by Doc – Owner, Founder, Surprised This List is So Long
This post was written for Amiibo Guru. You can find his channel here.
Pick your topic and be very specific to it – personal gaming channels don’t succeed anymore. At least, not without 5-10 years of full-time work behind them.
Your name is very important – it should immediately identify what your channel is. “Gentle Whispering ASMR” “Amiibo Doctor” “VidIQ” are all great names that grab the eyes of people who might be in their target market, because you know exactly what it’s about within seconds.
You are limited to the size of your topic. If you outgrow your topic, you’d better be ready to switch gears to something larger without losing your fanbase. This is very difficult and most people don’t pull it off successfully. Shesez did it well with Boundary Break transitioning to Region Break, despite Youtube algorithms taking a major dump on his channel.
Profile pictures and quality banner images are very important, and you’ve got to have a profile picture that stands out against Youtube’s white or black backgrounds.
Decide immediately what your channel is – are you a Shorts channel, a short-form channel, a mid-length channel, a long-form channel, or a livestream channel? My mistake on Amiibo Doctor was trying to provide everything because I was too creative to only do one thing.
Youtube’s front page isn’t affected by subscriberships, really. Youtube’s front page can be divided into a few categories:
- People you normally watch when they upload (which is where your subscriptions often show up)
- Videos that are gaining a lot of steam across youtube
- Experimental suggestions by youtube to see if you’d be interested in a channel
Subscribers don’t matter because of this reason: I’m not subscribed to any ASMR channels, but the 1-2 that I normally watch always show up on my front page anyway. I don’t subscribe because then Youtube would relegate them to the Subscriptions feed, which most people don’t check. Instead of gaining subscribers, you want repeat viewers – which aren’t the same thing.
If you can get someone to watch 3 of your videos in a row Youtube will forever recommend your channel to them. Youtube wants people to binge so they get more ad revenue.
The benefit of subscribers is gaining access to things like community posts, Youtube Stories, and monetization. And getting bitches.
Youtube Shorts and Algorithm
Short videos and very long videos are favored by the algorithm and by viewers. If you can combine a drawing factor and a short or long video together, you can have a real banger. See: my most popular video, where I uploaded modded footage of Pyra and Mythra in Smash Ultimate’s roster screen the day after they were announced. They weren’t in the game – the drawing reason – and the clip was 17 seconds long – the short video – so it took off and got like 165,000 views.
Youtube shorts are a different type of content. They’re a great way to gain subscribers, and a better way to lose subscribers – if you upload shorts and gain subscribers, those people will unsubscribe when they see your longer content. Do not rely on youtube shorts and think that they’ll catapult your channel to growth, because they’re infinitely competitive and getting harder to crack by the day.
However, Shorts are useful for getting people to watch your long-form videos when they go back to Youtube’s main screen. If you watch a short from someone, it’s more likely that one of their other non-short videos will show up the next time you open youtube.
The most important thing for Youtube shorts is by far, watch time. I have a Byleth amiibo short that periodically picks up a couple hundred hits from shorts because it’s at a 120% watch time, and I always gain some traction when that happens.
Youtube typically looks for new shorts shelf videos during low publishing times – if you release a video at noon that doesn’t do very well but has 120% retention, Youtube will often pick it up at like 8PM, when nobody’s publishing videos, because it doesn’t currently have enough new Youtube shorts content to supply everybody at home who’s binge-watching Youtube. I almost never have something hit the shorts shelf during the day, it only does so at night or in the early morning because I don’t have to compete with big creators releasing their shorts.
I’ve found the most success hitting the shorts shelf with 5-8 second videos. You’ll see on my channel that most of my shorts videos are battle clips that are that long, and that I edited in some sort of clicky title onto the video itself. Text keeps people reading, after all.
Know what the people in your topic are going to be searching for in five years. In five years’ time, when someone searches for how to train a mega man amiibo, they’re going to find my videos on it. That’s called evergreen content. Sometimes it’s the best source of growth that you have. Evergreen content is always useful and always interesting – that’s why podcasters do so well on youtube.
Youtube doesn’t give a shit about your channel. Its goal is to find the videos that are the most likely to get an individual user to watch as much as possible. It’s user-centric: if they know a user likes amiibo, they’ll recommend more amiibo videos. If they know a user likes guitar tutorials, they’ll recommend guitar tutorials.
Youtube measures what’s a good video and what’s not a good video by its watch time and click-through rate. Basically, how long do people watch this video relative to its length, and how often do people click on it relative to how often is it recommended. So if Youtube’s been spamming a video on your main page five times now, and you only click it once, that’s a 20% click-through rate. On a broader scale, if Youtube shows it to 1000 people and only 50 of them click it, that’s a 5% click-through rate. Having a better click-through rate gets you more views (which is achieved with better titling and thumbnail), having a better watch time gets you recommended more.
Anything can hit the shelf, if it has a high enough retention time.
One of the reasons that my videos get so much retention on youtube shorts is that it takes people about 2 seconds to realize a video has ended before they swipe – so if your video is 8 seconds long, 80% of people watch to the end but those 80% of people watch the additional 25%, you end up with like 105%-110% retention time, which is the retention that Shorts requires to be on the shelf.
If you want to rank in search, which you should, then you’ve got to learn everything that Youtube looks for. It’s called youtube SEO. Descriptions are the biggest part of youtube SEO at the moment.
Tags are basically irrelevant to how a video ranks, honestly. It’s kind of a shame. However, if you take the most commonly search terms in your topic and put them in the title and description, and make sure to say them in your video, you’re much more likely to rank.
Youtube’s search is tied in with google search, so if you rank in Youtube search you may just rank in google search. That’s a pretty good combo to have.
It’s better to have one good, solid video with great watch time and click-through rate per week than to have one every day with bad watch time and click-through rate. The absolute worst thing you can do on Youtube is to release two videos in the same day. Neither of them will take off, and they’ll both suffer tremendously for it.
Don’t release videos in the morning, shoot for about 2-4 PM CST, because that’s when most of the United States is either off of school or about to be. Upload the video at least a half hour before it goes live, because Youtube sometimes takes over an hour to process the 720p version of a video.
If you can find a low-competition search term, make a video on it. It’s better for a channel with 0 subscribers to be focusing on low-competition search terms and getting traction, than high-competition search terms and not getting any. However, you have to say the words of the search terms in your actual video, because Youtube checks transcriptions for relevancy.
Production quality is 100% required. If you’re not willing to put in the resources and time needed for production quality, you’re not cut out for Youtube.
If possible, have a facecam in your video, if it’s that kind of video. People need to see a face periodically or they lose interest. Doesn’t have to be yours, but it has to be someone’s.
Your voice doesn’t sound as bad as you think, and you’re not as ugly as you give yourself credit for. Get a better mic and a real camera, and make sure lighting is evenly applied. You’ll have to shell out a bit of money, my current mic and camera would run you about $70, plus the $200 cost of a capture card.
I use OBS Studio for recording videos, but not for streaming. It’s bad for streaming. OBS Streamlabs is best for streaming.
Your lighting is very important, and not hard to do. I bought a $10 light from Five Below that takes care of my bad lighting. Don’t be afraid to mess with built-in filters to correct for things, either.
Don’t make anything below 720p, and preferably export it in 60fps. Even if you recorded it in 30fps like I did, export the edited version in 60 fps. Youtube likes 720p 60 fps videos.
Stream VODs rarely work unless it’s game footage, but even then, they don’t work much. What Twitch wants and what Youtube wants are two different things.
Thumbnails and Editing
Thumbnails and editing are incredibly important. I slacked off on both until about last September, and basically didn’t grow because of it. Not making a custom thumbnail and not editing your videos is not an option. You can’t just upload 100 videos of game footage without commentary, without reaction, without anything interesting and expect to grow. And I know, because that’s basically what I did for the first 6 months. And no, the screen at the start of each Smash Ultimate battle that shows the characters isn’t a thumbnail. You have to actually make it.
When doing thumbnails, have no more than 3-4 words of text. If you scroll through my channel, you’ll see a lot of thumbnails that follow a pattern – that’s good. You want your thumbnails to be recognizable, so have the same theme across all of them. I experimented with the theme a bit, and you can see – in some, I have a Smash themed background with glowing gold text. In more recent ones, I have a gradient with text and images. In some of the really old ones, I have the “amiibo doctor dorito”, a red triangle in the corner. Red, light blue, and light green are your easiest “pop” colors. However, a gradient of nearly any light color will do fine, besides yellow and black-to-white.
I use pixlr.net for thumbnails – it’s free, and it’s basically photoshop.
Gradients are ridiculously good for thumbnails, when used tastefully and as background. You’ll notice that I’ve changed the thumbnails of most of my good videos to have something that pops more and uses a gradient. I try to avoid gradients that go more than three colors on the color wheel (like red-orange-yellow), and I make sure my text still pops against the gradient. That’s why I started using Impact font, and black outline.
Editing is tough and takes time. I’m cheap and have old hardware so I need something that’s light and easy to use. I go with Animotica, it’s $16 on the Windows Store and supports Youtube shorts formatting as well. It exports videos really fast and is far easier to use than something like Hitfilm Express or whatever Adobe products are out there at the moment.
When you’re editing, have things edited in so that people keep watching. Rule #1 of Youtube is that people are idiots, and you have to flash things in front of them to keep their attention for longer than about 3 seconds, especially at the beginning. That’s why you’ll see in my videos, I go for a piece of text or an image or a section or something to keep attention periodically. This is especially true in youtube shorts, where I keep text or something happening every 3-5 seconds.
Dividing videos up into sections is a great way to increase watch time. Youtube likes it when you have timestamps in your description, and so do viewers. Numbering things is always a bonus for retention.
Make a real channel icon and cover photo for your homepage that identifies at a glance what you do. My current one is straightforward: I have the iconic amiibo dots from the amiibo logo, and Dr. Mario’s stock icon. And the background is black, so it sticks out against Youtube’s white color scheme.
Make a real channel trailer, no longer than a minute that welcomes the viewer and tells them what you’re about. See mine for reference. Hell, copy it if you want. That should be the video viewers see when they first show up to your channel.
Playlists are great for getting people to watch 3 videos in a row. I have a playlist that’s labeled “Start Here, New Subscribers!” which is designed to A. educate people on amiibo and B. get them interested so they’ll watch the 3 videos and then I’ll show up forever.
Youtube doesn’t like livestreaming when you’re live, but it really likes livestreams after they’ve wrapped up and sometimes favors them in the algorithm. Livestreaming is a decent way to gain subscribers if your stream is 100% on-topic and interactive (such as when I do amiibo Q&A, or amiibo arenas).
Do not have uninteractive streams. People need to be able to see you, hear you, talk with you, and do something with you. You’re basically pretending you’re in the room with someone, like Dora the Explorer. If you just plug in a capture card and do something monotonous like amiibo training, you’ll pull maybe 2-3 viewers a stream. And you’ll bore the shit out of me, which is even worse.
Livestreams are easy watch time if you can pull a crowd. A 1-hour stream with average 20 viewers is 20 hours of watch time, and you need 4,000 hours in a year to get monetized.
Yes, I know Youtube livestream sucks.
Although the facecam on your livestream won’t be very large compared to the gameplay, you should be as interesting on facecam as the gameplay is. Move a bit. Smile. Look at the camera to simulate eye contact. You can lower the quality to 480p to save your hardware a bit of work, unless your camera is the majority of the screen.
I use Streamlabs OBS for streaming. It’s easier on the CPU and that’s a big deal for me. It has all the same options and overlays that Studio has.
Read and answer questions from chat as if they’re in the room with you. Remember who’s commenting and who the usernames are.
Livestreams need to have thumbnails and good titling too. Plan your livestreams out the same way you would a video – titling, description relevant to tags, etc.
For whatever reason, I get the best livestream traction when I’m just talking to the camera and not playing anything at all. I don’t know why that is.
Don’t copy other people’s ideas unless it’s infrequent and cleverly done. You saw how when the amiibo Raid Boss thing happened, and Alpharad did his Hard DK, suddenly all the other Smashtubers had their own strongest amiibo. That’s because they’re unoriginal and uncreative. Don’t be unoriginal and uncreative, or at least don’t do it often, and when you do it, do it well. You saw how I got onto the Raid Boss trend with my videos of how to train them, and how to beat them. I didn’t put out videos saying I had the strongest Raid Boss whatevers – because that’s unoriginal and uncreative, although it would’ve been easy growth.
There is something to be said about copying someone else’s idea – from a business perspective I don’t blame Smashtubers for having done it, because it was a trend. Hopping on trends is sometimes very helpful. However, when the trend dies, so too does the subscriber base. Just ask Captain Kidd.
People want things explained. How To videos are evergreens as long as people need to learn something.
Entertainment videos are much harder to build a channel off of than How To. With entertainment, you’re competing with half of Youtube, every streaming service in the world, and anyone else the viewer can text. With How To, your information is much more exclusive, but also significantly more niche. However, there’s an in-between. People seem to be entertained by facts, like channels such as DidYouKnowGaming and Boundary Break. I’ve tried to become more of that channel with the Amiibo University series.
Nobody comes to you for your “interesting personality”. You’re probably not as interesting as you consider yourself, and even if you are, that’s not the hook you want. Holding up a sign that says “I’m interesting please spend 20 minutes on this!” is a bad way to get people to watch. Instead, have reasons for people to watch outside of your personality, and then if they come back repeatedly they’ll start to watch for personality. In my case, that reason is amiibo.
If your niche is taken by omnipresent already-established fishes as was the case with amiibo, your content won’t do very well in comparison. If you can, collaborate with someone larger than you so that your channels come up in each other’s suggested videos. However, remember that most people on Youtube are opportunistic and don’t give a shit about you, so you’re unlikely to get something unless you can give something of roughly equal or greater value.
When collaborating, do it because you want to, not because you have something to gain from it. Don’t use people like that.
Clickbaity content doesn’t work that well, unless you deliver and if it’s something that hasn’t been clickbaited before. Remember the video I did where I trained a Ken amiibo wearing boxing gloves? No, you don’t, because it was a bit too clickbaity to be of interest, and there’s nothing thrilling about training amiibo with gloves.
Sure, you can title all your videos in caps lock, but you shouldn’t. I’ll judge you for it. And it’s repetitive.
Memes are good video growth, but people come back expecting you to produce the same kind of meme. If you don’t, they unsubscribe.
For some reason, numbers always get a lot more clicks. I’m not entirely sure why that’s the case, but lists of things help both retention and viewership.
When a breakout video does tremendously well, and I mean tremendously, follow up on it a few days later. Literally rearrange your schedule and drop everything to follow up on it. She’s not that important.
You need 1,000 subscribers and 4,000 hours of watch time to make money. Don’t pay for either of these things – if you only have 10 real subscribers and 4,000 hours of watch time, paying for another thousand is going to cost more than you’ll make.
It takes 150,000 views to make five bucks from a 15-second video, but it takes 5,000 views to make five bucks from a 3-minute video. Going for longer-form attention spans makes more money. However, longer spans are harder to obtain.
You can’t make a living on Youtube alone, and probably can’t do it with sponsors either unless you’re pulling a couple hundred thousand hits per video regularly. This isn’t a career option unless you’re already showing promising numbers. For reference, I make about a dollar per thousand views per day. After three years.
Your December CPM will be twice of your January CPM, usually. Don’t get monetized in December and think that you’re going to make a lot of money – come December 26, you lose all that CPM.