact DMAIPH now at analytics@dmaiph.com or connect with me directly so we can build a strategic plan to turn your company into analytics driven success story.