Board Thread:Consensus Track/@comment-20644-20180423192744/@comment-20644-20180427211548

Thanks everyone for your patience as I put together a response to your questions. It seems that the three big questions and points of feedback here can be summed up as:


 * Why is the player at the top, why does it follow you, and why does it autoplay?
 * How do the metrics that I referenced earlier justify success?
 * How can users be involved in a meaningful way?

I will address each one of those questions in their own section. If I missed anything, please let me know so I reply to that as well.

Why is the player at the top, why does it follow you, and why does it autoplay?

Featured Videos are intended to do two big things: introduce a new type of content for an audience that is increasingly looking for video (by 2020, 80% of online traffic is expected to come from video) and clean up the advertising experience on FANDOM. The size, placement, and autoplay are a big part of how advertising on Featured Videos has enabled us to reduce the number of ads per page.

One of the most important things to advertisers is what's known as ad viewability. Google has a great page here that explains the details of viewability if you want to read more about it. But for the purposes of the Featured Video, the long and short of it is that the placement and size of the player are key factors in determining whether an ad is viewable. That means we have to strike a balance between doing what can derive the most value for FANDOM as a business and what provides the best possible experience for the audience. If we went too far in the former’s direction, we could have a much larger video player that would achieve even greater viewability but with a very poor audience experience. And in fact over a year ago we tested various different size options, including a much larger player, but the one we have now achieved all of the goals while balancing the user and advertising experience. In the other direction, having a video that's too small wouldn't achieve enough viewability or the goals of cleaning up the ad experience.

Autoplay, meanwhile, has allowed us to achieve not just viewability, but enough views on advertisements in order to be able to clean up the ad experience. Without autoplay, we would not have been able to clean up the ads on FANDOM. Please note, though, that the player autoplays without sound. If you’re seeing auto-sound, that’s a bug and I’d encourage you to report it.

How do the metrics that I referenced earlier justify success?

To reiterate, here are the metrics that I talked about in my previous comment:


 * Completion rates - how much of the video did a user watch?
 * Page bounce rates - does a video cause someone to leave the page?
 * Videos per user - how many videos does each user watch?
 * User retention - does the addition of video get people to come back more often?
 * Rate of unmutes - are people watching the video with sound or just letting it play silenty in the background?

BerserkerPhantom suggested the numbers here wouldn't be meaningful because the video autoplays. That’s true, and it’s why I didn’t include views in those metrics. Because it's not enough to just say "we got a view!" and call it a day. We consider those to be video starts/plays. And because of the autoplay nature of the videos, it means that the bare bones metric of a video play/start isn't enough to judge success on. Starts/plays are important from a holistic standpoint, but on a community to community or video to video basis they're much less important in determining whether a particular video or set of videos were successful. I myself am much more interested in how unmutes and completion rates tie together, because those two together are a more accurate reflection of whether a user is engaging with a video and whether they like it enough to continue on with it. Page bounce rates are also important because they tell us whether people are leaving the page because of video, and our data shows that they are not.

The metrics I provided are all about engagement with the video and engagement with the page. And the numbers we see there have been successful thus far. It shows us that there is an audience out there watching these videos and engaging with them.

And that’s a key point. It’s true that wiki editors on some wikis have expressed concern about and disagreement over having videos on the page, including here on Narutopedia. But you also have to remember that editors, on all wikis, are just one part of the overall audience. And just because one part of the audience isn’t interested in videos doesn’t mean that we’re not going to provide them for the part of the audience that is interested in them, because that audience is growing larger and larger and it’s important that FANDOM provides the right kind of content for all audiences.

The idea of the ‘right kind of content’ brings us to the third question.

How can users be involved in a meaningful way?

That goes back to what I said in my last comment about script review or script writing. If the community wants to be part of script review, then any new script that we write will be sent to the user(s) appointed to review scripts before they’re produced so the user(s) can provide feedback on that script. If you want to actually write scripts yourself, you can submit those to us and then we will produce them.

If there are videos that are live now that contain inaccuracies, point those out to us via Special:Contact. We’ll take those inaccurate videos down and redo them. We can have you review the new scripts for those videos, or users here can write a new script for those videos.

Featured Videos are a global feature, which means that it’s not a feature that will be removed on a per-wiki basis. But we also want to make sure that the content is accurate and geared towards the audience of the wiki, which is where your involvement can come in to help us do that and help shape the Featured Video content.

Let me know if that’s something you’d be interested in.