How far is your goal? Would you know if you looked at this picture?
Spoiler alert: you wouldn’t be able to tell, because this picture (while showing the data) is a poor data point to answer your question. In the same way as your business data might be. And you might be taking business critical decisions based on it.
To avoid this, you can learn how to ✨validate your data✨ on the example of the data point on the picture:
It all starts with the preparation. Decide what you want to measure and how. Check a post on it here: https://www.linkedin.com/posts/ekaterinagm_how-to-measure-impact-activity-7055188247240392704-rESQ?utm_source=share&utm_medium=member_desktop
On the ski track: you want to measure your distance to the finish line to control your pace.
After the implementation is done and you get first data points, make a sanity check to ensure that the data makes sense given the circumstances.
On the ski track: if you run 5 km track, the picture should show distance less than 5 km.
Make a data unit check to determine whether the data describes the user or an action, whether it is in absolute units or not, and whether the time aspect is important for your problem.
On the ski track: this is where the data point fails the check - there is no starting point for the measurement. For this ski track one can enter from several points along the way and 3 km is a useless datapoint. Moreover it is unclear whether it is 3 km from the start or 3 km left to finish.
Make a trend analysis to understand why the data looks the way it does.
On a ski track: you can look at your pace and progress based on the distance poles along your way. You can make sense of the specific data point based on previous and upcoming signs. You can also get an overall sense of a pace per kilometer.
With this example, you can also see that even if the data is imperfect, it can still help in your decision making.
Now that you have answered all those questions and validated the data, it's time to celebrate! 🎉
Comments