💥 In This Training:

https://magnates.media/youtube

● How does the algorithm really work?

● What the algorithm optimises for
● Why most info about the algorithm is wrong
● Why it’s not as simple as CTR and AVD like many say
● The underappreciated importance of ‘channel reputation’
● Understanding analytics in context to make better decisions
● Analysing why some of my videos did better than others
● The traits of a viral video
● And how to actually get more views and subscribers.
🤖 How Does The Algorithm Work?
https://magnates.media/youtube

Algorithm 101: The Fundamentals

The job of the YouTube algorithm is to provide users with the content they want to watch so
they'll spend longer on YouTube.

The term YouTube use is that they’re optimising for viewer satisfaction.

YouTube is a business, they want to keep users there as long as possible so they’ll see
more ads, and YouTube makes more money - so of course this makes sense.

The algorithm is not rigged - it’s in YouTube’s best interests to promote whatever content is
keeping users on the platform more.

So if you make great videos on a new channel, you’ll grow. Most people just don’t know how
to make a great video.

The algorithm simply follows the audience. This means whenever you hear the word
algorithm, just think audience.

So if you’re asking ‘Why doesn’t the algorithm like my videos?’ instead ask ‘why doesn’t the
audience like my videos?’

As creators we think youtube should go and recommend our content to viewers, but youtube
actually does it the other way around; it starts with what do viewers want, and then tries to
pull in videos it thinks they’ll be interested in. So we need to always be thinking what the
viewer wants.

If you want to go viral, think more about human psychology than algorithms. You’ll probably
learn more from studying great marketing or movies & TV shows, and how they hook your
attention and keep you engaged.

But in very basic terms we’re trying to get people to click our video, then watch our video.
Which is why it’s not surprising that 2 of the main metrics the algorithm considers is
clickthrough rate (what percentage of people who are shown this video actually click on it?)
and average view duration (what is the average length people watch your video for?)

Because if people are choosing to watch your videos more often than other videos, and
they’re watching your videos for longer - of course the algorithm is going to want to push
your videos.

Many new creators - myself included when I started out - try to optimize for search so people
can find our content. However if you are trying to maximize views and reach as many people
as possible, you actually want to optimize for ‘browse’ and ‘suggested’ which is where
 https://magnates.media/youtube

YouTube recommends your videos on the homepage or the sidebar next to the video
someone is already watching.

Search can bring in a slow long-term trickle of views, but it’s just people who are actively
looking for specific content. A video goes viral because of browse + suggested traffic, as
people weren’t necessarily looking for that content, but YouTube has put it in front of them as
it thinks the viewer will like it. When MrBeast posts a video like 'I built Willy Wonka's
chocolate factory', people aren’t actively looking and searching for that, they find it because
of YouTube recommending it. This is how videos rack up a lot of views very quickly, and it's
what we want to aim for. So focusing on making videos viewers want to click on and watch is
the simple goal.

Remember YouTube has limited space for promoting videos, there’s only so many spots on
the homepage and suggested bar, and they want to fill those spaces with the options they
think each individual viewer is most likely to click and watch (i.e. which will satisfy the viewer
most).

So they’re essentially running continual split tests between lots of videos (meaning they
show users a bunch of options, and see which perform better). Your video is therefore in
competition with all the other videos the algorithm thinks a user may enjoy.

If initial viewers interact well with a video, the algorithm will start pushing it to a slightly wider
audience. If those also respond well, the algorithm will keep pushing it wider and wider,
which is how a video can go viral if it keeps winning those split tests against competing
videos. And if you use all the strategies in this course to increase watch time, clickthrough
rate, and engagement, your video will have a very good chance of getting chosen and
watched more often - and thus getting promoted more.

But if you want a video to go viral, the video has to appeal to a large audience (so the video
idea needs to have a large total addressable market of people who are interested in this
topic). Simply expanding the videos’ scope and picking ideas that could appeal to a broader
audience is key here - if the idea sucks and nobody wants to click and watch (or only a very
tiny niche amount of people do), it doesn’t really matter how good the video is.

Then can you package your idea in a way that both your main audience and totally new
viewers both want to click? If it’s going to go viral, it needs to appeal to people who aren’t
already fans of your channel.

However even though average view duration (AVD) and clickthrough rate (CTR) are
important metrics, viral videos tend to do well on both of these metrics, as it’s the interplay
between them that’s key - having a higher AVD AND CTR.

That’s because the ACTUAL key metric the algorithm uses to rank videos for viewers is:
Expected watch time per impression. (so it’s factoring in both how likely they are to click
AND how likely they are to watch - i.e. both AVD and CTR).

Because you could have a thumbnail and title that’s incredibly enticing to click on, but if it’s
then a bait and switch and you don’t deliver on that (so the thumbnail and title is misleading),
 https://magnates.media/youtube

people will click off the video early. And we know that’s not what YouTube wants - it wants to
keep people watching. So CTR may be high but if people aren’t satisfied with the video,
YouTube isn’t going to recommend it as much.

On the flip side if AVD is high but hardly anyone is clicking in the first place, it’s probably too
niche and not enough people are interested in the topic or the way its been packaged. So
YouTube is less likely to use up the very limited amount of spaces it has to push your video
as heavily to a wide audience when only such a small amount of people will click it.

So ideally you’d have a bold promise in your title or thumbnail that lots of people click on, but
then you actually deliver on it and keep people watching the whole video. That way the
expected watch time per impression is going to be higher and YouTube will recommend the
video more.

This video of mine makes a bold claim: how a kid with $0 created Nike - it’s also got
elements that hopefully create intrigue like the blood. But then the video delivers on exactly
what’s promised - it tells the story of how he did it. So even though it may seem slightly
clickability, you do find out what the title and thumbnail promised so you come away
satisfied.

There’s entire trainings in this course dedicated to thumbnails and titles (to get people to
click) and also retention (to keep people watching). But really the entire course is geared
towards getting you more views, as the course is largely about making better content. When
we think of the algorithm we often think about ‘algorithm hacks’ but it’s not; the answer is
making videos more people wanna watch, so everything we’ve covered previously like
coming up with better ideas and storytelling techniques - all of that is what’s actually helping
you optimise for the algorithm to push your videos more. As remember algorithm = audience.
When we make videos people enjoy more, the algorithm recommends them more.

People worry about the ‘algorithm changing’ but these are just minor tweaks, and the
fundamentals you should be focusing on don’t change. The algorithm is still going to be
optimising for viewer satisfaction. So you should too.
 https://magnates.media/youtube

Algorithm 102: What YouTube’s Research Paper Says

Even YouTube’s own developers don’t fully understand how the algorithm makes decisions,
as it uses machine learning to understand complex patterns.

However Google has published a research paper which tells us all we need to know to
understand what’s going on.

Firstly, each part of YouTube - like the homepage or the suggested videos in the sidebar -
technically each has a separate algorithm.

But overall what YouTube has is a recommendation system. It’s a matchmaking system to
match the right videos to the right viewers. And this diagram illustrates how it’s essentially a
funnel…

We start off with millions of videos.

But then at the candidate generation phase the system considers factors like:
● Past videos the user has watched (so it can see what they’re interested in)
● Videos similar to what the user has watched before
● Videos that other users with similar tastes watched (For example it may learn that
users who watch lots of historical documentaries also often enjoy lectures by famous
professors).
● Previous search queries and values that may be relevant to the user
● Demographic information
… And using all of this it can create a profile of the user's interests and 100s of potentially
suitable videos are generated.
 https://magnates.media/youtube

But then YouTube needs to rank them. The ranking stage will include factors like time of day,
device the user is watching on, how much interaction a user has had with a specific channel.

For example if you watch a channel a lot, you are more likely to get recommended videos
from that channel. If you don’t want a channel much you’ll get recommended it less (even if
you’re subscribed).

Likewise if you watch certain types of videos more on certain devices (e.g. shorts more on
mobile, longer videos on TVs) or you watch certain genres of content more often at night, it
should factor all of that in when ranking results for you.

The more a viewer uses YouTube, in theory the better it should get as it understands their
personal preferences more as it has more data.

But what is actually happening is YouTube uses a deep neural network - which means they
train the system by giving it a vast amount of data about users, and then this network learns
patterns to predict which video a user might watch next. The one with the highest predicted
watch time per impression is recommended.

So in summary: the system tries to predict how long a specific user would likely watch a
video. This involves factors like the video's content, length, and how well it aligns with the
user's interests. The idea is to suggest videos that users are most likely to find interesting
and watch for a longer duration, thus increasing their time (and satisfaction) on YouTube.

When making these calculations, there’s direct feedback the system uses (like thumbs up or
down, and surveys), but it also uses implicit feedback - for example when you complete a
video that’s a positive example that you were satisfied with the video. Whereas if you open a
video then quickly click off, that’s negative feedback. (Hence why high clickthrough rate
alone may not get the video recommended if people aren’t watching a lot of the video and
are leaving unsatisfied).

If you want more insights on the algorithm you can follow Todd (@hitsman) on X who has
been product lead for YouTube homepage and recommendations since 2014.

Although a lot of the advice will remind you that it’s actually doing the simple things that
matter:
 https://magnates.media/youtube

When it comes to the algorithm, you don’t need more info - you already know what to
optimise for - the hard part is actually executing on that… Making videos that people want to
click and watch, and then return for more of your content.

But a few things to note:

1. A channel with similar predictable content is a lot easier for the algorithm to
understand, categorise and recommend. This is why bouncing between many
different niches or topics that appeal to different audiences generally doesn’t work as
well. The clearer it is who your videos are for, and the more you can be consistent
with that, the better.

2. We know the system is trying to promote videos with high expected watch time, so
we need to focus on creating engaging content that keeps viewers hooked from start
to finish. (For example: a strong hook initially, then good pacing, storytelling, and
giving them reasons to stay engaged throughout).

3. Even though getting people to click (ctr) and watch (avd + watch time) are key things
to focus on, the algorithm also pays attention to engagement metrics such as likes,
comments and shares as they all signal that viewers are enjoying the content. So you
can encourage engagement by asking questions or asking viewers to like and
subscribe.
🤯 Why It’s Not As Simple as CTR +
https://magnates.media/youtube

Watch Time
Typically as a video gets pushed to more people, it will get pushed to ‘colder’ audiences who
aren’t as familiar with your content, and as a result CTR and AVD will decrease as the
audience isn’t as engaged or loyal.

When a video first goes live YouTube tends to recommend it to your most loyal audience
first, as their watch history indicates they regularly watch your content so this new video is
likely a good fit for them.

If they respond well, YouTube may push it to your ‘casual audience’ - people who sometimes
watch your content.

If they respond well, YouTube may push it more heavily to totally new audiences who have
never seen your videos.

The average view duration of one of your regular returning subscribers is likely to be higher
than someone who has never seen a video from you before. Same with clickthrough rate.

So just remember, it’s very normal that as the video gets pushed to a wider audience, AVD
and CTR typically decrease. That’s not a problem. And it doesn’t mean you should go
changing your thumbnail and title as the fact it is being pushed widely indicates the video is
performing well.

Additionally, the attributes of a video are going to affect the metrics. Of course it makes
sense a 1 hour long video may have a lower clickthrough rate but higher average view
duration.

It’s also worth noting that different traffic sources impact these metrics - for example external
traffic (e.g. your video gets shared on Reddit) usually result in lower average view durations
as people are less likely to watch for as long as someone who is on YouTube.

So, all of this is why you can’t just look at these metrics in isolation with no context.

My most popular videos have lower clickthrough rates than some of my least viewed
videos.
 https://magnates.media/youtube

That doesn’t mean I should go change the thumbnail and title on my most popular videos, it
just means its been pushed to such a wide audience outside of my typical audience - but
that’s a good thing. (Of course if something gets over 166,000,000 impressions - a lot of
those people aren’t going to be interested in the video).

Likewise, the average view duration on this video is 13:10 - which is decent but when you
consider it's a 33:50 video it’s not exactly incredible (average percent viewed is 38.9%). But
again when you consider the context that 6.5 million people have clicked on it, that means
it’s generated 1.4 million HOURS of watch time.

And thus the most important metric to tell if your content is doing well is simply
views. This is the metric that ultimately gives you the best idea of how your video did.
🧪 Analytics In Context To Make
https://magnates.media/youtube

Better Decisions
If a video is underperforming expectations, then changing the title or thumbnail can be one of
the highest leverage tasks, as it’s very easy to do and can have an absolutely huge impact
on how well your video performs.

But bearing in mind everything we’ve just talked about, clearly CTR alone in isolation doesn’t
tell us if the thumbnail is working well or not. We need context.

The metrics in your YouTube studio are influenced by so many other factors:
● Views and impressions (e.g. more views typically means a lower CTR & AVD as the
video reaches audiences less familiar with your videos)
● Content format and length (e.g longer videos are typically clicked on less but
watched for longer)
● Traffic sources (e.g. external views tend to have lower retention)
● Device type (e.g. mobile views are often shorter than views on a TV)
● Seasonality and current trends (e.g. certain videos are more relevant at specific
times which are outside of your control, like Christmas related videos being watched
more at Christmas)
● Supply and demand (viewers may have seen videos on this topic already so aren’t as
interested as the demand has been fulfilled or there’s so much competition)
… And many more factors which could influence your analytics.

So this is why you can’t just look at your clickthrough rate and go “oh that’s low, it must be a
bad thumbnail or title” - there’s a lot of variables that are unrelated to the thumbnail or title.
The truth is we can’t always fully understand all of these variables, but we can at least apply
context to try and make more informed decisions. Just like a scientist, we can put forward a
hypothesis and then test if it works.

If I upload something that’s a similar length and style to my other videos, and I thought the
topic was perfect for my audience, and I can’t think of any external explanation why they
wouldn't be watching, then I may try to switch the thumbnail or title if the clickthrough rate is
lower than normal. I would usually just change either the thumbnail or title so I can see the
effect that one change has.

The key metric I then look at is view velocity. (Because remember the goal is more views,
not a higher click through rate - and if the video does get heavily pushed CTR will likely go
down). So if I’m changing the thumbnail on an older video that’s already been published a
little while, I’ll note how many views the video got in the last 48 hours, then return exactly a
week later and see if the views in the last 48 hours are higher or lower. If there’s been a
notable increase, it likely suggests the thumbnail change has had a positive impact.

YouTube is rolling out thumbnail split testing features to make this more automatic, and
Tubebuddy also offers an A/B testing feature.
 https://magnates.media/youtube

But remember that thumbnails are subjective - which also explains why tests that run early
and are being shown to your core audience of viewers who know you, may give different
results than when the same thumbnail is shown to a more casual audience who’ve never
seen your content before and are less interested in the niche.

People will always have different opinions when it comes to thumbnails and titles - that’s why
netflix actually changes thumbnails to suit individual preferences, but since youtube dont
have that, we have to try and find which packaging appeals most to the widest audience if
we’re trying to maximise views. (Or the thumbnail which appeals to our ideal audience if
we’re trying to get targeted views).

Don’t always be too quick to change things right after you upload though - especially if you
personally felt really strongly about the original title and thumbnail.

YouTube shows you your video’s rank out of 10 (compared to your last 10 uploads) on the
homepage of YouTube studio, so a lot of creators panic if the ranking is low to start with. But
I've had videos that started as a 10 out of 10 later become a 1 out of 10. Sometimes the
video just takes a little more time.

I’ve found especially with longer videos, people are less likely to click it immediately as they
might not be available straight away. Of course if people still aren't clicking after a couple of
days, then I’d be more inclined to make a change to the packaging. Again, it’s important to
consider the external audience factors here and not judge metrics in isolation.

This is why it can be helpful if you upload at the same time, and focus on evergreen topics,
in a similar format - as then we have a lot of control variables that are the same every time
so we can genuinely make comparisons. (The more the variables are different, the harder it
is to make fair comparisons or draw accurate conclusions). The more you can compare
videos at similar view counts and similar time horizons and similar length / formats / styles,
the more accurate those comparisons are going to be.

This way you’ll learn what the baseline is for your channel so you know if a video is
underperforming or not.

Once you have posted a fair few videos you will also start to learn the average views you get
in the first 24 hours, and so if you have a video that’s significantly lower, you could then try
switching the thumbnail to see if the views notably increase.

I also recommend you only compare with your own videos. There’s no magic number of
what’s good or bad when it comes to YouTube metrics, and different creators with different
audiences in different niches will have totally different baselines of what’s a high CTR / AVD
for them. Focus on comparing your own data, and trying to compete against your own stats.

Honestly, I don’t analyse analytics too closely. It's so easy to draw the wrong conclusions.
The truth is you may change the thumbnail and the video starts doing better, but it could
have happened even if you hadn’t changed the thumbnail.
 https://magnates.media/youtube

So VIEW COUNT is the metric I track. I know how many views my videos average after X
amount of time, so if it’s underperforming, I’ll usually make a change as a test. Because
ultimately views is the most accurate way to say how well a video has done (which sounds
obvious but most people lose sight of this and get caught up in other metrics).

And if a video is already performing well, I find it’s best to leave it alone.

Personally, I would recommend worrying more about changing older videos of yours that
have clearly underperformed. Because there is nothing to lose there.

If it’s been a couple of weeks and the video has massively underperformed your average
views, that’s a perfect candidate for trying something different with the packaging.

And I can’t stress enough: a simple change to the title and thumbnail can cause a huge
increase in views (not because of the change itself, but because more people click on the
new packaging and thus the algorithm recommends it more). So definitely go back and
change older videos. (Of course if your older videos aren’t very good, there may be better
uses of your time).

But you can figure out where the biggest opportunities are; like high watch time but low
clickthrough rate and low views. If the video has lower views than normal but good watch
time, it’s likely the packaging was a problem - so could you reframe the title and thumbnail to
appeal to a wider audience? (Because if once people click on the video they are watching a
lot of it, the issue is probably that just not enough people are clicking in the first place, so it’s
worth testing different packaging variations).

Remember you can change the packaging as much as you want - but give it some time to
actually see if it makes a difference as it won’t be instant. Keep a spreadsheet of what you
changed and when. Then come back and see what effect its had on the views.

If it’s a video you felt had a lot of potential and should have done better, definitely don’t give
up on it - it’s always worth testing, as you can get a huge increase in views on past videos
for relatively little work.
⌚ Increasing Watch Time
https://magnates.media/youtube

Longer Videos Are Great

Technically, a video should be as long as it needs to be. You don’t want to stretch content out
for the sake of it if it’s going to make the video boring or feel like a waste of time.

But all the data I’ve seen suggests long videos do have an advantage (if you can execute
them well). This makes a lot of sense - it’s much easier to get a higher average view duration
on a 30 minute video than a 5 minute video. Even if someone only watches half of the 30
minute video, that’s still triple the watch time of if they watched the entire 5 minute video.
And we know that YouTube wants to keep people on the platform for longer, so of course
generating lots of watch time will be rewarded.

And exceptionally long videos (like an hour or more) are quite remarkable in themselves,
people may click out of intrigue wondering what this is - but then as long as you hook them,
they keep watching longer even if they didn’t intend to.

Longer videos also typically have higher RPMs as people see more ads when watching for
longer, which can give you the money to reinvest in longer videos.

And because it’s harder to do, there’s less competition. If you’re competing on shorter videos
like 5 or 10 minutes, there’s so much content that length that getting a higher average view
duration than other videos is actually really difficult. But with these longer videos I find it’s
easier to get above average audience retention rates as less people are making these
longer videos.

It’s no coincidence that of my 4 most popular videos - 2 of them are over an hour long (2 of
my longest ever videos by far), and even the other 2 are long (33 and 44 minutes).

So a cheat code for watch time is literally just making longer videos, that are naturally
stories. This way you hold people’s attention as they want to find out how the story ends.
 https://magnates.media/youtube

I’ll often say things like ‘get comfy and relax’ near the start to set the expectation that you’re
settling in for a long ride. Which brings us to…

Include more stories

The easiest way to make videos longer is incorporating stories. Just throwing more and
more information in is going to be overwhelming and probably unnecessary. But the good
news is, stories are your secret weapon. They can significantly boost watch time because
once you start a story, most people want to know how it ends. (My videos are literally just
one big story normally).

Even if you’re making more educational content, don’t just give dry information, always try to
attach a story to the point you’re making. For example, explain why this point matters, or
how you got to that realization… maybe give an example of the impact that doing this thing
you’re talking about had on you or someone you heard about. Anything that adds more
context (and ideally, adds more emotion) to the point you’re making. Because, the point
you’re making may only take 30 seconds to say, but the story behind it may take several
minutes, so this will easily allow you to increase your watch time whilst simultaneously make
your videos more entertaining and engaging in the process.

It’s also worth noting the power of emotion - YouTube videos that get millions of views will
usually exploit emotions triggers in some way. So remember to ask ‘what do I want my
audience to feel?’ when making a video. Viral videos almost always elicit some kind of
emotion.

Don’t mislead people with the title and thumbnail

If you notice a big drop-off in your audience retention right at the start of your videos, the
problem may simply be that people thought they were getting something else. For example,
if the thumbnail is not at all relevant to the video, or the title suggests the video is about
something totally different (or could be misinterpreted), some people will click on the video
then exit the video early, which clearly isn’t good for watch time / AVD.

It’s like a seesaw that needs to be perfectly balanced… if the the thumbnail and title is too
clickabity and doesn’t match the video enough, retention will be bad. But if the thumbnail and
title is too boring and literal, clickthrough rate will be bad. If one of these metrics seemed
skewed, you may need to adjust the packaging. As it’s no use having a high clickthrough
rate if it’s too misleading and thus people aren’t watching the full video. You have to find the
sweet spot.

Don’t give all the gold away up front

In other words, don’t skip the foreplay. Give people a reason to keep on watching longer;
don’t give all your answers and best parts of the video away right at the start. It’s also worth
dropping hints about interesting things to come later in the video, for example mentioning
something intriguing/unusual, and then saying ‘but we’ll get to that later’. Keep giving people
reasons to carry on watching, rather than giving everything away immediately.

Utilise playlists - If you’re a relatively new channel, make playlists with really searchable
titles and descriptions, and put a mix of other people’s videos (from well-known creators)
 https://magnates.media/youtube

along with some of your videos. So if you’re in the finance niche, you could put some videos
from well known creators in the finance niche, but sprinkle some of your content in too.
When people find the playlist, it not only helps people find your content, but it again trains
the algorithm that all these videos are connected and should be recommended together, so
you can hopefully start appearing in the suggested videos on some of those other videos.

Also, I mentioned ending your videos with a clear call to action to watch another one of your
videos, but a great twist on that strategy is to send people to a relevant playlist at the end of
your video, as you can add an end card that links to a whole playlist of your other content.
This is great because once they’re in a playlist it’ll autoplay with more of your videos, further
raising watch time, but also once again training the algorithm that these videos are
connected and should be suggested more. It shows the algorithm the pattern that when
someone watches one of your videos they go on to watch more of your videos, so this is a
channel really worth promoting. Again, it all goes back to your videos being responsible for
keeping people on YouTube longer which the algorithm loves. As if you can increase
‘session watch time’ for the user, of course YouTube is likely to start pushing your videos to a
wider audience to see if others respond the same. If you can keep people on YouTube
longer, the algorithm will reward you.
💖 Channel Reputation
https://magnates.media/youtube

When it comes to clickthrough rate, and people choosing to watch your videos - the most
overlooked factor is channel reputation, and the past experiences people have had with your
channel.

Everyone has ‘instant click’ channels where if they see a video from them, they’ll click.

When YouTube recommends you a video, contrary to popular belief it’s NOT just the
thumbnail and title that determine whether someone clicks, it’s also the channel name (and
the associations they have with that channel).

These could be good or bad. If a viewer enjoyed your previous videos they’ve seen from
you, they’re much more likely to watch this next video. But if in your previous videos they’ve
seen they didn’t like it or it was full of mistakes and lies, or they felt clickbaited and you didn’t
deliver what you promised - they’re much less likely to click in future.

As a side note, it’s also worth noting there are other factors affecting CTR as well. For
example the view count plays a small factor as higher views can give social proof and make
someone more likely to click than a low viewed video. Published date is a small factor too as
people are sometimes slightly more drawn to newer videos as we all like things that feel
fresh and more relevant. And finally there’s video length which impacts CTR too depending
on what length of video someone wants to watch right now and how much time they have.

So you can’t judge CTR purely on the thumbnail and title alone. Viewers see the thumbnail,
title, channel name, video length, and video views - and all 5 elements can play a role when
deciding whether to click.

But let's focus on the element we can control that has a huge impact and is often forgotten:
channel reputation.
 https://magnates.media/youtube

For example, this video got 23 million views! But it’s definitely not because of the title or
thumbnail - it’s because it’s from KSI, and people will click more if they’ve had a positive
experience with his videos in the past. Another creator could upload the EXACT same topic,
title and thumbnail, but the video won’t do anywhere near as well because of the channel
reputation.

Or another example: Ludwig posted a MrBeast video on his channel - and although it did
very well (15 million views), it would have got significantly more if MrBeast had posted it.
This proves it’s not just the title, thumbnail or even the video - it’s who is posting it as well.
People consider the channel name and their experiences with that channel in the past.

This again reinforces why the strategy of high quality content is so much more effective in
the long term than high quantity.

MrBeast gets so many views as he’s built up such a great reputation over many years.
People believe he will deliver. They know they’ll have a good experience based on all their
past experiences, so they click. (If you have a bad reputation - or even just a ‘meh’
reputation - it works against you in the long run and makes it so much harder to earn the
viewer’s click).
 https://magnates.media/youtube

So if you can do the same where every time someone watches your video they have a good
experience, they may start clicking your videos regardless of the topic, title or thumbnail -
simply because it’s one of your videos and they know that every time they have a good
experience.

Whereas if people feel tricked by a video and feel like they wasted their time by the end of it,
they’re not gonna come back for more of your videos.

Bad videos can chip away at your reliability or credibility in the long run. You can think of
your reputation like a health bar; and it’s hard to build up but very quick and easy to destroy.

Most people underestimate the importance of a viewer's experience with your last video. If
they liked the previous one, they’ll likely watch the next one or go back and watch others.
We need them to think ‘wow that was amazing, I need more of this’ so they’re excited when
they see you’ve uploaded and binge watch your content. That’s why so much of this course
is devoted to making better content your audience will enjoy more.

It’s a vicious cycle - if people watch more of your videos, they’ll get recommended to them
more. If they don’t click on your videos (as their recent experiences weren't that good), they’ll
get recommended your videos less by the algorithm. So don’t lose sight of how important
reputation is.
👍 How To Increase Engagement
https://magnates.media/youtube

● Slightly controversial / polarizing topics do work better. I’m not saying you have to be
over the top and deliberately controversial; but safe, vanilla generic content of course
does not get as much engagement.

● Ask your audience for engagement – there’s a good reason so many big channels
ask for likes, comments and subscriptions: IT WORKS! The reality is that if someone
is reminded to like the video if they enjoyed it or leave a comment with their opinion,
they are much more likely to do it. A lot of people just forget otherwise. It’s often as
simple as reminding people “if you enjoyed this video, feel free to hit that subscribe
button as I’ve got some more videos on the way I think you’re really going to enjoy”.
Or ideally try to make it a bit more fun; for example with Graham Stephan’s channel it
became a bit of a meme with him asking people to destroy the like button.

I’m personally a big fan of asking for them to hit the notification bell instead, as then
they get notified of all future uploads (subscribing alone won’t do this), so it
significantly increases the chance they’ll watch my future videos.

Note: it’s better to ask for engagement (e.g. likes / turning on notification bell) after a
good point in the video when they’re actually engaged and enjoying the content, don’t
just ask early on before the video has given them any value.

● Until you reach 1000 subscribers, you could reply to every comment. Not only does
this help build an engaged audience (as people much prefer commenting if they
know the creator is reading them and may reply) but it literally doubles the number of
comments on your videos, thus making things seem more active and raising
engagement. For example if you get 10 comments on your video, just by replying to
each, you now have 20 comments on your video. And if you respond with a question
(e.g. “Thanks for watching - what do you want to see next?” it’ll boost engagement
even further if they respond. It’s a simple way even a small channel can get more
engagement.

Even as your channel grows it doesn’t hurt to still leave ‘hearts’ on some comments
too to help increase engagement.
🔎 Search Engine Optimization
https://magnates.media/youtube

Throughout this course I’ve mentioned that many creators worry too much about SEO, like
keywords and tags - probably because a lot of free youtube advice talks about this. But in
reality these have very little impact. I spend virtually no time thinking about these things any
more, and yet get millions of views on videos. I assure you that going viral and getting lots of
views comes mostly from the other things we’ve talked about in this training and course -
focusing on getting people to click and watch your content. Spending longer on a better
thumbnail and title, or making a better video, would be a much better use of your time.

And to be clear, this doesn’t mean you won’t still show up in search. YouTube knows what
your content is about, so even if you don’t use a specific keyword phrase, if your video is a
good match for a search term, it’ll still show up regardless of the tags. If you search for the
subjects of any of my videos - for example “Rockefeller”, my video on Rockefeller will be one
of (if not the) top search results. And yet I spent no time on SEO research. YouTube just
knows what the video is about, and shows it because the video has high expected watch
time per impression (thanks to a pretty good audience retention rate).

But that’s not to say SEO can’t have its place - for creators looking to target a specific
audience, you may try to make content around specific phases, and using Google Trends
and Google Keyword Planner you can get an estimate of what the search volume is for
certain keywords or keyword phrases. You can also use YouTube’s search bar to get an idea
what people are searching for already - as if you start typing something in the search box it
will autocomplete with popular searches. There’s also tools like VidIQ + Tubebuddy to check
the competition and get an estimate of search volume.

The simplest tactic if you do want to try and get your videos to show up for a specific phrase
(e..g “how to lose weight fast” if you’re making weight loss videos) is repeat your main
keyword phrase early in the description. You want to write the description so it still makes
sense (as you’re not meant to just ‘keyword stuff’ the description), but you can write natural
sentences that incorporate the keyword you want to target and variations of it.

What I do for my videos is simply copy some of the script I’ve already written into the
description (I normally use my intro as this generally describes the content of the video). I
then may add a few extra lines with other relevant terms or questions that the video answers
and that people may search for. But I spend literally a couple of minutes on this at most, and
I don’t do any research - I just put whatever I think describes the video to help YouTube
understand it.

That’s because what you say in the video matters a lot more. People don’t realise that
YouTube puts more emphasis on the words you actually say in the video, as this allows them
to literally tell what you’re talking about, as opposed to tags or keywords in the description
which are purely on human input and so can be manipulated. So if you can just naturally
include your main keyword phrase near the start of your video, and ideally a couple of other
times throughout the video, that can all help you rank higher in search.
📈 To Get More Views & Subs,
https://magnates.media/youtube

Return To Fundamentals
If all of this sounds a little overwhelming just remember if you can get more people to click
your video, watch your video, and come back for more of your videos - YouTube is going to
want to recommend your content.

It really can’t be overstated how important it is to just focus on those things; spend more time
on making titles and thumbnails that people will want to click on, and spend more time
making great videos that keep people hooked all the way through.

If you keep your viewers satisfied, your interests are aligned with YouTube.

My focus - and same for most big creators I speak with - has shifted from worrying about
algorithms and seo and promotion to purely worrying about making better content. (The 10
items on the content scorecard).

Make content so good you don’t have to worry about the algorithm.