Simon, interesting post but I think it needs some nuance :-)

Happy to take this discussion further but I wanted to give some fast feedback already:

1. No questionnaires beyond 15 minutes!
I agree that survey length is an important one. We currently have a max of 20 minutes at InSites Consulting. Based upon analysis on the satisfaction survey after our surveys we saw that relevance of topic is often a bigger driver of satisfaction than length. Having a survey that is not within your interest should be shorter for higher satisfaction than having a survey on a topic that interests you. Eg youngsters taking a survey on festivals vs a survey on washing powder ad campaign.

But I agree on length being an important dimension and the shorter the better :-)

3. All questionnaires to be pretested by and on client and research agency staff - no exceptions!
Fully agree on this one. At InSites the survey is tested by our quality team twice in the process, tested by our clients and before it goes live we ask a group of members to test the survey. We call them our Watsons (cfr other post on Research Voice http://www.research-voice.com/discussions/sharing-ideas-and-initiatives-... )

Another one that might be interesting when testing is using google translate if you don't master the languages to test. This tool has helped us a lot already.

4. Feedback loop to a human being on every survey, including sensitive and creative satisfaction measure
Here again, I do agree, and often on surveys you don't find any support address in case you have feedback or issues. We have developed a nice system for this which is called our Shoutbox. This system gives participants the change to react on a question in their own language. The feedback goes to the person at InSites involved and is being translated by google translate to english immediatly.
(details http://blog.insites.be/?p=1802)

5. Variable pricing based on average participant satisfaction scores - the lower they are on your surveys, the more you pay!

Difficult one as I think there are aspects that you cannot influence as a 'client' or researcher. eg topic involvement that drives satisfaction or different satisfaction scores based on the country (some countries score higher)

6. Rewards menus attuned to what is actually important to the participant

Can you give some more details on this one?

7. Dynamic profiling to reduce the burden of repetitive questions to which we already know the answers

Here we have a nice suggestion, our passport system that is offering the socio-demo questions that come back every time. The participant can retrieve the data from the database and update if necessary. This way they don't have the idea of being in the survey already and we real time update the data of the variables we select on the most.

8. Participants to decide by which mode they would like to be contacted (participant-driven multi-modal research!)

How do you see this?

I hope more people join this and are open about how they approach things.

sam@insites.eu

Some nice touches here, Sam. Here are some clarifications that you sought:
6. Rewards menus that are actually attuned to what is important to the participant: By using either the rewards mechanism of the site at which you found the participant or offering a menu of rewards (e.g. air miles, donations to charities, coupons..) you can make the monetary value of the reward leverage much higher in the participant's mind and so bring them more satisfaction.
8. Participants to decide by which mode they would like to be contacted: some people prefer online; others mobile; others landline. Why not let people choose how we actually contact them for our surveys?
Thanks for your feedback, Sam!

Конец взлетной полосы казался неприятно близким Подъемную силу, достаточную для полета, "Ибис" развивал при сорока пяти километрах в час