TubeTalk: Your YouTube How-To Guide

YouTube Algorithm Myths Are BUSTED with Mark Robertson

June 04, 2020 Liron Segev; Mark Robertson Season 3 Episode 215
TubeTalk: Your YouTube How-To Guide
YouTube Algorithm Myths Are BUSTED with Mark Robertson
Chapters
TubeTalk: Your YouTube How-To Guide
YouTube Algorithm Myths Are BUSTED with Mark Robertson
Jun 04, 2020 Season 3 Episode 215
Liron Segev; Mark Robertson

There is so much information about YouTube and its algorithms. But what info is correct and what is false?
I speak with Mark Robertson, one of the very first SEO industry leaders to go all-in on digital video and the first to coin the phrase “social video marketing”.
With over 2 decades of experience in digital media, Mark is a pioneer, and a digital marketing expert who really understands that ins and out of the YouTube system.

On today's episode you will learn:

  1. Does YouTube prefer long videos?
  2. Does changing the video metadata after a video is published good or bad?
  3. Do Descriptions, Tags, Titles matter, or is YouTube AI smart enough?
  4. Do Likes, Dislikes, and Comment help or hinder your video?
  5. Should you or should you not link to external websites off YouTube?

If you want to finally hear what is FACT and what is FICTION when it comes to the YouTube algorithm and how it impacts your video performance, THIS is the episode for you!

Show Notes Transcript

There is so much information about YouTube and its algorithms. But what info is correct and what is false?
I speak with Mark Robertson, one of the very first SEO industry leaders to go all-in on digital video and the first to coin the phrase “social video marketing”.
With over 2 decades of experience in digital media, Mark is a pioneer, and a digital marketing expert who really understands that ins and out of the YouTube system.

On today's episode you will learn:

  1. Does YouTube prefer long videos?
  2. Does changing the video metadata after a video is published good or bad?
  3. Do Descriptions, Tags, Titles matter, or is YouTube AI smart enough?
  4. Do Likes, Dislikes, and Comment help or hinder your video?
  5. Should you or should you not link to external websites off YouTube?

If you want to finally hear what is FACT and what is FICTION when it comes to the YouTube algorithm and how it impacts your video performance, THIS is the episode for you!

Liron Segev:

In order to succeed on YouTube, you must create videos that are over 10 minutes. In order to succeed on YouTube, you must never touch your thumbnails, your titles, your description. YouTube is going to re-rank your videos and that could tank horribly. You should never use external links because that's going to make YouTube hate your videos and therefore not promote them. Are these true? Are these myth? Well today, we're going to unpack all of this. Let's do this.

And welcome to another episode of TubeTalk. My name is Liron Segev, I'm a tech blogger, a YouTuber and the director of customer success here at VidIQ, where everyday we help creators big and small level up their channels, get more subscribers, more views in less time. So one of the things that we specialize at VidIQ is understanding YouTube algorithm. That's why it drives me absolutely insane when I see online social media, there's so much misconception that's out there about the algorithm. So today we have an expert in. Today we're going to bust some of those YouTube myth.

Liron Segev:

I've got Mark Robertson in the house. He's one of the very first SEO industry leaders to go all in on digital video. The first to coin the phrase, "Social video marketing," two decades of experience in the digital space. So of course we had to make sure he was part of VidIQ family, which is exactly where he is now. He now heads our Product Marketing Leader at VidIQ, is a digital marketing expert, speaks around the world at the various events and he is going to bust at a lot of this myth for us. Mark, welcome to TubeTalk.

Mark Robertson:

Stoked to be a part of it. And those two decades are because I started at age 10. Don't want to give away my age, there. No, not exactly.

Liron Segev:

But you've been around for video for a very long time. You have seen video, you have done video. Give us the Tweet. Who is Mark in a tweet? The parts that I didn't cover.

Mark Robertson:

I hope they did expand the character limit, there. I'll try and as brief as I can, but I'm left brained and right brained. So I'm a bit creative, but I have to analyze a lot of data. I come from a family. Actually my grandfather invented color film at Kodak. My father was a cinematographer. My father filmed plenty of movies. Did Michael Jackson Thriller. He did his thesis actually with George Lucas at USC. And I went into music. And about 10 years into that, I got a job as a marketing person, got into SEO. And in 2006 started seeing these videos showing up in Google and thought, "Wait a minute, how do you do this?"

Mark Robertson:

And that's when I created a site called Real SEO at the time. It was really a source for me to learn myself, a bit about the optimization within Google video and quickly learned about that. But also really learned that to focus in on video SEO, and I'm using air quotes, we'll get to that in a second, that you really have to understand video. How to produce, how to do storytelling. It's a lot more than what I was used to as an SEO person. So anyway, I'm excited to be a part of VidIQ. I'm excited to be on this podcast and we'll do our best to, I guess I'll say dispel myths. But I think, in a lot of cases, information stays online for a long time and yet algorithms change constantly. And so, sometimes some of these things we may talk about may not have been myths at one point but they probably don't need as much focus.

Liron Segev:

And I love that. The idea of that algorithms changed. And we use algorithms in the plural because [crosstalk 00:03:55] made up of multiple algorithms. So, let's start off right at the bat. And this is the big one. We hear this at events around the world. We hear this every time we do live training, every time we do audits with individuals, does YouTube prefer long videos?

Mark Robertson:

Oh, that's a tough one. So yeah, what I've heard is, and this is based off some data where the methodology could be questioned, that in fact, not only do they prefer longer videos, but that you should be uploading videos that are longer than 10 minutes in length. And I think when YouTube launched watch time, there was a lot of confusion around what watch time is. And just to go back in time a little bit, I believe it was 2013, when previous to that videos were ranked in YouTube, according to how many views they got. And so, they switched to something called watch time, but they made it very clear in all their training around watch time that it wasn't about longer videos. It was about encouraging viewers for creators to create content that that has viewers staying on the YouTube platform longer.

Mark Robertson:

And so, here's a direct quote from YouTube. It says, "We want viewers to discover the videos that are most satisfying to them, whether they're a few seconds long or a few hours." And as recently, I believe it was about, may even be two years ago now, but I want to say a year ago, YouTube mentioned when they were interviewed that they're now focusing on what they're calling quality watch time. And there hasn't been a lot of communication that I'm aware of from YouTube around what that means. But I think we could bust the myth that they prefer longer videos. And I think we could bust what I'll call the myth that you need to be uploading videos longer than 10 minutes. I'll try not to spend too much time talking through this, but I think what it really boils down to is the viewer intent and whether or not your video satisfies that intent.

Mark Robertson:

I saw this with Google, with real SEO actually, years ago. When I started in 2016, you could upload a video. You could say, "Hey, this is a Colorado Springs Ford dealer F-150 and it would show up in Google. What they started to realize over time is that when people are searching things like Colorado Ford, they want to know a website for the local Ford dealer, phone number, maps, perhaps inventory, but they don't really want to see videos. And so, they've really moved more towards intent. Producing results in Google that align with the viewer's intent when they do a search.

Mark Robertson:

Now, Google isn't YouTube. It's definitely different, but it's owned by the same company. And the theory of satisfying the user intent rings for both companies. It rings true in most discovery algorithms. So again, if you've got a video, that's, let's say it's a personality driven documentary. More than likely, that should be longer than 10 minutes assuming you're doing a good story, it's well-produced, et cetera. If you're doing a video about how to tie your shoes in five seconds, then doing a video that's 10 minutes long is overkill. So, and you and I were talking about this earlier. I don't know if you want to [crosstalk 00:07:28] some specific advice here, but there's things to look at in YouTube analytics to guide you.

Liron Segev:

So I think that's important. The important word here is intent. It's very much like in the old SEO blogging world. It's, "What's the intent of the reader?" Well, very much, let us carry it forward to the intent of the viewer on YouTube. Now, if someone has a toilet that's clogged. I don't know why, but I always keep going back to the clogged toilet example, but that's what it is. But if someone has a clogged toilet, they don't want a 25 minute vlog of what you ate for breakfast and where did you shop? They just want a solution.

Liron Segev:

They have a problem. They need a fix. The intent is to get in and get out. So to Mark's point, you can say it in four minutes, don't make it 24 minutes, say it, give the user satisfaction, let them love what they watching and they'll be sure to come back again and again. And this was subjects that we were referring to, it's, look in your analytics. Now, if you don't have the VidIQ tool, that's cool. Go look into the YouTube analytics. Look at the individual video, look at something called retention. And then Mark, within retention, what does that mean? What should we be going for?

Mark Robertson:

Sure. And again, I think we should be focused on user intent and consider the fact that there's various niches across YouTube. Different types of content are going to have different answers, here. But if you go into YouTube analytics, there's a audience retention reports and there's a relative audience retention, and you can do this, I believe on a per video, perhaps you can group it as well. But if you go into a specific video in creator studio, click on the analytics, look for something called audience retention reports, and it'll show you, for a given video for example, let's say it's 10 minutes long, the average amount of time that people are watching it and it also show the average percentage watched.

Mark Robertson:

And if you've got a six minute video and you can see that the average retention is 75%, that could be an indication that you could have produced a longer video. If you see something, maybe you've put up a 10 minute video and everybody's abandoning it at 10%, well, then you've probably created a video that's either not getting your story across, or it's just not resonating with that audience and it's too long.

Liron Segev:

So I think it's an important metrics, which a lot of people don't really spend time in. I love the audience retention reports because if I go at a video and I see that I'm getting 40% overall retention, so I'm aiming for 50%, but if I'm getting 40%, I'm getting close. Well, I want to dissect that report to see where people are dropping off. If they're dropping off at the beginning, well, my hook wasn't strong enough. I didn't grab their attention. If they're dropping off in the middle, well, maybe my content isn't great, that wasn't getting their attention. If I dropped off at the end, well, means I've done my job to get them all the way to the end. Maybe I need to do a better job on the outro. So, I would encourage everyone to do spend some time in that retention report and really dissect each individual video.

Liron Segev:

And as Mark says, if you're getting very, very high retention rate, people are loving your content, that's a great indication that people want more and more and more. Maybe your next video needs to be an extra couple of minutes long. Go look at your retention audience. Is it still 75%? Amazing. Maybe make your next video even longer. Until you reach that 10 minute mark. And for those who don't know, the reason that 10 minute mark seems to be the golden standard and that's maybe why some of these misconceptions are coming out, any video over 10 minutes allows the creator, allows you, to place ads in the middle of the video, not just at the beginning and at the end.

Liron Segev:

So people love it as creators because it allows us to run more ads and therefore earn more revenue. The flip side of that, if you've run a 15 minute video, that's filled with lots and lots of ads, what are you going to do to your audience? So you always got to find that balance between you, and what's good for you, and then what's good for your audience, because your audience is the one that's going to keep coming back again and again and again.

Mark Robertson:

100% and the entire goal here is YouTube wants to satisfy the people, the viewers that is going onto YouTube. And let's say they do watch your two minute How to Tie a Shoe or your three minute How to Unplug Your Toilet, more than likely they're going to spend more time on YouTube as a result.

Liron Segev:

So one of the things that we spoke about just before going on air here, is a case that proves once and for all that 10 minute videos are not a must. We spoke about the case of Dad, How Do I? That's a new YouTube channel. It's been spoken about at length. And if you look at that channel, it's got something like 2 million plus subscribers and each video is about four or five minutes. So, do not have to aim for 10 minutes. You want to go for quality. That is what you go for, not quantity and certainly not length. You've got to remember the viewer.

Liron Segev:

I think we've covered that myth about you absolutely must make longer video. That is truly busted. What you must make is videos that people want to watch. Which leads me on to point number two, or myth number two, which is, "Look, I've made this video. I uploaded it. It's not doing so great. Well, I'm definitely not going to change the metadata. I'm not going to change the title. I mustn't change the thumbnail. I mustn't change the description or the tags because YouTube hates that. I'm going to step away from that." True or false?

Mark Robertson:

False. In fact, it was said to me at one point that if your video is showing up, say number one in the search result for a given query, and we won't go deep into that because in many cases, those are personalized search results, that you don't want to edit that title or description or any of the metadata, because it causes a re-index by YouTube and you won't be showing up as high as a result.

Mark Robertson:

And I did do a little research here and pulled out a quote directly from YouTube. It says, "It depends on whether the updates make them better for the viewer or worse or about the same. There's no inherent disadvantage of the system related to updating these information. So if you think you have an improvement, give it a try." I believe this was something that could possibly have been the case several years ago.

Liron Segev:

Exactly.

Mark Robertson:

Because there was a bit of evidence that showed that that could be the case. I've never once heard YouTube suggest that specifically. And, in fact, they say the opposite. So if you have a video, whether it's doing great or whether it's doing poor and you think it's more relevant or less misleading to update the title, do it. If you think you can add a thumbnail that's better. There are actually strategies and reasons why you may want to do that. Certainly don't avoid changing your video metadata because of this myth that it'll take some time for it to get re-indexed.

Liron Segev:

I'm a bit on the cautious side, so if a video is really, really performing, don't touch it. YouTube loves it. The audience loves it. You're not going to make it any better. Leave it alone. However, if there's a video that's not doing well, it's getting zero views now, what harm could you possibly do by not optimizing that video? Go for it. Stop being afraid of the YouTube. Stop being afraid of the algorithm. It wants you to succeed. The more you succeed, the more ads YouTube serves, the more money it makes. So I'm glad we busted that myth. It was a big one floating around for years, and that is absolutely false now.

Mark Robertson:

Yep, absolutely.

Liron Segev:

Now, speaking of metadata, our next big one is creeping its way up. I think it's becoming more and more popular. Look, the message is, YouTube's AI algorithms really understand my video. They know what it's about. They're doing captioning. In which case descriptions are not important. Optimizing my title, not important. YouTube just knows this stuff. I don't even need to bother with us. What do you think about that?

Mark Robertson:

Well. There is some truth to part of what you said. They are getting much, much better and probably in our lifetime, will be so good at it that they won't be looking at the metadata specifically with regard to understanding what our video is about. But, let's talk about descriptions specifically. And there's something I'm going to ask you to bring up in a second because you've been diving deep into it, that's related to descriptions and something they recently enabled or started suggesting.

Mark Robertson:

But when you think about the description. One, it's an area in which you can increase engagement with your users. You can provide them with additional videos related that they could watch. So there's ways in which the description can be used for engagement. There's ways in which it can be used for SEO purposes in YouTube, which we'll talk about in just a second. But the one thing that always irks me on this, and this is probably because of my traditional SEO, is that there's also Google.

Mark Robertson:

A lot of people forget about Google or Bing, and they definitely look at on page text. And so, when you think about the fact that Google has 15 times more search volume than YouTube, YouTube still has a lot of search, but Google has 15 times. And YouTube videos rank on 20% of all those searches. And 2% of all Google search clicks go to YouTube videos. They are looking at those descriptions to help understand the content and how to rank that in Google specifically. And in many cases, let's say you're doing a video about a recipe and you've put a recipe in that description and you've linked it to a blog post where that blog post has a well-written recipe in it, say it's well structured. Maybe it even has schema markup.

Mark Robertson:

Google will look at both of those, tie them together and you make it a featured placement in Google with a suggested clip based off that natural language processing. The most important thing with optimizing on YouTube is the content, is your storytelling, is are you retaining the viewer, et cetera, but metadata is still important and they wouldn't ask for it, if not. And it's certainly important for traditional search.

Liron Segev:

I'm of the opinion that we have to feed the beast. So, YouTube has given us places to allow us to give it information. Well, let's use it. Remember your first two lines or so on your description are critical because they get seen as a preview. If you want to capture people's attention, remember to focus on those first two lines. Attention, attention, attention is what it's all about.

Mark Robertson:

100%, actually, and I'm glad you brought up the human versus what some people would think of as SEO. By the way, SEO is really about following best practices and optimizing the content you're putting up on a given platform for discoverability. It's not about hacking, it's not about keyword stuffing. And so I'm glad you brought that up. You want to write this for the viewer and to your point, you do want those first two sentences to be a bit of a summary for when it shows up in various search placements across the web.

Mark Robertson:

But this shouldn't be something where you copy a list of keywords from someone else's video and stuff it in your description. In fact, I think on the description screen, you'll see a notice from YouTube saying, "Don't do that." So, it's not an area where you need to optimize it from the standpoint of hacking it and adding terms that aren't related and that sort of thing. That's certainly not going to help or work.

Liron Segev:

Correct.

Mark Robertson:

But it's a valuable area to help certain users who find it useful to engage better with you and to have a better experience.

Liron Segev:

So let's move on to the next one, which is again, we get told this often. Likes, dislikes and comments don't matter. So, who cares if people love my videos and they give them lots of thumbs up, or if they hate my videos and give it lots of thumbs down, or if they put anything in the comments? Who cares? YouTube doesn't look at this, so why should I even bother?

Mark Robertson:

So, this is an interesting question. And the myth is that they're ignored by the algorithm. Now, certainly likes, dislikes and comments are something you should be looking at because it is an indication, in many cases, of how your video is resonating and does the algorithm look at them? My suggestion to you in a second, I do have a quote from them as well, from their official blog.

Mark Robertson:

But, my is any and all feedback that a platform like YouTube can gather user feedback retention, whether they commented on your video, perhaps even the sentiment of the comment, your metadata. There are thousands of signals they could be looking at in order to improve what they serve to people. And in most cases, why would they ignore any of that? Now, there is the question of how much do they look at it? For example, are dislikes 80% of everything they look at? And I think that's not really the question we're addressing here. Obviously I don't work at YouTube and we don't know that.

Mark Robertson:

But to say that likes, dislikes and comments are ignored by the algorithm is a myth. And here's the statement from the official blog. "We've responded by updating our system to focus on viewer satisfaction instead of views, including measuring likes, dislikes, surveys and time well spent." So it's right there.

Liron Segev:

Mark, if you think about it, YouTube Rewind, one of the most disliked videos of all time, was still ranked number one for many months. So, if they didn't want to surface videos with dislikes, surely they wouldn't have surfaced that video at all, because it just goes against the thinking that dislikes impact the videos performance. So, I think we've got to look at this as again, viewers intent, because that's how YouTube is looking at this. Put yourself in the eyes of the viewer. They come to your video, they love what you have to say, they give you a thumbs up, they write a lovely comment.

Liron Segev:

Well, those are many, many, many signals that YouTube is looking for. A long word. Did they pause? Did they rewind? Did they bookmark it and watch later and come back to see it? All of those are signals saying, "Hey, this guy, this girl, loves your content." Therefore, when you put up a new video, guess who's going to get that served? That viewer. Because they've shown YouTube how much love they have for your content. So similarly on the dislikes, if they keep continuously dislike one of your videos, well, they're probably being very troll-like in which case, YouTube is not going to suggest your videos to that person.

Mark Robertson:

That's exactly right. I was just going to say. We just said earlier, there's a lot of different algorithms and it may not be that they look at likes and dislikes for search, but it may be that they've noticed a certain user every time he's served a certain video, or type, or channel, or a word is said, dislikes that video. And therefore let's say I'm that user. I may start to see less of that channel's videos and suggested as an example. These are just ways to think about it. Again, we don't work for YouTube.

Liron Segev:

Correct.

Mark Robertson:

But to say that they would discount user feedback signals entirely, doesn't sense.

Liron Segev:

All right, as we move on to the fifth one, and this is a very controversial one, I would like to say, it's something that's been debated at length. It's all about external linking is bad.

Mark Robertson:

And you're right, this is a complicated one because in many cases it's not a myth. So, let's say it's not a myth. Adding external links can actually hurt your videos performance, but not by default. It's about keeping them happy and engaged. I know I keep saying happy and engaged, but it's true. In fact, here's another quote. This is directly from YouTube. Does adding external links in your descriptions or [inaudible 00:24:24] your performance? And they say, "Not by default. Period. Yes. Keeping viewers happy and engaged," that's funny, I just said that. "If people are driven off YouTube, it may indicate the loss of the session. Most watchers don't actually click the links.

Mark Robertson:

And I think what they're referring to there is, let's say you've got a bunch of links in your description., A lot of viewers won't click those. But you do want to consider the fact that when they did outline the watch time algorithm, it's around overall session watch time, not video watch time on your own video. And so, if somebody's on YouTube and they're enjoying their experience and they watch your video and then they leave to go to your website, that interrupts that session watch time and that could be a negative signal. So, I think it's a myth to say that adding them anytime is always going to hurt performance. But I think there's some truth to the notion that YouTube wants people to stay on YouTube and have a good time on YouTube and not necessarily be taken away.

Mark Robertson:

So, there's lots of reasons. And I want to hear your thoughts on this, but there's a lot of value and a lot of reasoning as to why you would want to put a link somewhere in your video or somewhere in the description, either to provide more context or even as simply, as I said earlier, if you've got a recipe video and you want it to show up better in Google, maybe at the bottom of your description, do it. But I think at the end of the day, the most important advice I would say here is, save your calls to action until the end of your video. Utilize the tools within YouTube to add links and don't put them all over, littered. I'm curious what you'd say.

Liron Segev:

Yes. The reason it's complicated because the answer is, it depends, right?

Mark Robertson:

Exactly.

Liron Segev:

So, if you're using YouTube to be able to showcase a product that you've built and you have a store and you make your money not off YouTube ads, but off selling goods or selling a course or selling something, well, then you don't care what YouTube doesn't surface your video for longer. And I'm saying you don't care. It's not your main priority because what you want is you want eyeballs on your website. That's your main mission to get leads so that you can sell your product or service. In which case, stick links everywhere. Right? Because that's your mission. However, if your mission is to get more views and more subscribers, become a YouTuber, grow that community, well, then it will be silly for you to send people away from YouTube, which is what you're trying to build.

Liron Segev:

In which case, rather have links to other videos, have links to other content that is going to satisfy the viewer. And therefore, they're going to stick around for longer. We love building mailing lists and I think it's critical that we build mailing lists so we have ways to communicate with our audience.

Mark Robertson:

I was going to say, especially if you're not satisfying their intent [crosstalk 00:27:23] in the first place. If they are asking you, "How do I unplug a toilet?" And you show them how to unplug a toilet and then towards the end of your video, you say, "Hey, you want to learn more about me, or how I unplug toilets? Click here," because that helps you. I think that's fine. Now, I wouldn't literally your video, ever.

Liron Segev:

Exactly. [crosstalk 00:27:40].

Mark Robertson:

And I think you know that. That's what advertising is for. And YouTube advertising is incredibly effective and it's not that expensive.

Liron Segev:

Absolutely. I think it's about building relationship with your audience, but it's also about using the platform to build relations with other videos within your catalog. Remember YouTube's main priority is to first look at your own catalog. Do you have other videos that this viewer might like? That's the question the algorithm is asking itself. First, if you don't, then it will look elsewhere on YouTube. Now, YouTube rewards you, gives you credit. I don't know what the right term is, but, for keeping a viewer engaged and keeping a viewer satisfied, love thy viewer.

Liron Segev:

And because you are doing that and you're not sending them straightaway off the platform and they're able to watch more and more of your content, all these are more and more signals, more and more reasons for YouTube to surface more and more of your content. We have seen channels very, very big dirt, but you know what? We've also seen very, very small channels, really understand their audience intent and do this so effectively with playlists and additional videos and additional resources. What does that do? Earns trust. Earns trust, more viewers, YouTube surfaces your video even more. And then it's a rinse and repeat cycle.

Liron Segev:

So it does depend what your intent is. And as Mark says, just be specific with what you're doing. Don't litter your videos with just links. Because that just screams spam and none of us want to be surrounded by that anymore. All right. Well, we're definitely going to have you on, because these are only five myths that we've managed to speak to. There are many, many more, as you know. So Mark, thank you very much for spending some time with us. I know you're pretty busy here, but I appreciate you doing that. So thank you for hanging out here.

Mark Robertson:

Thanks for having me.

Liron Segev:

And for the rest of you guys still hanging out here and listening, all this goodness is of course available on the VidIQ website and share this episode with at least one other creator. Let's help each other. Let's understand how YouTube works. Let's dispel some of these myth. So we all grow and grow our channels together. And we'll catch you guys on the next episode of TubeTalk. Thanks for hanging out.