YouTube just released a feature that creators have been asking for for over 12 years and they have been promising it to us for almost just as long it's finally here for most creators and it's rolling out really quickly so I don't normally make videos like this but we really felt like we just had to get this one out to update you guys on what is going on okay so the feature is AB testing if you aren't familiar and it's actually ABC testing now because they gave us the option to run three thumbnails so how it works is you are able to upload three different thumbnails and YouTube will test them with your audience to figure out which one performs the best now this is big I mean you can imagine what a big difference this is going to make for creators if it works and so we're going to get into that and I'm also going to show you how to find out if you even have it because most people have it at this point on June 13th the time of recording this it seems like about 50% of channels have it so to check just go to any of the videos on your channel and click the little details button it can be a video I've already published or one that you haven't published yet scroll down to where you normally upload the thumbnail and there's three little dots if you click there at the bottom there will be test and compare of this little scroll down list so if you don't see it you don't have it yet or you haven't turned on Advanced features make sure you have advanced features on and here's how to do that just go check over here okay then some other really quick details first of all it's only on the computer version right now if you're looking in the app you're not going to see it no matter what and you can't test thumbnails on videos that are made for kids for mature audiences or on private videos I'm not really sure why they're doing that with made for kids or mature audiences I don't know maybe that'll change later if you don't have it yet don't worry because they say they are rolling it out to all creators in the coming weeks and they said that three days ago I mean maybe with their track record you should worry a little bit but I imagine that you're going to get it really soon because we're seeing it on new channels every day but here's the kicker we have had it for over six months and we've been beta testing it and I have some serious doubts on how well it works for all creators especially smaller creators and honestly guys I really hate to tell you that because you've probably been waiting for it a really long time if you knew about it and if you didn't know about it prob really excited to hear about this feature cuz it could be gamechanging and the feature looks so cool I mean when we first got access to it we were so excited we were running tests left and right and then we started to get really worried about some things but we're going to cover that in a future video coming really soon and maybe those problems are why it's been in the works for so many years and they keep promising it to us and telling us it's coming soon it's coming soon and we're just now finally getting it while researching this video I found a hilarious Reddit post about just this thing how long we've been waiting for it so I'm going to put that at the end of the video okay let's dive into the nitty-gritty on exactly how this works and all the little nuances with the features so the first big thing to cover is that they are measuring with a metric called watch time share now if you ever keep an eye on your analytics that's probably really surprising to you because thumbnails and clickthrough R kind of go hand in hand but instead they're tracking it by looking at how much watch time each of the thumbnails got you at the end of the day so here's an example let's say that YouTube showed all three of your thumbnails to 200 people each and for thumbnail a 100 people clicked on it and thumbnail B 50 people clicked on it but with thumbnail a those 100 people didn't stick around very long because maybe the thumbnail was a little bit click baity and the video didn't seem to deliver on what they expected to get and then on thumbnail B it was only 50 people but those people actually watched a good percentage of the video so overall those people spent more time on your video than the hundred people then YouTube would say that thumbnail B is doing better now a lot of people are really upset about this and I kind of get it I would love to see click-through rate and other metrics but if they can only show us one thing at this time then I get that watch time share probably does make more sense than click-through rate there's a lot of things with this that I hope get updated in the future now the second big thing to cover is that you won't see any percentages at all on how these thumbnails are performing until you have a lot of impressions for a while they were saying that it's 5,000 and I can't actually find anything documented on that so they might have changed it or they might just want to have the flexibility to change it but based on a test we ran just the other day it does seem like about 5,000 is when you get to see those percentages so will it work on really small channels maybe not so they let you run the test for 2 weeks but if you're not getting 5,000 Impressions within the span of 2 weeks you're not going to be able to see any data at all and the test will just end and it will automatically select the top thumbnail and I don't mean the top thumbnail that performed the best it will just be the number one the first one that you uploaded so will this work for very small creators I don't think so but let's say you get those 5,000 Impressions within the 2 weeks that you're running the test there's still three possible outcomes there is a definite winner label a preferred label and an inconclusive label now I'm not going to read the entire help page I will link it in the description or even the entire paragraphs on each of these things but I do want to point this out and quote them verbatim because I think we have data that proves otherwise and so if I'm going to disagree with them I want to at least make sure that I quote them accurately so they say once the test is complete you'll see a report with your result you'll see a winner label if a thumbnail clearly outperformed other thumbnails based on watch time share and we're sure that these results are statistically significant okay moving on cuz we're not getting into that yet so if that happens YouTube automatically sets that as your thumbnail you can change it manually but they will just kind of automatically set it okay let's keep going they say more you might also see a preferred label which means the thumbnail likely outperformed other thumbnails based on watchtime share but we're less sure that viewers prefer this thumbnail compared to one that got a winner result now I like that they do this because if they are not sure a really should be telling us that they are not sure and if you run this test on your own I really don't want you to give too much credit to this preferred thing especially especially if the thumbnail test isn't done running and they haven't given you any label listen to me listen to me I do not care if one of your thumbnails says it's getting 58% and one of your thumbnails says it's getting 29% you cannot be sure that that 58% one is any better or any more appealing to your audience or the larger YouTube audience that doesn't know about about you yet than the 28% video trust me on that okay sorry I'm getting off topic again we're saving that for later okay they go on to say if there's no clear winner the first thumbnail would be selected and shown to your audience if there's no clear winner trust them on that there is no clear winner and don't give any weight to it even if one shows a higher percentage than the others okay so then they give this little disclaimer remember that regardless of test results choosing the thumbnail is up to you you can always manually select the video thumbnail you want to use even if it wasn't the winning preferred option which is cool I'm glad that let you ultimately make the decision because ultimately you really should be okay a few more really important things so like I said we have had this feature for over six months and if you want to ask us questions about what we've been testing we are going live on Monday June 17th at 8:00 a.m. Mountain Standard Time daylight time whatever Mountain Time Nathan's going to be back in the office and we're just really excited to do that for you but I don't even know if we're going to keep that video listed we might just completely unlist it after the live is over so make sure you're there but we're also going to compile all of our data and tell you everything in a video that's coming up really soon probably next week if not our normal publishing day on Monday the 24th if that video is already out by the time you're watching this you can watch it here and if it's not here's some more thumbnail tips right here and as promised this hilarious Reddit post hope you have it all the best on testing it we're so excited to explore this new feature with you guys [Music] oh [Music]