So What? YouTube SEO and Content Strategy

Well, hey there. Whoops. We're doing good on Thursday. Welcome to So What, the Marketing Analytics and Insights Live Show. I am Katie, the uh, you know, incompetent button pusher.

Uh I am joined by Chris and John. This week we are talking about YouTube SEO and content strategy. Uh you know, Chris and I have covered YouTube SEO before. However, it's actually a really interesting topic, and we wanted to recover it and refresh it because there's still at least, you know, Chris, correct me if I'm wrong, there's still this misunderstanding that you can SEO optimize your YouTube videos because YouTube is a search engine. It's a social media platform, but it's also a search engine.

It it is a search engine. Uh it's the second largest in the world, right behind Google itself. And uh people go to YouTube for lots and lots of things. I mean, anytime that you break something in your house, you go to YouTube about how I fix this thing. That's me.

Or if you get a new oven with an air fryer, you go you YouTube how to use it. Exactly. So there's a lot of of that in there. Now we covered the basics about a year ago. Uh and the basics haven't changed, right?

Uploading. Basically, if you if Google gives you a feature, you should probably use it, you know, when it you when you're loading up a video. Like recently, it yeah, they added if you have if you flag your video as an educational video, it'll ask you what's the topic of the video. Uh, what are the is it associated with a university course? Things like that.

So the the basics are are pretty much still the same. Uh what I thought we would cover on today's show is trying to get an understanding of what's working from a topic perspective, and then doing some forecasting basing on that, because what we know is that people use use YouTube differently than they use Google search. Um the habits are different, uh, the types of searches and time of year are different, uh, and what people like uh is different. So uh I thought that would be a good place to start. What do you think, Katie?

No, I think it makes sense because um, and maybe you'll get to this, but one of the things that we know is that if you're using Google Trends, for example, to try to figure out what people are searching for, Google gives you the option of looking at the Google search versus YouTube search because they know that it's two different systems, and they know that the behavior is different, so therefore you should treat them as differently as well. Using your SEO keywords that you use in Google search doesn't necessarily translate into YouTube. Exactly right. So the first thing we should probably start with then is our own videos and just get a sense of what's working, right? So we've been publishing, we've got a hundred and seventy, I think, different videos on our YouTube channel of you know, uh a variety of different topics and things like that.

And if you recall back to about, I think it was the fall of last year, uh, we looked at how do you just organize and understand uh your videos by topic. Well, one of the things that I I put together was I wanted to understand uh is there a relationship between the different topics in our YouTube videos and the outcomes we care about. So putting together a correlation matrix, we have all the different topics we've we've covered. Google Data Studio, Data itself. These are all keyword-based, right?

Influencers, Google Tag Managers, so on and so forth. And we have our major metrics like watch time, views, likes, and impressions. Views is one that I think is probably the one of the more helpful metrics to look at. But obviously trying to dig it out of a correlation matrix is a is a sure way to get an eye doctor's appointment. So what the easier way to look at that, it oh, I went to the wrong window.

Here we go. See, we're all having trouble with the buttons. Exactly. Is okay, obviously views correlates perfectly to itself. What are the other things that correlates to?

And this is just a standard spearmine regression. You can do this in any statistical tool. Obviously, strong relationship to watch time, to likes, to impressions, to subscribers games. But look at the we start getting into the topics. Google all things Google, Google Analytics, Analytics, the So What Video Series, all of these uh have done reasonably well in terms of getting uh views, right?

Uh there's a there's a a statistical correlation. Then we get down here into SEO, and now we're starting to get into the the statistically less valid section, right? And then you go all the way down here, and then you get towards the end, you know, the podcast is uh is negative. Uh podcasting itself is as a topics in negative, change management is negative. Um average viewed percentage, so that's interesting.

There's an inverse relationship between the number of views and how much of the video people watch, uh, which basically says shows people are dipping in to check it out and then bouncing right out, right? They're not sticking around. Uh and then the Trust Insights podcast has actually a very strong negative correlation, uh, which I found interesting and and and slightly concerning because it essentially is saying uh people don't watch. Well, it's a podcast, and so I you know, we record it so that we can put it up on YouTube, but as a podcast, if people weren't downloading it and listening to it as an audio file, then I would be concerned. I think YouTube is a nice to have for our podcast, but you know, I for the most part I don't sit and watch people record a podcast personally.

Yeah, that's interesting. You know, um, because that's you just put the whole episode up there, right? It would be interesting if you took one minute clips and tested those and see uh like because I think the length is what's killing you. I would imagine the average podcast is longer than a lot of that other stuff up there. Sounds like we need some shorts, John.

Damn it. I was waiting for the chance to get shorts in there. I'm all about the shorts. I feel like I missed something, but I'm also okay with that. Google shorts are the uh Google announced that this year part of their strategic drive is to get these YouTube shorts going.

And so they're trying to go toe-to-toe with TikTok and with uh Insta videos. So yeah, so it's all shorts all the time with uh with YouTube. I thought you literally meant like short pants. I know, I it's very confused. The jokes write themselves with this, it's unbelievable, you know, like you you're gonna spend all day in your shorts and you're gonna be working on shorts.

Well, it works with the RFP request for pants. This is it. All right, and you know, we're back. All right. So the next logical question is um how do we then decide what if we know broadly topics about stuff like analytics and things are working?

How do you go about getting uh a sense of what could actually work? Uh it's not going to be from YouTube itself, but one of the places you can go is your uh SEO tool of choice. And this is true for pretty much all these tools. Uh it's not uh anyone we're using the Atrefs uh tool, but again, it's Semrush supports us, Moz supports us. Uh and like you're saying, Katie, in the beginning uh of the show, you can choose which search engine you want to use.

So for keyword scoring. So we have Google, YouTube, Amazon. If you're doing Amazon shopping stuff, uh Bing, Yahoo, Yandex, Baidu, et cetera, so on and so forth. So let's just take a real quick look here. This is our keyword list, the trust insights keyword list.

Now we see uh on regular Google search, in terms of the things that people search for the most by volume, Google Analytics tops that list with 813,000 searches a month, followed by Google Trends, Search Console, SEO, so on and so forth, right? This is uh a pretty this is exp I what I expect to see. Um when I flip over to YouTube. I didn't know that this was something you could do with your SEO tool. This is this is a learning moment for me.

Um yes, it is. Um with YouTube, a the volumes are much lower now, 41,000 instead of 800,000. And look at like Google Analytics drops down to 6,000 searches a month, right? From 800,000 in regular Google search. So we have very different topics here of the keywords that we all like to think about uh and and talk about, you know, AI topping list, affiliate marketing, machine learning, SEO, data science, social media marketing.

So on a on a forecasting basis, or at least on a uh on I guess in a more of an explanatory uh analytics basis, we see that this is very different. This the things that people search for on YouTube definitely looks different than regular Google search. And some of the other topics that uh we talk about a lot. Let's go uh, like you're down here management consulting, only a thousand searches a month for that change management, a thousand searches a month for that. So it's not a surprise to me also uh when we think about the Trust Insights podcast, we've been covering a lot more of stuff like change management and things, which on YouTube doesn't do as well.

And it makes sense because some topics lend themselves better to written content, some topics lend themselves better to more visual and video content. So that doesn't surprise me at all. Um, you know, to have a talking head explain a change management process, probably not a super fun video to watch, versus to have you know, uh data visualization explained on a video, probably a little bit more interesting. Exactly. So we have now this quick look at our same keyword list.

So this is the same keyword list that we're using for our regular search up engine optimization efforts now, uh mixed up for YouTube. So the next logical step would be can we forecast out these terms and get a sense of when we should be creating content around these things? Uh, of course, the answer is yes, uh, because this would be a terrible show if the answer was no, I'm just guessing. That'd be a very short episode. It would be also it would be very short.

So here, this is the Google searches, and again, we can see the volumes that you know, week by week. Uh this Google Analytics uh peaked at the beginning of the year, and then we see spikes uh you know, May, June, July, September. So there's peaks towards pretty much the end of every month when people are trying to figure out how to use their reporting uh for regular Google search. But if we flip over into YouTube, different story, different look. Uh right now, so the general term marketing obviously uh has the highest search volume on YouTube uh for the for the month.

We see AI dropping down here in mid-March, uh coming back up in April, and then uh June. It's interesting. Uh, we may want to talk to our friends over at the Marketing AI Institute because their conference is in August, and August is when this thing is down kind of in the valley of despair. Uh, that they may want to move their promotional efforts uh at least for video back all the way back into the into June. You know, uh, because we had done an exercise like this last fall, when I'm planning out our editorial calendar, depending on the kind of content that we're creating, I use a different predictive calendar.

So I have the YouTube predictive calendar, which I use to plan out the episodes for the live stream, versus the regular Google search calendar that I use to plan out podcast episodes, blog posts, those kinds of things for this exact reason. If I was just using the Google search calendar to plan out all of our content, I'd be missing a lot of things. Exactly. So, in terms of like what we should be doing, um, we have SEO. Uh SEO is actually an interesting one.

Uh, it has a for YouTube, it has very strong seasonality, right? Sort of in the middle of the year, is when you see a nice big bump uh for that term specifically. So it's something to keep in mind. Uh we have affiliate marketing, which always seems to be showing up uh from time to time. Uh machine learning.

So let's go into machine learning here. Again, it's interesting that all the AI and machine learning stuff has this sort of mid-year peak and then just goes away. So uh to the extent that you know we should be starting to think about shows uh on YouTube in May-ish, uh, to try and catch that wave. Uh, we definitely should be. Uh let's see.

Other things that uh some stuff I would ex I would have expected to do better, stuff like social media, for example, kind of just meh. Uh well, I mean, it really what's missing from this key this keyword list is any sort of context around the kind of thing people are searching for, like how do I put together a social media calendar versus social media, which is you know, you could talk about uh a thousand different things about social media. So that's sort of what's missing here is you know, do I want to watch a demo of yet another social media platform, or am I looking for a specific thing? Exactly. And so that brings us to um your general content strategy around YouTube.

So one of the things that YouTube itself recommends is sort of this hero hub help model. So uh you create a lot of help content, and help content is uh content that you are making that just answers generally, you know, how to, how do I do these things? Um hub content, sort of like a bigger idea, sort of a monthly promotional idea, and then your your hero content is like once a quarter, a big thought piece, you know, uh sort of a why kind of uh a really broad appeal. And when we think about now these different uh topics, let's take artificial intelligence. Well, let's see if I can find my mouse first.

Um take artificial intelligence here. If I were creating uh help content, it'd probably be just a year-round thing. How do I, for example, vectorize text uh for natural language processing? How do I uh build a confusion matrix? How do I interpret the results of a confusion matrix?

Like, like you said, Katie, very tactical, uh straightforward stuff. The hub stuff, I would think we would be trying to aim for, you know, some of these minor peaks, uh, something uh something from when you see a little bit more strength uh from search to say, okay, we're gonna spend some money, we're gonna spend you know some ad dollars, maybe uh engage our design team to make something that doesn't look like I put it together in the middle of the night while you know, uh just half asleep. And then the big peaks are the ones when the your hero content should go out. So you I I would say if we were going to do like a piece on uh on hero content, we would do uh launch at mid-May for something on AI uh, because that's when we know for YouTube at least, that's when people are gonna really be paying attention to this topic. So it's not just a question of the hero hub help model arbitrarily, just throwing out content as you create it or as you have a budget for, or maybe just on a calendar, but trying to time it to when you know interest in that topic is going to peak.

Um, so we have a couple of questions, Chris, and John, I apologize. This means your face is gonna disappear for a second. Um there it is. Uh, so Jesse commented, I think it's both content, it's both about content fit, but also about audience. Um now, Chris, you've done some extensive research on audiences on these different social platforms.

Uh he says, I would expect YouTube is going to be younger, which means different interests, which is true, different interests on YouTube versus Google search. Something like change management is going to be searched by a manager, while artificial intelligence could be searched by students. What do you think about that angle, Chris? The audience on YouTube, uh, again, because if it it's massive footprint, is pretty much just the audience of the internet. Um, there I I there's not really a major demographic skew on that platform, as opposed to say a TikTok, where there is definitely an under 30 skew.

YouTube is pretty much used by everybody, not necessarily just the platform itself, but when you're browsing around from page to page, like if you're on the Trust Insights website and we've got videos embedded, you're using YouTube, and that that data shows up in YouTube. But the point about interests is definitely on target. Uh, you know, someone again looking for change management. Probably YouTube is not the first place you would go to think for for change management stuff, right? You it's you're probably thinking more a book about change management.

It's just not super uh at first blush, it's not the sort of thing that you would go to YouTube unless you were thinking about can someone explain how this works or what this thing is, sort of educational content, like from like my my uh one of my kids. Uh, whenever there's a topic in school that you know their their online course is doing a terrible job of explaining, send them to YouTube and say, go find another video that explains it in a way that you can understand it. Um, so there is definitely that intent uh angle that I think is 100% correct. Yeah, I mean, when I think about YouTube, like or search in general, if I'm searching what is change management, I'm probably starting with a Google search. And if there's a relevant YouTube video, that will come up in the search.

But again, I personally don't think that a talking head of someone explaining what change management is is gonna be a really compelling piece of video. I could be wrong, but I don't really think that it's gonna do that well. One of the things that you can also do in uh in your SEO tools, and again, this uh I'm gonna flip over to back to Google search here. So I've got my Google search for our keyword list. If we want to get a sense of maybe what some of the topics that we might want to create videos about because of what you just said, Katie, I'm gonna turn on cer go to search features and I'm gonna say, show me search terms where video is included in the search result.

Can you see video, Chris? The uh the screen. Oh, whoops. That would help, wouldn't it? Uh let's go back.

Makes for compelling video. It makes for compelling video when you can't see it. Okay, so I've gone back to my Google search. I've turned off the YouTube search here and gone back to regular Google search. Uh and now I'm gonna go to SERP features and I'll say show me only uh search terms where videos are included in the search result.

You'll notice it's not exact, it's not identical, but there is more of an overlap here with the YouTube list than pure Google search, right? So we do see you know, social media affiliate marketing, artificial intelligence and stuff, um, showing up here. In these are the terms where there is a video thumbnail or videos themselves in Google search results. So again, if you think that there's an opportunity um here, you might want to create videos for this. Now, if you want to get even fancier, uh, let's cap our keyword difficulty at 50.

So basically, we're saying I'm not going to try and compete with the big boys, right? Um, so these would be the terms where I could show up in search results with a video, but I'm not in a knife fight with McKinsey, for example. Uh so the management of change, 7400 search Google search results where a video is included as well. Prescriptive analytics, descriptor, TikTok analytics, which we'll have to save that for when we put our TikTok paper out in a couple weeks. Um Google Analytics for beginners, R for data science.

So again, we're just in a straight up uh SEO tool seeing okay, these are the kinds of things where we might stand a fighting chance of showing up for these particular topics if we can create YouTube content and optimize it. Yeah, same with CPC too. It's crazy. Management of change is not even a tenth of what you know paid for change management is exactly. Um Dagmar was asking uh where where do we find this particular calendar?

We make that. Um, so that is something that we have our own software for. Um, and it is a paid service, and John will be happy to take your number after the show. Um Jesse said, shorts are insane, definitely worth trying them out. I wear shorts all summer long.

I don't. We'll be here all week, ladies and gentlemen. Try the news. Oh, we are the worst. Uh, we really are okay.

So uh when we think about the Hero Hub Help model, we want to try and time the different components of the Hero Hub Help model to when we know search volumes are going uh going to be working in our favor because why why otherwise? You know, would you not do that? Right. Um the other thing that is really important um with YouTube, and it's not a search thing, but it is an effectiveness thing, is to the extent possible doing collaborations. YouTube is a platform that really does a great job in encouraging collaboration and helping you work with other groups um to sort of share audiences.

So if you were to go to YouTube, uh, let's go to just straight up regular YouTube here. Whoops, awesome YouTube. Um let's look for uh musical group first to 11. Uh you will notice that a fair number, and this is something that musicians do really well. A fair number of these uh are collaborations with people, and uh when you do your promotions with them, you can obviously tag people in uh and you know stuff your things full of links and stuff and and share audiences and YouTube really works well for cross promoting for helping you cross promote um on your videos with other audiences.

So if you've got something that you've identified another YouTube account that is at at parity with yours, right? So if you've got a thousand subscribers, they got a thousand subscribers. You can you can probably team up pretty easily. It's not gonna work out so well if you have a thousand subscribers, they've got a million. There really wouldn't be a reason for them to work with you.

But when it comes to collaboration, this is a way to build and grow your YouTube audience to get links to your YouTube account. So you can see here this uh Brooke Surgeoner is is linked up within this video. So remember that YouTube is run by Google, Google's a search engine, and it's predicated on the network graph. So if you can essentially get inbound links to your YouTube channel from other videos, uh it will benefit your channel. So all the same pillars of SEO that we talk about with Google search apply then to YouTube search.

Many of them do, many of them do. YouTube apparently, at least from what I can tell, has not yet inherited um the BERT language model for you know more in-depth queries. Uh that seems to be a Google search. YouTube's search engine still is appears to be very token heavy, meaning individual words and sort of biograms. The last um fully revealing paper we have about YouTube's algorithm is still from 2016, and it's individual words and stuff that are factored into into how uh it serves up requests.

And when you say words like that too, so that's then based on title field and meta descriptions, right? It's not on automated transcriptions or anything like that. It is so it does include the the automated the closed captions. It's one of the reasons why in our basics of YouTube video from a little while ago, we said every video that you load up, um, if there are terms and jargon that you use in your industry, you should be loading your own closed captions. Interesting.

Um one of the things that I've noticed uh and has actually gotten really good if you are in the video production space is Adobe Premiere now has built in uh transcriptions so that as you're editing your videos, it can create the transcript for you, publish the closed captions file, and you can also burn the captions into the video itself if you want to have um captions. So, like when we produce our our our shorts uh for TikTok and Instagram, uh I am burning those captions now straight into the video, and it's just it's built right into the tool. It's built into your shorts. I I I'm loving it because I thought I would be the one that would be abusing this to death. Umsight Chris is the one that's really can't stop.

Oh my goodness. Uh yes. Uh well, you know, I I feel like a lot of times I've seen a lot of videos, you know, uploaded to YouTube from other B2B companies, and the lack of any kind of transcription or just general description of the video seems to be a missed opportunity because that is a big part of how people are finding your videos, not just the title of the video. Exactly. The other thing is that YouTube's recommendation engine is really fast, it changes very, very quickly.

So if you watch just two or three videos from a creator, for a little while, you see a lot of video from that creator. Um, you you see it um and it gets shown pretty prominently. And so one of the things that I've been trying out recently, and I've seen some really good success with um is and this is actually a recommendation from folks uh who who subscribe to my newsletter is turning my newsletter into a video. Uh let's go ahead back here. Order entries do.

Oh, I make a lot of stuff. Let's see. Apparently, I need to fix the search on my website. I says moment of truth. Here we go.

Exactly. Every sales guy hated demo. No, it's totally fine. Um let whoop. Let me grab this week's newsletter.

Because what I've been doing is putting the newsletter as a YouTube video. Um so when you go to the page, you can see watch this newsletter on YouTube, right? And you pop on over. Now on the site itself, um, in previous issues, I've I'm playing on putting the video player straight in. What you'll notice is that this cranks a couple hundred views of of a video on my channel, you know, within 24 hours, right?

It helps get content seen on YouTube really quickly. Logically, if I can get you to watch one video a week, right? Um through the newsletter, then you're gonna see my other videos from that week. Since I'm publishing a video every single day, if I can have that one, it's not true here or Hub Help, but there is sort of that tent pole strategy. If I can get you to see at least one video week and get you keep coming back, then I'm gonna take advantage of YouTube's recommendation engine to get you to be able to see the other videos on my channel.

So this I found I've been doing this now for about a month, and it it does add time. It adds about an hour to the production time of each week's newsletter. But the results are really good because it's boosting my channel substantially compared to just publishing and doing the regular stuff of you know having it in the blog post and then having it on social media by having a lot of traffic get funneled to my YouTube channel. I I do see really good results from it. Is it literally just you reading your newsletter?

Yes, I read basically the cold open. Um and then I very quickly just narrate anyway. And in this week's section, we have you know the news, there's this, this, and this, you know, check out our trust insights uh uh courses and so on and so forth. So there's the it is mostly the long cold open. Ah, gotcha.

It's great. You can get more of the cool disdain that you really can't get from the newsletter, you know, from a dry read. Um what's also interesting is I've I've got an audio version too, because I just ripped this from um video to audio, and even that gets you know 80 or 90 uh listens a week because people want the different modality, but we'll we'll save newsletters for another show. Just in terms of how do you get more people to see your YouTube content um and to link to it and so on and so forth, putting it in your newsletters is is a pretty straightforward win. Yeah, it's just amazing because you've got that right side, you know, content there.

It's like that's just such a huge honeypot. That's fantastic. It is. So we're on the Trust Insights YouTube account, right? So that um this is literally um, I I typically don't visit my own stuff from the from my my company profile.

So just in getting this here, we've got a bunch of other stuff on my channel, and you know the show that we're literally recording right now. So again, if you can get people in, um they're gonna see your other stuff relatively quickly. Uh Dagmar says uh most B2B companies are mature and allergic to YouTube, they use a video storage location to embed in their websites, so transcriptions are after the omitted. They are. Um the other thing that I I think is unfortunate um is that when you load YouTube video to YouTube and you don't provide closed captions, YouTube tries to do it itself.

Uh and that never really goes well. Uh there's an awful lot of uh really, really uh poorly interpreted words, especially if uh you're like me and you talk really fast and YouTube just starts making hash uh of the words that you have to say. And Jesse says uh recent 739 to embed YouTube videos in your blogs and product pages. Absolutely right, as long as the videos don't suck. Well, yeah, but that's subjective.

It is, yeah. So that's sort of the content strategy side of YouTube, right? Take your predictive analytics for your content, use make sure it's YouTube specific, apply your forecast to the Hero Hub Help model, look for collaborators uh to generate those links from other channels to your channel, and then leverage the heck out of your website content, your email content, and so on and so forth to get people to your YouTube channel and get them watching your videos. That's the key, is you've got to get people to watch the videos 30 seconds, is all you need because that's what YouTube counts as a view. And once you've got a view on the clock for somebody, the algorithm is gonna temporarily work in your favor.

Well, and you know, we talk a lot about the transmedia framework where you record a video and then you can start to pull apart the pieces. So you can create your shorts you can create your audio files, you have your snippets your so are shorts what snippets are basically like is it the sort of the same it's like a 30 second like T Yeah 30 to 60 seconds. Oh yeah okay you didn't need a whole announcement about that Google. Um that already exists but basically so create the video and then repurpose it for a bunch of other things the other thing so because you're talking about the hero hub help and how your help content should be the most kind of content you're producing I like to use uh websites such as Answer the public that will give you all of those what we're how why when questions that people are asking about that topic and then you can just literally start ticking down the list of answering all of those questions. Exactly and the other thing that is a a slam dunk and very straightforward to do is ask your community right well we go to our Slack group and we ask them questions all the time like hey uh somebody tell us you know what how how would you think about this thing or what questions you have about this thing and then you're getting authentic natural language from other human beings um to that you can use as a basis for content.

Mm-hmm John thoughts check out my shorts that's all I can say that statement is not endorsed by Trust Insights they just the jokes just right themselves. I think we're done for this week. I think so. Thanks for tuning in. We'll see you next week.

Thanks for watching today. Be sure to subscribe to our show wherever you're watching it. For more resources and to learn more, check out the Trust Insights Podcast at Trustinsights.ai slash TI podcast. And a weekly email newsletter at TrustInsights.ai slash newsletter. Got questions about what you saw in today's episode?

Join our free Analytics for Marketers Slack group at TrustInsights.ai slash analytics for marketers. See ya next time.