Data integrity and accuracy of 100% is impossible in most situations. Yet decisions have to be made daily. Here's what helps me handle this challenge:
1) Collect data from all different sources (people's personal experience, open-ended, surveys, analytics etc.).
2) Triangulate where the truth could be based on the varying data points.
3) Establish a wide range that gives you a high degree of statistical confidence.
4) Be conservative in your assumptions and pick the low end of the range.
5) Make a call and be prepared to be proven wrong.
You will constantly need to make calls based on imperfect information in all kinds of areas:
- Growth projections
Sometimes, you'll have the luxury of waiting as more information becomes available as more time passes by. Use that time in those situations.
But in many cases, the calls must be made. You'll get some wrong and you'll get some right. Just a reality of life :).