Are we teasing out the right thing?
I recently launched a program to help customers see the value of using the software - I work for a software company now. The program was for showing that more is possible if you choose to act. We knew why were we doing it, but we had a hard time figuring out, "did this work?"
What was so difficult?
It wasn't collecting data. We had lots of data.
What was the challenge was pursuing the right data point.
What was the data point that helped us tell our story? Was it the trend of how many people logged in? Did that increase? How do we make an apples-to-apples comparison against the people who did not use that program? Oh wait, that data doesn't work.
Figuring out what to build is half the fun. The other half is figuring out how you will tell the story that what built made the change you sought to make.
Don't forget to do the second half before you start building!