Where is the hard data (or as close to it as possible) regarding the effect of autoplay videos?

[Netmark]

Where is the hard data (or as close to it as possible) regarding the effect of autoplay videos? I seem to be in agreement with the other "opinions" that I've found, about them being annoying. But does anyone have statistical evidence to either support or refute this stance?

[RyanPurkey]

CrazyEgg--the heat mapping / user interaction site--wrote this blog post back in 2013: http://blog.crazyegg.com/2013/08/12/reduce-your-bounce-rate/ with #7 being exactly what you're asking about, "Nothing’s worse than being blasted with background music or a voice when you’re not expecting it – especially if you’re at work or you have multiple tabs open and can’t figure out which one the sound is coming from!"  Many people do break time shopping / research while in the office so having an audio play video or song is almost a guaranteed bounce. CrazyEgg didn't put the stats up to go with this but they're certainly speaking from experience.

[EGOL]

Here is my hard data.

For autoplay ADS, I close the window without reading the article 100% of the time.

For autoplay content videos, when I am looking for items to link to from my news page, I find the same topic on a competitor website and link to it 100% of the time.  I am not sending my subscribers to any type of autoplay.  They can cuss about me for other reasons.

[RuthBurrReedy]

Honestly I think this is a question that is going to vary based on a lot of factors, including:

• Where the autoplay video is happening (on Facebook, Twitter, YouTube, or your site)
• Who your audience is and what they're doing when they see the video
• The context of the video (is it an ad, or is it part of your site content?)

Some things to consider when using autoplay videos:

• Mobile users tend to find autoplay videos especially obnoxious because they use up so much data.
• Autoplaying videos may inflate your video views count, since the view isn't user-directed; autoplay video engagement should be measured using additional metrics such as time viewed, additional videos viewed, or clicks-through.
• Sites such as YouTube and Netflix have seen increased engagement through strategic use of autoplay, starting another video as the first one ends.

I would recommend testing this out for your site or sites and seeing whether autoplay increases or decreases engagement and conversions. Anecdotally, a lot of people say they dislike when videos autoplay, especially with sound, but anecdotal evidence is no substitute for a good test, and you seem to be correct that the results of this sort of test aren't readily available online. If you do run a test, it would make an excellent  YouMoz post!

[RyanPurkey]

Ruth brings up a good point, autoplay can work when it's expected, but most sites aren't Youtube, Netflix, or Vimeo. Still if your site is in the category of "all we do is videos" then autoplay may work for you.

[AlanMosley]

They turn me off,