Here's an interesting anecdote from one of the members of my real estate newsletter service in Ontario, Canada.
For the first seven months he was a member, the newsletter was delivered in the local community newspaper as an insert. However, he received only two responses from it.
Kind of disappointing.
After doing some door-knocking at Christmas and seeing the number of houses that had the paper still sitting on the front porch days after it was delivered (plus some delivery issues with the paper itself) he decided to switch to direct mail through Canada Post.
Since January – 4 months – he has received 18 responses.
Looks like more people are reading his newsletter now.
Of course, it doesn't necessarily mean that direct mail is better than newspaper delivery (although, personally, I think it is because you are connecting directly with your reader), but it shows the importance of testing. If one distribution method is producing disappointing results, try another.