Biggest Failures of Data Project
Recently in my training, I have been asked what IS the biggest reason why data projects fail. I can think of two and from my consulting experience, I felt it is a split between the two.
Data Collection
We all know that quality of data is important. But data collection for most companies are an afterthought rather than something at the top of their mind. As such data is collected haphazardly resulting in poor quality data that even if it looks decent, may create downstream impact such as wrong decisions made because data did not present the actual reality.
Companies need to plan how to collect good quality data, else the time spent to collect data will be down the drain and companies will need to allow time to pass by, to collect enough data for decision making.
If you have benefited much from the newsletter and will like to join my cause, consider making a “book" donation. Link at the bottom of the newsletter or here. :)
Cost of Project is Greater Than The Value it Brings
To be fair, gauging the value a data project can be difficult whereas the costs of a data project can be more precise. There are many ways a data project can go wrong, hurting the value it can bring.
Most companies being new to data tend to overestimate the value of data projects, and this is compounded by the lack of knowledge on how to protect the value potential of these projects, exacerbating the challenge further. As such, the value is severely stunted and the project failed because the costs is bigger than the value it brought.
What is the solution? Again, I recommend getting someone more experienced in data projects to help with the planning, execution and management of the project. They should have enough implemented projects under the belt to realise where a data project can go wrong and provide the necessary direction and execution.
In summary, the two biggest reasons for failure are
Lack of Data Collection Planning
Allow Cost to Exceed the Project Value
What are your thoughts? Do share them in the “Comments”, please! :)