Candour

Blog

Episode 55: The moving target of keyword research

Episode 55: The moving target of keyword research ,
06.04.2020 • 25 min read • Podcast

Play this episode:

Or get it on:

What's in this episode?

In this episode, you will hear a SearchNorwich recording of Jon Falgate giving his talk on the keyword research process. Jon explores how keyword research is not just a process you go through once when you make a site or new page, but something that needs to ride the wave of intent and respond to ever-evolving demands.

Video

Slides

Transcription

MC: Welcome to episode 55 of the Search with Candour podcast! Recorded on Sunday the 5th of April 2020, where we're going to hear a SearchNorwich talk from Jon Falgate, ex-head of organic at Further, who's going to be talking about keyword research.

I picked this talk for two reasons; firstly, I think at the moment it's really timely because a lot of the takeaway you'll hear from Jon's talk is that keyword research is not this one-off process that you just go through when you're building a site or making a new page. As the world changes, as people's wants and needs change, keyword research needs to move along with that and I don't think there's any bigger example of that than what we're going through now. CloudFlare is reporting, in some areas, there's a 40% increase in internet usage, that means whole new audiences are coming online to e-commerce that has traditionally avoided it before and those products and services that were predominantly served offline are going to be forced through an evolution to become digital and all of this brings a groundswell of new search intent.

Secondly, next week I'm going to get e-commerce consultant Luke Carthy to join us. He's been giving some brilliant advice online to retailers about how to navigate the current conditions and I think listening to John's talk before this is some really great groundwork. Here's John, I hope you enjoy it.

JF: Cool, so the other thing that would be useful for me to know in advance is who out of everybody here actually does keyword research or if not does it kind of in-principle, understands why we do or how to do it. Cool, that's most of you. And then does everybody here roughly knows what Google search console is or ideally how to use it or what it's for? - okay, maybe I'll just do a little quick filler - alright, I'll do it as I go along.

But I'm going to talk about some processes that we would use Google search console to do and other paid keyword tools as well but predominantly because it's free Google search console. So if some of this doesn't make a lot of sense, I'll definitely share some blog pieces on how to use it. So I don't want to talk too much about the how to do keyword research because that's a whole different talk, I want to talk about that the why and the when so I'll skim over some of the keyword processes and then I'll share some blog pieces on how to do some of that stuff separately. I'll talk about it, but I'll share some more in-depth stuff.

Okay so like I said I want to talk about the when rather than the how and the why - I want to talk about how we approach keywords. So I think probably quite a lot of us here are a little bit guilty of this, some of us very guilty, someone was not so guilty, we all do it differently, we all do this sometimes and sometimes maybe this is what people think we do, but we approach keyword research as this checkbox activity that we do in order to inform the build of a new page or the reason for a new page to exist. We gather a whole bunch of keywords, use it to optimise a page and then move on. The activity stops once our page goes live. So I want to put to you that keyword research is a lot more than this, that it's an ongoing thing and there are loads and loads of times when we should be looking to gather keywords or measure our keywords or dig into this idea of keywords beyond just when the page goes live. So we can plan and create our page, but we can also measure it once it's gone live and use this information to then inform incremental improvements to our page as we go and we can also use keyword research to agree with ourselves that we should stop broadening the keyword targeting of that page and actually we're at that point when we should actually maybe now create another page to target a new set of search intent.

Okay, so if we approach keyword research like this then we can get to start to get the sense of a web page having a life of its own that it's something that we can nurture and it's something that we can - I think I already said incrementally improve on over time. So we can keep what new keywords are found then we can use those to inform how we might improve the page or add additional content or targeting to the page. And, not just keywords as well like we also have lots of other user metrics we can get from search console and we can get from Google Analytics; I’ll definitely be scheming over these, all of this is important but I like keywords a lot so I'm going to talk about those.

Cool so let's create a page together. let's all start with a new company and talk about and sell or let's do something with static caravans - the talk I did the other month was about boats, so this one is gonna be about static caravans. Cool, so we've got a keyword; we're gonna make our page, we're going to use that keyword as a stem so then go and gather a whole bunch of information about other related keywords and what we're probably going to find is that mostly they're the same, they definitely have the same intent and they're just some other variations of the same keywords and really there's like some other ones with some slightly different intent and we're going to ignore this anyway and that's about as much information as we've got for most speculative keyword research in advance. If you're going to be picky, there are loads of things we can do - we can steal competitor keywords and all this kind of stuff but if I'm being really simplified we've got these keywords some keyword planner and they're a bit sh#t.

So we can use those keywords to generate an idea of how we might rank that page, whether we need links to do it, the length it needs to be, the quality it needs to be etc. etc. and we've got loads of tools for that SEMrush, href, all this kind of stuff that - Moz if you use it, whichever tools you are using, there's a whole bunch of free ones. Then we make the page live, we pat ourselves on the back because we did a really good job at building a new page for our website, we're all super happy with it. And that's for me is when the problem starts because all of that is a really good hypothesis, that's really good guesswork, we did a really good job at reverse engineering Google and figuring out what's probably going to make that page rank, but we definitely feel it, we feel that content grindstone we feel that pressure to keep making new pages all the time, we might have promised a client ten new pages or we might subscribe to this myth that Google needs to crawl fresh pages to understand that your website's constantly updated or you know whatever these reasons are, we feel like we just need to keep making pages.

But I would say that really it's only once that page goes live that it starts to really get good and actually that's when you should really start paying attention to your page. So this is a little fact, Google themselves confirm that 15% of all of our searches every single day are brand new, as it never been searched before, so Google receives 3 billion searches a day, forty thousand keywords a month I think - I did Google it before I came in so I didn't forget that, I might be wrong but it's something like that, something ridiculous. The 15 per cent of all of those keywords have never been searched before.

So let's take an example you have a page talking about the top 10 gadgets for kids and you've worked that 5 years ago and it ranked really well and it got some links and it's still kind of ranks quite well because of those links, but if you wrote it five years ago and you've never touched it again, it didn't talk about fidget spinners - that trend has come and gone, where you didn't use Google Trends to know what was up-and-coming and what was new and how to make that page relevant and how to make that page topically relevant to that subject right now and actually for all the links in the world, that's just going to you know that's going to die over time. And the really important point that I want to make, the one that I'll probably label a lot from now on is that, ok so this is a really common scenario you've probably seen it on all your websites, we see on all sites and we definitely see it on our own website, that a very very very small percentage of the articles or all the pages in our blog or on our website generate the line's share of the traffic. So on our own website out of 214 blog articles, 53% of the blog traffic is generated by just five pages, just five articles and those five pages rank so ridiculously well for some really really great keywords that it just overshadows every other page that we've got. There are some legitimate reasons for it because some of our pages aren't designed to rank, some of them are just talking about that we've got a new starter or we've got a new client or something, so there's we're not expecting traffic from those and then a whole bunch, 10 or 15 of those articles sit just below those their top 5 pages and they also generate some half-decent traffic as well, they do relatively well, but it doesn't change the fact that these are that those top five places are super important.

So sure laughs now. What does that mean? It doesn't mean 42, but that's an inside joke or maybe not but to me, that means that for starters they have five pages on my website, on my blog, that drives half of the traffic. I care a lot less about everything else than I do about those five pages and if I'm not spending half of my digital content resource, making sure that those pages are protected, making sure that I'm monitoring what other competing pages that are around it, what new pages are coming, as soon as the new page ranks, I can steal their keywords all this kind of stuff and unless I'm housekeeping and maintaining and protecting those pages, I'm definitely wasting or risking a massive proportion of the traffic that comes to my website. And then even after that, I've then got those kinds of 10 or 15 articles which should just below it which are driving ok traffic, which is bringing hundreds or thousands of visitor month, but just kind of not quite nailing it. I've got all those and now I've got data for those pages and I've got data for those; I can be spending all day working on improving those and I'm much more likely to get those pages to rank better, based on the data and the hypothesis I can build on how they perform than I am to just go back to throwing sh#t at a wall again and hoping with some of the stuff I do that I do rank. But of course, we need to do some of that but it just needs to be proportional to what we think is the best strategy for getting more traffic.

So I didn't press start on my timer - on my run through before I went a lot over, so this is why I'm talking quite fast. So okay I'll skip through this page, but what I want to say is basically once this page is live, you've got a whole bunch of data that you can then start to use build really much, much, much better hypotheses about how it might rank better compared with when you've got no data at all and you're just guessing at reverse engineering how Google works. And of course, if it's not an informational page or if it's an informational page which gets loads of traffic and it does make conversions, then instantly for all the people they're telling you to keep that content machine churning, instantly you've got a business case to suggest how much more money you might make if you could spend the time improving that one, as opposed to going back to machine-gunning pages onto your blog. But the other thing, the bit I really enjoy is there is the keyword part to this; if you're using search console or you're looking at your top organic keywords list in href or SEMRush or with Moz, or whatever this is or ideally you're doing both. In both ideally, you've got Google tools and you've got your paid tools, if you're looking at how at the keywords which rank just outside of those traffic-driving positions, those kinds of peripheral keywords, I've got some examples so don't worry too much but if you're kind of measuring and researching what those keywords are that you almost rank for which could potentially bring you an additional set of traffic and then figuring out how you can optimise your page and then get it to rank for those keywords as well, this is how you can take a page that gets some traffic and then and then grow it to a page that gets more traffic.

Right, so let's go back to our static caravans article, we did a super awesome job of getting it to rank for those keywords in the first place and it does that and we're happy right but if we use our tools to measure, we can see some other kind of slightly related keywords coming through, there's one here about how to drain down a caravan and there's one here about static caravan boilers and some cushion ones about cushions. So I'm keen in these drain down Caravan keywords - when we wrote the article in the first place we knew that when you drain down a static caravan that's one of the main things you do when you prepare for winter, which is the original keyword. But we didn't talk in-depth about it, it was just one of the things that we talked about in the article, but we can see that we're ranking 15th and 18th for this keyword and it's quite high volume and super low difficulty. So, in theory, we could just add more content for that particular keyword and then rank for a whole new set of keywords right, still totally related to our original intent or our original topic.

So little game, our original keyword ‘winterising a static caravan’ and this is the new keyword that we spotted ‘how to drain down a caravan’ - by the way, draining down a caravan is if you're gonna leave your static caravan for the winter, you'd take the plug on the bottom and all the water drains out so the pipes don't freeze, that's basically what it is. I should probably I should have said that earlier. Right, so who thinks that the original key with the winterising one has got a higher volume as a keyword, among higher monthly average volume, who thinks that one's higher? Does everyone think that the new ones higher? Does everyone think that? So you are wrong, but it gives a takeaway. They're both quite low volume keywords but they're both quite similar and the good thing about it is that this drained down one topically has loads and loads of volume, there are loads of variations of this keyword which mean the same thing; so if we can rank for that main one, then we will basically rank for all of them.

So just by adding or improving one section of our page, we can rank, we can then move that page into a traffic-driving position for all of these keywords and that would never have happened if we'd have just - when we first did our research we got all our related keywords for our main keyword but we didn't get these secondary or these subsets of keywords within that topic, we just had the keyword and tools didn't provide them. Okay so and I wanted to include another quick example because I realised as I was putting this slide together that I was just talking about our blog pieces over and over and over, but I want you guys to know that this doesn't just work for blog pieces, this works for our commercial landing pages too and this was the page we worked on, we audited for a client which is for convicted driver insurance. So as we talked about, insurance has high volumes and is really competitive right so we pulled in competitor keywords. I should add this a good point for agencies this is the page we inherited from a client, we didn't create this in the first place, we just needed to go to it to optimise it, so the keyword research maybe wasn't done very well in the first place, but if we pull the competitor websites, if we steal their keywords, if we pull all the data out of search console and we take every single piece of keyword data we can find, put it all together, we find that there are a whole bunch of nuanced keywords, which are relevant to that parent keyword but have all these different nuances and irrelevant to all these people with very different specific circumstances but are all still under that umbrella of that parent topic. So not only now have we got a whole set of additional keywords that we can target by adding these sections on the page, but we've also got all this great opportunity to really hone in and target and engage people about their actual specific situations and it looks like the same thing right, but car insurance for people with points on the license isn't the same thing as car insurance for people who have lost their license or have been previously banned, they are different people, they've got specific sets of concerns and we can talk to both of those people with our page. I wouldn't suggest that either of those need different pages, we're still under the same umbrella of that original keyword and all of these pages, I don't think what would you be you'd be pretty doorway ish if you started building pages for each of these I would say. So once we have this data, once we can just collect it and understand it, then it's just a simple case of maybe retargeting the page, adding some new sections or images, definitely adding some new internal links and then awesome. The case was a static caravan article, that was a page that we updated and we did double the traffic, it is possible, it’d one of the blog pieces I'll show you, I’ll share with you but hadn't actually how to do this stuff, it's a blog and their claim in the title of it is that they doubled the traffic of their blog in three months without creating new pages so it's a bit clickbait, but it's a case study, so this does work so it is important.

The point I want to make really with this slide is, so for all of the kind of vanity that we get when we pull all these really big great big keyword data lists out and we talk about how we can reverse engineer Google's search algorithms and how great we are at predicting how a page will rank, is that it’s not up to us. It is up to Google to decide how they rank their page, it's up to them to test what their bounce rates are and to figure out which pages engaged best for which topics and for which intents and how our page helps people do the things that they want to do and for every hypothesis, we can build as to why it would do that, really, until it goes live we can't actually see whether it worked or we can't see how it worked for how we wanted it to work and maybe it works slightly differently and we have to measure our keywords and understand how it works per keyword to understand what the page does successfully and what it doesn't and once we have this data we then it's way easier to to to help it improve.

So did any one time? Am I doing all right? Okay cool, I can take a breath.

So the final step of this end-to-end process is the idea that we've taken our page that we've built and we've kind of incrementally improved that over time but we want the page to still remain super relevant to the main intent that that page fulfils, so we want to know at what point do we stop build broadening the relevance of that page and actually then create a new page instead? So if you go back to my static caravan keywords, if we remember back from there, from those measurement slides I had, there were keywords about static care about maintenance and static caravan boiler problems and things like this. So for me, they don't feel like they were still relevant for our original page; if somebody wants to know how to get their static caravan ready for winter, while they might be the kind of do it themselves kind of people, they're not asking me for this generic maintenance advice at that particular point, so they don't really feel that relevant for that particular page but they definitely feel relevant to that person. I mean, for all the kind of cool and expensive tools we have, there isn't a much better way of going about this than like I said before, just asking Google what the right thing to do is. So, these extra keywords that we've got, we can Google them and start to see what Google thinks is the right content to answer for that keyword. So it’s known that Google prefers really, really, really long substantial pieces of content. We know because we can correlate the ranking factors, we know that word count generally equates through a high-ranking position, that doesn't mean we can just write sh#t on a page and expect it to rank better, it needs to still be relevant to the search intent.

So while we know that our page needs to be really long, we definitely don't want to water it with stuff that's not relevant to the topic. So we can ask Google, do we need like a really big long generic piece of content that covers winterisation and generic maintenance or is a generic maintenance article that we need to target that separately with a new page? We just need to literally ask Google, we just make these searches to figure that out. Not only we can see what ranks, but we can also see what other kinds of content we need to make and all this stuff, Google’s there to just tell us if we ask.

So for me, the most important thing as SEOs, the most important skill that we can get as SEOs is to learn how to stop being an SEO for a little while and learn how to be the person that's performing the search. So if I needed to know how to drain down my static caravan, would I care about generic maintenance tips? I would care about them, but I wouldn't care about them in this right moment when I'm trying to do something specific - this is where the relevance of internal linking comes in. So for me, it's not a page which covers both, it's a page which covers the thing that I need it to do right now and then a bunch of links to relevant content which I can assume that person would be interested in, based on what they've already told me and what they told me, is what they've searched into Google.

This is how important it is to understand which pages are driving and which keywords are driving traffic, so you can then build an idea of who it is you're talking to and what they're trying to do and that in turn then helps us lead on to understanding what kind of what this mythical topical authority is right. So if we can take the page that we rank for, that we get some traffic for and we can grow that to the point where it completely fulfils that intent for that person, at that time, for that topic and then used that to understand, right this is the next relevant topic for them, add an internal link to a new page and grow that up until we've completely and utterly fulfilled everything that that page can do for them and then we internally link again and we go to the next page, you can kind of see this united front, the f zip-locked topical authority right and these pages are ranking well because we've put a lot of hard work into them and they're getting the links naturally that we want them to do and they're getting the traffic and they're not bouncing.

This is that mythical SEO topical authority that everybody talks about, we don't really know exactly what that means and actually all that takes is for us just to be super organised on how we approach our content and really just keep a firm grasp on what it is that works for our website and what we can learn from that, all the failures but what we can learn from that and incrementally improve on now over time until till we've got a page that ranks first position. I do firmly believe that everybody can rank the first position for something. It's just a case of being honest with yourself, what is it that you deserve to rank for, not just what the high volume keywords that you want to rank for, but if you could just rank for something you can get that wedge in and you can build on that over time until you are topically authoritative on a topic right and everybody can rank for something, but you know it hardly ever happens by just throwing sh#t at a wall, basically. Awesome, thanks.

MC: I really hope you enjoyed that talk, you can see the video of John doing that talk along with his slides at search.withcandour.co.uk along with the transcript of this episode. I mentioned a while ago since John actually did that talk, our keyword research tool came out which is AlsoAsked.com - A-L-S-O-A-S-K-E-D .com - it's available, it's still free and it helps you explore the people also ask questions from Google by returning all of the related questions as Google categorises them, so it's a great way to start your keyword and intent research. You can just give it to content teams and it will give them an idea about what questions people are asking about specific topics and how you should group them in terms of pages and articles.

If you are listening online, don't forget to subscribe with your favourite podcast app, if you like it that much. We'll be back, hopefully speaking with Luke, in one week's time on Monday the 13th of April and until then I hope everybody is staying safe and sane.

More from the blog