Data quality.
It’s a big issue with lots of ways things can go wrong and only a handful to get it all right. One of the worst things you can do for your clients is give them a recommendation that is based on bad data.
Let’s define bad data. It’s not just wrong numbers. Bad data can be a number of things:
- Using the wrong data source.
- Math errors.
- Wrong numbers.
- Leaving out data.
- Misinterpreting results.
- Wrong calculations (the math is right but the actual manipulation of the numbers shouldn’t have been done).
Each of these is a unique way of getting the answer wrong. And it’s possible to do some of these without even realizing that you’ve done it. It’s very difficult to test for not using a data set or whether you are using the right kind of calculation. That can only really be done by bringing in someone else to check your work and methodology. Not easy to do if they don’t understand everything you are trying to do.
When doing your QA and final checks on any solution you are giving a client, be sure you run through all of the possible things that may have gone wrong. There’s nothing worse than giving the wrong answer.