We're not told that the top 1000 would be the majority. IF that is true then the sample might be a good indicator for the company. However, if it is not true then it's probably not representative of all the customers and could lead to strange projections.
One question that I have is: Will these sample statistics accurately estimate the purchasing habits of all our customers? Can you share some of your thoughts on that question? Why or why not?