SEO for technical articles

Denis Wilson Souze Rosa: We are going to talk about SEO for technical articles here. Essentially, I think that this is a very essential topic that many people overlook. If you but first, let's talk about myself first. I am Dennis. I work for a company called Couchbase, which is a NoSQL database.

I have been working as a software developer for fifteen years. I'm mostly a NoSQL and Java special specialist. SEO is not usually a topic that I talk about. I mostly use SEO. And the reason why I'm talking about SEO today is because I actually had an SEO startup in the past, and that's pretty much where my my whole knowledge of SEO comes from.

So, before we start, I would like to start with, just I have just one rule, which is don't use those techniques here against my articles. So if we ever are writing about the same stuff and see, oh, there is already a good article. There's already wrote a article about about this topic. Please write some about something else because if done right, you can really out you can actually outrank pretty much every single content out there. So, well, one of the things that I think that are important when you write write in a content is to understand what you expect from this article.

Right? It sounds like it's it might sound like a dumb question, at first, I don't think that this is a it is a question that should be asked because over the years, I noticed that some people want to write an article when they actually want to, just talk they want to write news, and news is a different branch of SEO. Or some other people just want to send their message. Right? So they don't really care deep down, they don't really care about how many people will read the article.

They just want to discuss some specific topic or find their tribe. But for me, naturally, the the word I have been working on, I really care about readers. That's when I'm writing something, I want to maximize how many people will read my article. Just giving you guys here an example, this is one article that I wrote in the past. It's actually, divided into, two, parts.

And you can see that, okay, we have some broken analytics here. This, print screen is actually from early, this year, so January. Now we have, much more traffic, Much more number of views on this on this page. But you can see that we have roughly 200,000 page views. And if we do some quick math and we multiply the number of views, let's say, 106,000.

Right? And multiply the number of views versus, the average time on page, we will end up with something like two hundred six, six thousand minutes that people spent reading this article. If we do some quick maths, again, we can actually, say that, okay, this article is as good as 200 more than 200 talks. That's how powerful, writing content is. Right?

And since since I wrote this article two years ago, Ray, it has been it it was mentioned in in a few talks in some other people that I never met, so they mentioned this talk. Some other people that I met in the booth, they mentioned that, oh, yeah. They have heard about Couchbase because of those. They read some article about the specific topic that I wrote in our blog. And even surprisingly, some people even come to the meetups where I usually speak to to ask questions about this specific topic.

So this is just a few examples of how powerful blogging actually is. And, I mean, it kind of blow it blows my mind as well when I see that this kind of things happening. I I know that we are living in strange times, and we won't be able to travel to conference and speaking with people in the booth for for at least for a few months, I would say. But the point that I'm making here is that we can make a great impact by just blogging guys. That's really powerful.

And, of course, SEO is just it is an important part of that. One of the problems here with SEO is when we talk about SEO, people usually still have in the back of their minds this SEO from the early two thousand ten. Back then, SEO was a very technical topic. So, but when I started my my startup on this, area, most of the things we were doing was friendly URLs, meta tags, tree using the the forward proper forward pro protocols, some small growth hacks. And we used to do things that are even considered black hat today, but they were not back then.

So that's how SEO was in 2010. If we fast forward to 2020, well, Google got far smarter now. Some experts said that Google use should use something around 400 different signals to to rank your page. So they have, like, more than 400 items that they check to rank your page. This is just a graph from Moss.

So Moss is some other companies are kind of a common thing to try to reverse engineering some of the Google algorithms to understand how their weight is distributed. And one thing that I noticed since I started was that in every single Google update, because Google updates the content the algorithm every four months, let's say. In every single Google update, one thing that was clear was that, hey, we want to make good content rank higher. And that's why we actually see this mantra, this content king content is king mantra being repeated over and over in the Internet. Because now, yes, there is some technical stuff that still mirror, but, if we're writing good content, it will probably rank really well.

However, the natural question is, okay. How do I write a good content that also ranks well? And, well, if I have to give just one tip for you guys is to to to spend 20% of your time on keyword research. That's what I usually do. If I have five days writing an article to write an article, I will spend one day just, doing keyword research.

And quite often, the right keywords for your article are not the obvious ones. So for instance, Couchbase is in the SQL database. We also had support for two sec full text search. We also had support for transactions. And potentially, when I write an article, that's not the keywords I'm gonna focus on.

You have the the gold mine is on the long tail. And luckily, we work in the IT field, most of us, of course. And the IT field is has plenty of keywords, and that's exactly what you should focus on. Like, go after the long tail keywords because that they those are the quick wins. Right?

If okay. You can, of course, rank for to rank as the first pursuit to no sequel database, right, or DevOps or all sort of things, but it will cost you a huge amount of money to do that. You have to to spend a lot of time and money to rank as the first results for those main keywords, while you can easily rank for all those other long tail keywords and do as good as ranking as the first result for this specific keyword. However, when you are doing keyword research, you should not pick keyword researches you should not pick your keyword based on guess. There are some specific metrics that you should use.

And the most important then of them is keyword volume, which is how many people search for these keywords monthly. Right? And keyword difficulty, which is how hard is to rank for those keywords. For instance, let's say I want to write an article about Pele, the the football player. Right?

Potentially, there's already a a Wikipedia page talking about Pele and all of his career. So for me, to write an article talking about Pele and ranking higher than Wikipedia is, is very unlike unlike, unlikely that we'll be able to outrank, Wikipedia. And that's exactly what keyword difficult is. So if there is already good content out there that has tons of links that people like, and I want you you focus on the same keyword, then it will probably, be very difficult for me to outrank this this, specific keyword. I highly recommend you using paid keyword researches for that.

I can actually demo one of one of those to you guys. Let me come here to now that's now that all my Chrome tabs were closed, I have to reopen here. So let's go to some rush. And then we'll say rush here, but because that's the one they have access to, but could be any other. So we could let's say I want to write a blog post about microservices and Java.

Right? So let's say Java microservices. That's it. I can come here and say Java microservices. And then note that I'm only searching in The US database.

So this is some estimations of how those keywords look like on US. And you see that, okay, those are the keywords that people usually search for. So I have Microsoft's in Java, Java Microsoft framework, Java Microsoft tutorial, Java Microsoft's example, and so on. So those are somewhat key related keywords that I could use on my content. Here is the volume.

So naturally, Microsoft's in Java is not Java Microsoft that I search it for. The most popular keyword is Microsoft's in Java. So that's already a change that I would need to do on my blog. There is also keyword difficulty here, which is, okay, how hard again, how hard is to rank, for, for this keyword. And what I need to find here is, okay, what is the keyword that has a high volume and a low keyword difficulty?

So anything that should that is under below 64 here, at least from from the Semrush point of view, should be reasonably reasonable. It should be easy enough to to outrank. If we have something 90 or something else, that's that's already super difficult to outrank. So that's pretty much what I what I do before writing a content. I come here and say, hey.

Java, Microsoft's example, GitHub, that should be a very seems to be a very difficult keyword to use. So I should not target this keyword on my content. And then and, of course, when you once you click on this keyword, most of the those two, they usually show you some stats. So what's the global volume? What is the some related keywords as well to this one.

So let's see some other variations of this keyword. And that's pretty much what I do before writing a content. I think, guys, this is one of the most important steps. Before writing something, try to understand what people search for, how hard it is to rank for for this content. Okay.

I'll I'll only have twenty five minutes, so I can't make a good walk through this whole tool. But another important thing here as well is you have to target multiple article multiple keywords in the in the same article. In fact, that's even recommended. So for instance, let's go to article of mine here called fuzzy stream matching. Right?

So I wrote this article, like, just one day. It was super fast. And if you search here keyword, of course, is fuzzy matching because I have here fuzzy matching, fuzzy matching the title. I still have, like, fuzzy matching here. And but if you search in the content, you see something like fuzz let me fuzzy string matching.

What else? We have some other variations here like, let's say, fuzz fuzzy lockups and a lot of other variations of this this keyword. In fact, levy's time distance is also another variation of this keyword. Edit distance is also another variation of this fuzzy matching. So that's exactly what I'm doing here.

I have my target keyword, and then I use, throughout my content. I use some variations of this keyword as well. Okay. Let's go back here. Another very important, easy way to add more extra keywords in your article is to add code.

Because I am a developer, and quite often, I search for blocks of content of code on Google just to understand maybe I'm not sure how do I instantiate a new class or what are the possibilities that how can I use h t p let's go back here? How do I use h t p this entity here or this object? How do I manipulate this object? So quite often, I would copy a block of of code and just place in Google and see what are the other codes or other tutorials that are using this specific top this specific block so I can understand how they are using it. And one advantage here is that, okay, although GitHub has already a lot of content there, so you can find you can go on GitHub and find the same class, Google don't index GitHub very well.

I I think you most of you have never seen GitHub class rank as the first result. Right? Most of sometimes you see the the the GitHub homepage or of a project because this is kind of relevant, But some of the internal classes, never. So if you add code to your page, that's definitely help you to rank for other keywords as well. I did I have some examples like I just I wrote some coding, like, three hours.

I thought, okay. This could be a tutorial. I just literally almost place a couple and paste the code, some base explanations, and boom. A solid number of views per month on catch based blog. I mean, we also talked about keywords, but I think it's essentially to talk about what a what a good content is.

I think, the most basic definition of what good content is is it has to answer the user question. That's something you should always bear in mind. So if the user is searching for the specific keyword, you have to understand, okay, what they what he wants to accomplish here. And I want to make sure that this content answers the question that he he is searching for. Another important thing that I noticed that a lot of people forget about is if you want to be the first result, your content should be as good as the other top 10.

Right? So if I come here and search for, Java, micro services, right, My content, in order to rank as the first result, should be as good as the other top 10 results. So what I usually do is, okay, I go through all the top 10 results. I read all the top 10 results, and I make sure that my content has nearly the same content. We cover the same stuff plus something else.

And that's how I am sure that, okay, Google is already telling me here what he thinks that is relevant for this keyword. He's already telling me when I search Java microstats, he he tell me, hey, those are the top pages, most more relevant pages for this specific keyword. And what I want to make sure when I write something is, hey, my content is as good or better than all those top 10 results. So and that's something that people usually totally overlook. You also want to make sure that your content is shareable, so people want to share your content.

And you also link to other related content as well. Back then, we we used to think that, okay, linking to other pages will make you lose the juice of your page. That's not true. In fact, a good page will link to other related, and trusted resources. And, of course, a good content has also images and videos.

And here comes something that I think that's super, super important. If you make videos, add the transcripts because Google don't index the content of your of your video. So Google does index, the title, the description, but doesn't it does not index it it don't index, the the transcript. So if in the middle of the video, you're talking about something really, really awesome and and somewhere is is someone is searching exactly about that on Google, your video won't appear on the result because you don't have transcript. Like, Google doesn't index the content.

Right? So what I usually do is whenever I record a video I know that this is painful. I know it takes some time. But when I whenever I record a video, I make sure that I add the whole transcript of the video because that's not for the user. That's for Google to understand what the content of the page is about.

Right? And that's really, really powerful. I have seen, like, okay, videos have not huge improvement just because we a huge increase in traffic just because we added transcript for the videos. Other common mistakes here is I I I see that other people like to add some people like to add catchy titles. So for instance, they want to write some they they give the article some smart title that doesn't use the keyword, and that's literally screw up with the with the relevance for Google.

We also see other people that in the beginning of the article, they don't describe what this article is about. And Google usually have this content snippet here, which you when you are the result, it basically shows, okay. This is what this talk this page is talking about. And if you don't have that, yeah, that might be not interesting for a user to click on. So, having a good description or or if you have meta description or a first paragraph is a very good description of of what this page is about, then we'll help to increase your click through rate.

There is also no magic template. So I have seen people say, hey. Do you have a template of a blog that I could follow that is SEO friendly? There's not such a thing as that. If someone tells you that, you can say, hey.

Dennis told me that this doesn't exist. It's also important I know that it's difficult, but it's also so important to ask other people inside the company to review your content. You can get amazing feedback on that and make your article even more powerful. It's also important to do keyword research before writing the content. Right?

Because sometimes, if I I'm I want to write about Java and microservice, maybe when I do keyword research, I might find out that, okay, I should have written about distributed microservice in Java or some or, I don't know, some something some other topic which is easier to rank and have a higher volume. And don't rely on on on SEO plugins as well, and fire and forget is one of the common mistakes. So congratulations. Once you publish your article, you you're just half done. There is still a lot of work you should be doing after you publish your blog.

One thing that I see people doing is, okay. Yeah. I published my blog, so my company just share on Facebook and Twitter. That's not content distribution. What you should be doing instead is, hey, let's create a list of relevant places and communities where we we should share this content and make people read them.

Like, we want to find the the the readers. Right? And then you let the the authors to go there and manually share the article because they this give it's it's a good thing when the author goes there and and share the item instead of doing this via some automated tool. And, of course, you have to revisit other related blog posts and link to the your new article because link building is something that matters for ranking article well. And if you have budget, also, please please try to promote your content.

I am not saying for you to spend lots of money on a campaign for content, but spending like $5,100 is already good enough to give this initial this initial boost that the article need. It's also worth to highlight that your article will take a while to rank well. This is some strategy of the Google of from Google to avoid spammers. Right? So if Google doesn't allow you to rank.

Even if you wrote the best article in the world, it do it should it should take a couple weeks for you to appear as a first result. So bear in mind that content is a long term strategy. It will take you three to six months to see some good results. And articles are also living content. Right?

So you have to share in in other places after a few months. You have to syndicate your content in third party websites as well. You have to update your old products after your old blogs after a pro a new product release. So for instance, Couchbase didn't have in the past transactions. And in a lot of those those contents, we make sure that, okay, we don't support transactions.

And now we do we do so. We added support for transactions. So I have to revisit all those bugs and say, hey. Guess what? And now, Couchbase supports transactions.

Right? Because the people the person landing the space, they don't know that, okay. We just launched this this support for transactions now. So it's very important to update the old content as well. And it's okay if the if after three months, your article is not ranking well.

It's okay for you to to also go there and update and try to change the keywords and change the article to see how it work, if if you have any problems. And, of course, if you speak another language, be make sure to always translate your content if possible. I just had two slides left. I have some other few tips here. In the titles, make sure sometimes it's cool to add numbers.

So this is more of a psychological pattern. People like to see numbers and titles. So, five common mistakes, for microservice, five common microservice mistakes or how to speed up your performance in 20%. This is are usually very good titles. People like numbers there or dates.

So for instance, this is the title of my talk, state of no sequel in 2020. Make sure that, your, titles are not longer than 14 words. And it's also cool to add some call for action. So find, download, watch, and whatever you want the user to do. In the content, make sure to add, videos as well in inside your content and add out, the out fill the out, attribute in an images because it might help you to rank in the Google images as well, and people will could come from, they might look for images and land in your blog.

Image sizes, usually, page speed is a ranking factor as well. Make sure for that your articles is easy is easy to share. So add the the social share buttons. And, of course, take leverage features snippets in Google. So for instance, when something when you search for something and you see this other cards in the left, so this card can be can come from any result in the page.

So even if you are in the fourth or fifth place, for instance, you can if you have feature snippets, people can still see the snippet there and they will click on your blog instead of clicking on the first result. This is a very this is a I mean, I have a lot of stuff that I would like to talk about, but I probably far over the time that I have. So, if you have any questions, I I really appreciate. Feel free to reach me out. And, yeah, thank you for having me.

Speaker 2: Alright. Thank you. I know we've had some some tech hiccups. So sorry about that. We do have a question.

And, any questions that we don't get answered, I'll, you know, direct you to answer them on Slack so that we don't bleed too much into the the rest of our our agenda for today. But the first question we've got is from somebody out on YouTube. It says, are there any differences between a technical documentation and a technical article when it comes to these SEO recommendations? And how do you guys handle this at Couchbase?

Denis Wilson Souze Rosa: So documentation has some one of the problems with documentation in general is the version. So one thing that you should be aware of is quite often people can land in the wrong version of your documentation. And that's some of the we had some problem in past, for instance, where old pages on the documentation would rank higher than the actual the the the current ones. And that's something they should be very careful because that might be be misleading. Some some people might might not notice that.

And so we did some great work now to make sure that, okay, whenever you search for something in coach base, the first page that will appear is their most recent documentation. So I I think that's the most important thing. Apart from that, I do think that documentation is in general, is it should be as you should follow the same steps when you write an article or documentation. But the documentation has only this issue with versioning. In fact, one of the popular pages on on couch on our documentation is an article about NTTP, which is nothing related to NoSQL.

But the way we wrote the article was really good, and every time we we even laugh about it because it's one of the top pages on the documentation.

Speaker 2: I think you mentioned what the Google uses 400 different signals or ways, and some of it sounds like it's a little bit magical, and you're not really sure how it happens, so you laugh about it. I know I had a question. It seems like you've you've kind of got yourself in this mindset of thinking about target keywords and keyword variations. Can you talk a little bit about how you developed that skill, and was it from using tools like Semrush and seeing the the other words that it recommends?

Denis Wilson Souze Rosa: So I think that, in general, keyword research is something that you have to practice a little bit. But in general, I what I usually do is when I when I'm about to write about I have an idea of what should I write about. So I know, let's say, in this fuzzy matching article, for instance, I would like to talk about catch based full text search. But what I usually try to do is, okay, let's try to find a keyword related to something that people are already searching searching for. I can, give them the answer they are searching for and then talk about what I want to talk about.

So I'm not usually just talking about the product itself. I am, in some some extent, contributing to creating some grow some great content around a common question. And then if if you are searching for fuzzy string matching, let's say, you are so interested in full text search because that's the those are very related very common terms. And then it's okay for me to also talk about full text search after that because that's one of the ways where full text search is used for to fuzzing to do fuzzing magic. So, in general, what I try to do is, okay, let's, I have just an general idea of what I want to write about.

I do some keyword research and say, okay, seems like people are more interested in those specific topics and that's and then I wrap up and I write my content around those main keywords that I have found. So that's why I think that it's important to do keyword research first before writing a content because then it's too late. Right? You just have to if you do keyword research after your content is written, then I mean, you have to find some keyword that match exactly what your text is already talking about. When I do the opposite, I I have more much more flexibility.