Posted by randfish
The long tail of search can be a mysterious place to explore, often lacking the volume data that we usually rely on to guide us. But the keyword phrases you can uncover there are worth their weight in gold, often driving highly valuable traffic to your site. In this edition of Whiteboard Friday, Rand delves into core strategies you can use to make long tail keywords work in your favor, from niche-specific SEO to a bigger content strategy that catches many long tail searches in its net.
Video TranscriptionHowdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're going to chat about long tail SEO.
Now, for those of you who might not be familiar, there's basically a demand curve in the search engine world. Lots and lots of searchers are searching for very popular keywords in the NBA world like "NBA finals." Then we have a smaller number of folks who are searching for "basketball hoops," but it's still pretty substantial, right? Probably hundreds to thousands per month. Then maybe there are only a few dozen searches a month for something like "Miami Heat box ticket prices."
Then we get into the very long tail, where there are one, two, maybe three searches a month, or maybe not even. Maybe it's only a few searches per year for something like "retro Super Sonics customizable jersey Seattle."
Now, this is pretty tough to do keyword research anywhere in this long tail region. The long tail region is almost a mystery to us because the search engines themselves don't get enough volume to where they'd show it in a tool like AdWords or in Bing's research. Even Search Suggest or related searches will often not surface these kinds of terms and phrases. They just don't get enough volume. But for many businesses, and yours may be one of them, these keywords are actually quite valuable.
2 ways to think about long tail keyword targeting
#1: I think that there's this small set of hyper-targeted, specific keyword terms and phrases that are very high value to my business. I know they're not searched for very much, maybe only a couple of times a month, maybe not even that. But when they are, if I can drive the search traffic to my website, it's hugely valuable to me, and therefore it's worth pursuing a handful of these. A handful could be half a dozen, or it could be in the small hundreds that you decide these terms are worth going after even though they have a very small number of keyword searches. Remember, if we were to build 50 landing pages targeting terms that only get one or two searches a month, we still might get a hundred or a couple hundred searches every year coming to our site that are super valuable to the business. So these terms in general, when we're doing this hyper-specific, they need to be...
- Conversion-likely, meaning that we know we're going to convert those searchers into buyers if we can get them or searchers into whatever we need them to do.
- They should be very low competition, because not a lot of people know about these keywords. There's not a bunch of sites targeting them already. There are no keyword research tools out there that are showing this data.
- It should be a relatively small number of terms that we're targeting. Like I said, maybe a few dozen, maybe a couple hundred, generally not more than that.
- We're going to try and build specifically optimized pages to turn those searchers into customers or to serve them in whatever way we need.
#2: The second way is to have a large-scale sort of blast approach, where we're less targeted with our content, but we're covering a very wide range of keyword targets. This is what a lot of user-generated content sites, large blogs, and large content sites are doing with their work. Maybe they're doing some specific keyword targeting, but they're also kind of trying to reach this broad group of long tail keywords that might be in their niche. It tends to be the case that there's...
- A ton of content being produced.
- It's less conversion-focused in general, because we don't know the intent of all these searchers, particularly on the long tail terms.
- We are going to be targeting a large number of terms here.
- There are no specific keyword targets available. So, in general, we're focused more on the content itself and less on the specificity of that keyword targeting.
Niche + specific long tail SEO
Now, let's start with the niche and specific. The way I'm going to think about this is I might want to build these pages — my retro Super Sonics jerseys that are customizable — with my:
- Standard on-page SEO best practices.
- I'm going to do my smart internal linking.
- I really don't need very many external links. One or two will probably do it. In fact, a lot of times, when it comes to long tail, you can rank with no external links at all, internal links only.
- Quality content investment is still essential. I need to make sure that this page gets indexed by Google, and it has to do a great job of converting visitors. So it's got to serve the searcher intent. It can't look like automated content, it can't look low quality, and it certainly can't dissuade visitors from coming, because then I've wasted all the investment that I've made getting that searcher to my page. Especially since there are so few of them, I better make sure this page does a great job.
A) PPC is a great way to go. You can do a broad-term PPC buy in AdWords or in Bing, and then discover these hyper-specific opportunities. So if I'm buying keywords like "customizable jerseys," I might see that, sure, most of them are for teams and sports that I've heard of, but there might be some that come to me that are very, very long tail. This is actually a reason why you might want to do those broad PPC buys for discovery purposes, even if the ROI isn't paying off inside your AdWords campaign. You look and you go, "Hey, it doesn't pay to do this broad buy, but every week we're discovering new keywords for our long tail targeting that does make it worthwhile." That can be something to pay attention to.
B) You can use some keyword research tools, just not AdWords itself, because AdWords bias is to show you more commercial terms, and it biases to show you terms and phrases that do actually have search volume. What you want to do is actually find keyword research tools that can show you keywords with zero searches, no search volume at all. So you could use something like Moz's Keyword Explorer. You could use KeywordTool.io. You could use Übersuggest. You could use some of the keyword research tools from the other providers out there, like a Searchmetrics or what have you. But all of these kinds of terms, what you want to find are those 0–10 searches keywords, because those are going to be the ones that have very, very little volume but potentially are super high-value for your specific website or business.
C) Be aware that the keyword difficulty scores may not actually be that useful in these cases. Keyword difficulty scores — this is true for Moz's keyword difficulty score and for all the other tools that do keyword difficulty — what they tend to do is they look at a search result and then they say, "How many links or how high is the domain authority and page authority or all the link metrics that point to these 10 pages?" The problem is in a set where there are very few people doing very specific keyword targeting, you could have powerful pages that are not actually optimized at all for these keywords that aren't really relevant, and therefore it might be much easier than it looks like from a keyword difficulty score to rank for those pages. So my advice is to look at the keyword targeting to spot that opportunity. If you see that none of the 10 pages actually includes all the keywords, or only one of them seems to actually serve the searcher intent for these long tail keywords, you've probably found yourself a great long tail SEO opportunity.
Large-scale, untargeted long tail SEO
This is very, very different in approach. It's going to be for a different kind of website, different application. We are not targeting specific terms and phrases that we've identified. We're instead saying, "You know what? We want to have a big content strategy to own all types of long tail searches in a particular niche." That could be educational content. It could be discussion content. It could be product content, where you're supporting user-generated content, those kinds of things.
- I want a bias to the uniqueness of the content itself and real searcher value, which means I do need content that is useful to searchers, useful to real people. It can't be completely auto-generated.
- I'm worrying less about the particular keyword targeting. I know that I don't know which terms and phrases I'm going to be going after. So instead, I'm biasing to other things, like usefulness, amount of uniqueness of content, the quality of it, the value that it provides, the engagement metrics that I can look at in my analytics, all that kind of stuff.
- You want to be careful here. Anytime you're doing broad-scale content creation or enabling content creation on a platform, you've got to keep low-value, low-unique content pages out of Google's index. That could be done two ways. One, you limit the system to only allow in certain amounts of content before a page can even be published. Or you look at the quantity of content that's being created or the engagement metrics from your analytics, and you essentially block — via robots.txt or via meta robots tag — any of the pages that look like they're low-value, low-unique content.
A) This approach requires a lot of scalability, and so you need something like a:
- Discussion forum
- Q&A-style content
- User-posted product or service or business listings. Think something like an Etsy or a GitHub or a Moz Q&A, discussion forums like Reddit. These all support user-generated content.
- You can also go with non-UGC if it's editorially created. Something like a frequently updated blog or news content, particularly if you have enough of a staff that can create that content on a regular basis so that you're pumping out good stuff on a regular basis, that can also work. It's generally not as scalable, but you have to worry less about the uniqueness of quality content.
B) You don't want to fully automate this system. The worst thing you can possibly do is to take a site that has been doing well, pump out hundreds, thousands, tens of thousands of pages, throw them up on the site, they're low-quality content, low uniqueness of content, and Google can hit you with something like the Panda penalty, which has happened to a lot of sites that we've seen over the years. They continue to iterate and refine that, so be very cautious. You need some human curation in order to make sure the uniqueness of content and value remain above the level you need.
C) If you're going to be doing this large-scale content creation, I highly advise you to make the content management system or the UGC submission system work in your favor. Make it do some of that hard SEO legwork for you, things like...
- Nudging users to give more descriptive, more useful content when they're creating it for you.
- Require some minimum level of content in order to even be able to post it.
- Use spam software to be able to catch and evaluate stuff before it goes into your system. If it has lots of links, if it contains poison keywords, spam keywords, kick it out.
- Encourage and reward the high-quality contributions. If you see users or content that is consistently doing well through your engagement metrics, go find out who those users were, go reward them. Go promote that content. Push that to higher visibility. You want to make this a system that rewards the best stuff and keeps the bad stuff out. A great UGC content management system can do this for you if you build it right.
All right, everyone, look forward to your thoughts on long tail SEO, and we'll see you again next week for another edition of Whiteboard Friday. Take care.
Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!