Did you have any thoughts regarding the attractiveness of the people in the videos?

E.g. the video in another commenters post (the wooden remote) obviously hired models to appear in them.

Awesome article!

It says you asked 100 campaigns. Does that mean you got 100 answers or you only 100 campaigns? I'm curious as to how the bias of who responded to you would affect your data.

That's great to hear! We contaced many campaigns, and received answers from 100 of them. We are very grateful to the companies who shared their video info. This kind of transparency is much needed, in particular in video production. In terms of methodology, one third of the sample is from randomly selected campaigns on Kickstarter / Indiegogo, one third are campaigns that made videos with Videopixie creators (I am a co-founder), and the remainig third we targeted specifically to cover the full range of campaign sizes. Obviously we'd love to do this with even more campaigns. This was kind of a proof of concept for us. People seem to benefit from the data, and we have a methodology in place, so I hope we get to do more of it!

Did you do any studies on any pure software products, such as a videogame?

Only a few datapoints, but that's a good area to explore next. My hunch is that the findings will be quite different because there's a wider range of video formats for games. In particular much longer intros to set the stage.