With “not provided” taking over our keyword data, the Keyword Planner replacing the Keyword Tool, and Hummingbird changing the way Google interprets queries, it’s no secret that SEO’s emphasis on keywords has diminished. In fact, the notion that SEOs can’t look at keywords the same way ever again has been pretty much done to death. “That’s so October 2013,” right?
But now that we have some perspective, almost half a year of it, maybe it’s time we revisit the subject with some wisdom.
The keyword is anything but dead, provided you know how to use it. Let’s talk about why, and how you can still use them to expand your reach.
Hummingbird’s Real Relationship with Keywords
As you already know, Google’s core algorithm was rewritten and replaced sometime in August of 2013, as they announced on September 26th. And, as you already know, one of the primary reasons for this update was the growing prevalence of vocal search. Vocal search means longer queries, and that means Google has to adapt to make sense of a much wider variety of queries.
And here’s something you probably think you know.
If keywords matter less, long tailed keywords matter more… – Asher Elran at KISSmetrics
Long-tail keywords, which are normally added to pages during the natural process of writing high quality content, may generate more traffic for a website in the future. – Huffington Post
This popular opinion seems it makes sense at first glance, but a deeper analysis of what’s really going on, reveals that this isn’t quite how things are working. Too explore why, we first need to respond to another popular myth:
By and large, there’s been no major outcry among publishers that they’ve lost rankings. – Danny Sullivan
…we haven’t seen a big impact of Google Hummingbird on website traffic yet, not even a lot of complaints from webmasters, which is very common whenever there are updates to Google’s algorithm. – SEM Rush
But there was. The update came sometime in August, not September. Don’t you remember the reports of an update back on August 22 and 23, when virtually all of the keyword tools suggested a major update hit the SERPs. While it’s true that the forums were relatively quiet, there was no shortage of complaints. There were even speculations that it was a Penguin update.
The comment sections on all of these old posts reveal that some people certainly did lose their rankings around the time that Hummingbird came out.
Why? – Because pages that target long tail keywords matter less.
Here’s an example. This is what I get if I search for “how to make carrot cake”, a popular search term with 2400 monthly queries according to Google’s Keyword Planner:
And here’s what I get if I search for “bake carrot cake”. which only gets 10 monthly searches:
So we get only one difference in the first four results.
In effect, Google isn’t even really showing the results for “bake carrot cake”. It’s showing carrot cake recipe results, perhaps with a bit of preference for articles that happen to use the word “bake” somewhere in the text.
This is bad if you’re targeting specific long tail keywords.
How This Changes Keyword Research
Some people might say that all of this means keyword research doesn’t matter anymore, but I wholeheartedly disagree. That being said, it’s certainly different from how it used to be. Let’s explore how.
Let’s use the “bake carrot cake” example above. As it stands, there’s currently only one result on the first page with an exact match in the title:
More importantly, the exact phrase doesn’t turn up anywhere else in the next 10 pages.
This used to mean that it should be easy to make the front page if we just used the exact phrase “bake carrot cake” in our title.
That simply isn’t true anymore.
There actually are places on the web that use this exact phrase in the title. Here’s what I get if I search “bake carrot cake”, this time surrounding it with quotes:
That’s right. Lifehacker has a page about baking a carrot cake. But unless I wrap “bake carrot cake” in quotes, even they don’t even show up in the first ten pages of Google for the phrase.
So even though “bake carrot cake” is only searched for 10 times per month, and even though the exact phrase only turns up once on the front page, it’s actually a very difficult phrase to rank for.
That’s because your competitors are no longer the sites that use the exact phrase in the title. Your competitors are now whoever is showing up on the front page.
So, how should this affect our keyword research?
Let’s use another example. Suppose we were originally trying to rank for “make a wood chair”. Once upon a time, this would have been a pretty easy phrase to rank for. There aren’t any exact matches in the title on the first five pages and beyond:
But, once again, we discover that that the pages do exist if we wrap our search in quotation marks:
Instead, we need to analyze the competitors who actually show up in the search results. We can quickly realize that the results are a good match topically, even though they don’t use the exact same phrasing. We can also check up on their authority in Open Site Explorer:
That’s too much competition for a puny search phrase with only 10 monthly searches. So do we move on?
Not necessarily. Remember, the key change made by Hummingbird is that niche keywords get replaced by more popular queries. The results aren’t really for “make a wood chair”. The results are for pages about making wood chairs.
So we need to start looking at related keywords. To do that, we can take a look at the results in the Keyword Planner, and find more popular keywords that mean something similar:
And look at those results:
And what do you know? The first search result is exactly the same.
Such predictable unpredictability lends a whole new leverage to the use of keyword research tools beyond Google’s Keyword Planner.
Now that we’re talking about a phrase with 720 monthly searches, we might be able to justify the investment it would take to rank for this phrase.
The point here is that you can’t simply target more rare or awkward phrases in order to reduce your competition in the search results. Google will merely reinterpret the rare query as a more popular query.
When you perform keyword research, you need to look at the most popular search phrase that means essentially the same thing. This is going to give you a closer estimate of the amount of traffic available in the space, and the strength of your competition.
How do We Use Keywords in Modern SEO?
In the previous example, this would be a natural question to ask: “Wait, so which keyword should I be using, “how to build a chair” or “make a wood chair”?”
The answer is…that’s not quite the point.
The point is that a page with the title “Build an Elegant Dining Chair – Fine Woodworking” is ranking for both of them, even though it doesn’t use the words “make” or “wood” anywhere in the post, and it doesn’t use the words “how to” in the title.
And if you need a reminder that Google is…unpredictable, here’s what happens if we search for “build a chair”.
So now the number one result has “how to” in the title? Seriously Google? Maybe that’s because “build a chair” is actually a less popular query:
In any case, I hope you can see my point. As important as keyword research is, keyword usage is no longer about choosing a specific keyword and putting it in your title. It’s not about targeting “how to build a chair” or “make a wood chair”. It’s about targeting both of them at the same time.
This Shopify page might have “Small Business POS System” in the title tag, but that doesn’t mean it isn’t also targeting “pos systems for small businesses”, “small business pos software”, or any other query that means essentially the same thing.
Now, as with keyword research, I’m not saying that keyword usage doesn’t matter. I’m simply saying that it’s changed.
The title tag is still a crucial part of rankings. As far as SEO goes, it’s almost always going to be better to be descriptive and straightforward, as opposed to getting too creative. (And if search results are anything like email headlines, that’s probably good advice from a CTR standpoint as well.)
Also, keep in mind that virtually any time you perform a search in Google, the results on the front page will almost always have at least one bolded matching word (or synonym) in the title.
However, it’s not really important to use any specific matching phrase, or even necessarily to get all of the words in the title. All you need to do is make sure that the title contains a phrase that means essentially the same thing as all the various keyword phrases you are trying to target.
For example, if I search for “online personal loans”, not all of the result titles are going to contain the word “online”:
However, you’ll notice that the word “online” does still show up in the content. This is important conceptually. If the word online, or a synonym, didn’t show up anywhere in the content, then it would be a page about personal loans. That’s conceptually different from “online personal loans”.
For another example, while a search for “google hummingbird” primarily returns results that contain both “google” and “hummingbird” in the title, the front page still has this result:
Which only mentions Hummingbird in the content, and this result:
Which only mentions Google in the content.
To sum this section up, I would draw three conclusions:
- You still need to get a few words into the title that are relevant to the query
- Any conceptually relevant words that don’t make it into the title should be used somewhere in the content
- You don’t need to target any specific keyword phrase, but you do need to use enough relevant words to mean the same thing conceptually as a group of keyword phrases you are trying to target.
Keywords have changed, but they’re as meaningful as ever. In fact, they’re probably more meaningful, now that the skills necessary to use them properly are in shorter supply.
Here’s what we’ve covered:
- The direction Google is heading is bad for SEOs who specifically target long tail keywords. Hummingbird essentially replaces long tail queries with more popular queries, allowing pages to rank for multiple search phrases, even if the exact wording or phrasing for some of those queries isn’t used in the content.
- The best way to evaluate the potential traffic and the competition for a group of conceptually similar keywords is to look at the most popular query that means essentially the same thing.
- It’s no longer necessary to pick a specific keyword phrase and use it in the title. However, you still need to use some variant from your group of conceptually similar keywords, use at least part of it in the title, and use the rest of it somewhere in the content. Anything less is probably putting too much faith in Google at this point in time.
And that wraps this up. If you plan to use any of these ideas, I’d appreciate it if you passed this along. If you have something to add, we’d love to see it in the comments. Thanks for reading.
Note: The opinions expressed in this article are the views of the author, and not necessarily the views of Caphyon, its staff, or its partners.