Is that percentage for DNW survey against individual accounts or individuals as real people?
If it's against accounts then it's easy for people to register a survey response against multiple accounts and put a real nasty skew on the results.
What percentage of the DNW population actually answered the survey? Do you have the nay-sayers being the only people that answered in force?
Interesting that only percentages are given and those are only based on the results. It would be more worthwhile to see those figures measured against active accounts. and paid versus fr00b.
You may get more fr00bs going paid if the community was bigger and more active.
At the end of the day, unless you slice the figures correctly you are looking at misinformation which can potentially harm your business. Give us all the data and we'll tell you what needs doing