Focus groups

Tend to be small gatherings of people who answer questions and give opinions on topics that the facilitator puts there.  Focus groups can give good answers to questions, but the setting is itself artificial, and does not necessarily reflect the opinions that the members of the group will have in the real world.

Direct mail trial runs 

Perhaps the most common approach in terms of research, direct mail trial runs involve producing the material you are considering sending out, and then only sending it to a small sample, just to see what the response rate will be.  It is common to send out two or more trials at once to different groups drawn from the same master list.  In one experiment that has been written up many times since, we wrote two different versions of a leaflet - the difference in fact being the addition of two extra sentences in one of the leaflets.  Both were sent out on the same day to schools randomly selected from the master list of 20,000 potential buyers.  One version got 2% response rate, the other zero - and 2% was more than enough for a profitable venture.   On other occasions we have taken the same leaflet and sent it to different people in the selected establishments, just to find out exactly who might be the most likely to place an order.

On-line surveys

Are lower cost to run than questionnaires, but they tend to be a little less reliable, in that the distance between the respondent and the person asking the question means that some will be tempted to put down any answer, just to get through the questions quickly.