00:00so if you don't like this video are you
00:01just going to not recommend it to
00:02anybody uh it's not my decision hello
00:04insiders I am Renee Richie your friendly
00:06neighborhood Creator liaison and today
00:09we are talking to Todd opre Todd what is
00:11it that you do at YouTube so I lead the
00:13growth and Discovery team so there's a
00:15couple outliers that I often hear from
00:17creators and one of them is they have a
00:19channel they're doing a certain kind of
00:20video and then they just decide to do
00:23another kind of video and it takes off
00:24it it blows up it becomes viral they get
00:27so many views from it they start to
00:29worry that it's going to distort how the
00:31algorithm sees Their audience yeah so I
00:33think it's important to understand that
00:34for the most part the algorithm for
00:37Discovery is focus more on individual
00:39videos than like trying to come up with
00:42some comprehensive average performance
00:44of a channel similar to how you don't
00:47want your videos to be shown to the
00:49wrong viewers such that they won't
00:51engage and it'll hurt your performance
00:53we don't want to be showing the wrong
00:55videos to the wrong viewers either and
00:57so we try as much as possible to
01:00understand that video's audience
01:02obviously if there's high demand for
01:04that topic yeah um and you choose not to
01:07make videos on that topic more then
01:09maybe you're you won't continue to have
01:12that outside performance um that's
01:14something you just need to decide um
01:17based on what your interests are and and
01:19what the audience interests are and find
01:21the best match uh for you so I don't
01:23need to go through and delete all those
01:24videos no I wouldn't recommend deleting
01:26those videos because there will be a
01:29subset of that audience who will be
01:30interested in your other content yeah
01:33generally I wouldn't advise deleting
01:35videos um because I think you're just
01:38removing opportunities at that point
01:40because I was afraid of this at first
01:41too that if I do one bad video it begins
01:43like the downward spiral of my channel
01:46but it sounds like you're not going like
01:48based on channel and the Creator but
01:50more on those video by video basis yeah
01:52if if your last video wasn't so great
01:54and your next video is great we want to
01:57we want to realize the potential of each
01:59video yeah we wouldn't want the system
02:01to just decide oh one bad video this
02:04this channel is not able to create
02:06anything because that's not the reality
02:08next video could be the best video in
02:09the history of YouTube it could one of
02:11the other things that you said to me
02:12that really really changed my
02:14perspective on YouTube was that YouTube
02:17does not push videos for creators
02:20YouTube pulls videos for viewers could
02:22you share that a lot of creators their
02:25kind of uh interaction with YouTube is
02:27they make a video and then they upload
02:29it and and then out from the system they
02:32get impressions and Views and things
02:34like that so I think it's natural for
02:36creators to think of like oh well I give
02:38YouTube my video and then YouTube pushes
02:41my video out to a bunch of people but
02:44it's actually uh more the reverse so yes
02:47you upload your video to YouTube but
02:49what triggers a viewers seeing an
02:51impression of your video isn't some you
02:55know YouTube plan that says oh look
02:57here's a new video from Renee I'm going
02:59to I'm going to give this to 10,000
03:01people and then see what happens that's
03:03not how it work for T ISRO all the
03:05videos yeah instead what happens is we
03:08don't we don't calculate recommendations
03:10until a viewer actually uh appears on
03:14YouTube they open up the app or they
03:15open up the YouTube homepage it's at
03:17that moment that the recommendation
03:19engine is engaged and the challenge that
03:23the recommendation engine is given is
03:25given this viewer who's coming to
03:30YouTube from this type of device at this
03:32time of day with this history of of
03:35videos they've watched before what are
03:37the best videos that we could show them
03:39right now and so we are very centered on
03:44that viewer and finding the best videos
03:46for that viewer that also ties into
03:48something else you told me which I
03:49really loved and that is the algorithm
03:51doesn't give up on video so creators
03:53shouldn't give up on them either yeah I
03:54see creators focusing so much often on
03:58their real time Analytics imediate yeah
04:00and I get it like you spend a lot of
04:02effort to like make a video you wanted
04:04to you know reach as many people as
04:07possible and you're looking at it and
04:09you know I'm always hearing from
04:10creators talk about oh this video was a
04:12nine out of 10 and yeah you know and
04:15they're just kind of down about it what
04:18I often try to say to creators to
04:20reassure them is like how many times
04:21have you seen where a video started out
04:24as maybe a nine out of 10 nine out of
04:26your last 10 but then you know a few
04:28weeks later ended up being a 5 out of 10
04:31or a four out of 10 so while I
04:33understand the focus on wanting to get
04:36the immediate uh excitement from the
04:38audience when you upload it one of the
04:40great things about YouTube is we don't
04:41limit recommendations to just videos
04:43that have been uploaded this week for
04:45example and so over time there's always
04:47going to be more opportunities for your
04:50video to get you know connected again if
04:53if there's like renewed interest or just
04:55kind of a different Trend flows through
04:57the ecosystem sure look at your early
04:59dat data maybe uh if something's
05:02underport forming to your expectations
05:04consider whether you might package it
05:06better the better thumbnail that conveys
05:08the value of the video or better title
05:11But realize that the ultimate success of
05:15that video isn't always determined in
05:17the first 24 hours there's this other
05:19concept that creators have sometimes
05:20about the algorithm being punitive like
05:22your channel starts to get fewer views
05:24or you take a break that suddenly you're
05:26put in the recommendation Penalty Box so
05:29to speak I'm sure that's something that
05:30you hear often how do you address that
05:32when creators bring it up we want to get
05:34every video out to the audience that
05:36would be most interested in it we don't
05:38want to be putting in logic in the
05:40recommendation system that that draws
05:44the wrong conclusions about a video that
05:47just came out and so we want to be led
05:49by the audience and so we aim to not
05:53overemphasize historical data if that
05:56data isn't particularly predictive of
05:59each video and so that's why we tend to
06:00lean more on the per video you know
06:03performance and feedback uh than try to
06:06like come up with these like penalty
06:08boxes for channels we don't want there
06:10to be a a player in the Penalty Box when
06:13they could have been scoring goals yeah
06:15right and so uh so we really try not to
06:18to add those sorts of rules and and
06:22enable videos to freely find their
06:24audiences I've heard you talk about the
06:26algorithm follows the audience but do
06:28creators have to think about that as
06:29well like reinvention and updating and
06:32trying new formats do creators have to
06:34follow the audience as well yeah I think
06:35that's really important I think shorts
06:37is a interesting case study in this
06:40space a lot of uh creators they may say
06:44like oh well YouTube is pushing shorts
06:46on viewers and YouTube wants shorts to
06:49be more successful than long form that's
06:51not why YouTube has focused on Shorts
06:53YouTube is focused on shorts because
06:55audiences have let us there they've told
06:57us that they want more videos that are
07:00more efficient and get to the point
07:02faster they want more entertaining
07:04participatory content and we see them
07:07you know audience is voting with their
07:09behavior of the apps they're using to
07:11get that content and we feel obligated
07:14to offer that to the audiences and we
07:16want audiences that want shorts to be
07:18able to find them but also if an
07:21audience member is happy with long form
07:23we want them you know to find that
07:25easily as well so um that's an example
07:29of you YouTube following the audience as
07:31the competitive landscape changes as a
07:34Creator I would recognize that there is
07:37a lot of attention some of which is
07:39shifting from long form to short form
07:41some of which is new uh opportunities to
07:45recognize that as part of their content
07:47strategy and then within a format there
07:50are also Dynamics there's there's you
07:53know creators who might say hey you know
07:55I've been doing this the same content
07:57for a couple years now and it's not
07:59performing as well anymore and sometimes
08:01I'll look into you know work with them
08:04and encourage them to look at their
08:05audience Tab and see well what are some
08:07other channels your audience is watching
08:09and we'll find that oh there's a new
08:11channel that came out in this that
08:13that's also serving the audience and you
08:17know that channel wasn't around two
08:19years ago and and now you have to
08:22recognize that there's a bit of supply
08:24and demand going on and when there's you
08:27know more Supply you know with the same
08:30amount of demand you may get less views
08:32and then you got to figure out well
08:33where is there unserved demand that you
08:36can maybe tap into um or where can you
08:40do deliver better on the demand that you
08:43started with if creators looked at their
08:45content strategy similar to like any
08:47small business or or business that's
08:49trying to figure out well you know how
08:51do I maximize my Revenue I I think about
08:54things like market share and competition
08:55and things like that so those Dynamics
08:58are are very uh relevant in the YouTube
09:00ecosystem as well I make a video and
09:03then I see someone else has made a video
09:05that to me feels really similar but they
09:07get 10 times the views that I get or
09:10maybe like the topic I see like gets
09:12tons and tons of views but I'm not
09:14getting them when you hear about those
09:15kinds of things what are some things you
09:18look at to understand performance one of
09:20the things that I'll look at and
09:21encourage creators to look at on their
09:24videos is how did the uh viewers who saw
09:28and impressions of the video in the
09:30subscriptions feed react because that
09:34feed it doesn't have any like special
09:36recommendations algorithms it's pretty
09:38much always the people going into the
09:40subscriptions feed you could go into
09:42traffic sources browse features and then
09:45subscriptions and most of the time when
09:48I do that um I find that videos that
09:52underperform also underperform when
09:55they're shown to subscribers in that in
09:58that feed which suggests to me that it's
10:02some issue with the content or the way
10:05the content was packaged you could also
10:07look at the the average V view duration
10:09for that audience that subscriptions
10:11feed offers a bit of a control over
10:13other variables that can lead
10:16click-through rates to for example go
10:18down with a broader audience I think
10:21sometimes too like when I look at those
10:22videos I see that they're topically or
10:24structurally the same but the feeling I
10:27get from them are very different and
10:28that's not always as easy to measure
10:30yeah it's hard to measure and uh you
10:33know viewers can be finicky and and
10:36quirky and it's sometimes hard to uh to
10:40understand why some things are
10:42successful when similar things are not I
10:44think this is something that's always
10:45been the case with with media uh even
10:48before YouTube DOD thank you so much if
10:50you have any comments if you have any
10:52questions please leave them down below
10:53and then keep it real