How to Do Keyword Research: Go Beyond Search Volume (Part 1/3)
This lesson is the first one in our brand new keyword research tutorial series using our Keywords Explorer tool.
We’re going to cover quite a few things a lot of people don’t know of, think of, or fully understand when it comes to keyword research.
So let’s dive right in.
So we’re inside Ahrefs’ Keywords Explorer tool and we’re going to search for the keyword phrase “how to grill steak.”
And here, you can change the country, but we’ll leave it to find data from US searches.
Down below you see some deep metrics on this search query.
The first part is the keyword difficulty metric.
This score gives you an estimate of how competitive the top 10 results in Google are for a given keyword based on the average number of referring domains they have.
This is the only metric we use for KD so you shouldn’t treat this one or two digit number as the answer to know which keywords to target. It’s just a proxy to link popularity of the search results.
There are a lot of other variables to gauging keyword difficulty. You will be able to read how to analyze whether you can rank for a keyword in the third article in this series.
Next, you can see that this phrase averages 11 thousand monthly searches. And this big number is based on the country that you selected in the last step.
You can see the global search volume as well as the breakdown of searches by country which is ordered from the most to least monthly searches. So for this particular query, about 90% of searches come from the US.
But before we continue, what the heck is search volume anyway?
In short, search volume represents the average number of monthly searches on Google for a given query.
So technically, if the same person searched for the same keyword a million times, then the keyword volume would go up too. So you can see how this metric can be easily manipulated and isn’t exactly the most reliable one.
The only source for Google search volumes has and likely always will be from Google themselves. But there’s also an additional source called clickstream data.
So here at Ahrefs, we use both of these sources to make sure search volume estimates are accurate and most importantly, updated every single month. And this is the exact reason why we add an extra source of data to our search volumes.
If we jump back to the global search volumes, it’s pretty clear that this query is way more popular in the US and then there’s a tiny bit of traffic everywhere else.
But other times, you’ll see the opposite effect where a keyword is significantly more popular in other countries.
For example, if we type in “rugby” you’ll see that this query is the most popular in France, the UK, and then the US.
This is a good way to decipher whether or not your keyword is worth targeting for your business. If you find that search volumes are low where your target audience is, then it may not be the best keyword to target.
Search Volume trend
If we look at the search volume for the original country we selected, then you’ll see this cool graph where you can see the volume trend by months.
So you’ll see that people are barbecuing less in the winter months and more in the summer, which tells us a little bit about keyword intent. It seems like common sense, but things like this often get overlooked during people’s keyword selection process.
This graph is really helpful and often the first place our eyes gravitate to when we’re doing our keyword research.
If we look up the word “Christmas” and look at the search volume, you’ll see that there are over 800,000 monthly searches.
Well...not really. Search volumes are rounded averages.
So if you look at this graph, you’ll see a trend where people start searching for this keyword in November and December.
But the rest of the year? Pretty much crickets.
Look at the results for the search query “presidential election.” There’s a huge spike during the time the event was happening.
And then crickets.
The number promises you 615,000 monthly searches next month, but the trend shows that you’ll likely to get a few million visitors only during the time of the election, which happens every four years.
We have a great post on the Ahrefs blog on this topic “Keyword Search Volume: Things you didn’t know you don’t know”, which we recommend you to read.
Next, we have a very important section, which is “clicks”. This number reflects the total number of clicks that actually happen on the search results page. This often tells us whether a searcher clicks on multiple results or maybe none at all.
And again, this tells us a lot about keyword intent, which can’t be ignored.
With this example, and actually most “how to” search queries, the clicks are normally close to the actual search volume.
We’re going to show you two drastically different search queries to better illustrate our point.
We’ll open up a new instance of Keywords Explorer and for the first search query, we’ll type in “time in new york” and for the second, we’ll type in “chicken soup recipes.”
Next, we’ll click on the metrics tab.
Now take a look at these results.
Even though the query, “time in New York” has 100,000 more monthly searches, than “chicken soup recipes,” the latter gets more clicks.
How is that even possible?
The best way to illustrate this is to put yourself in the driver’s seat. If we Google “time in new york” and you see this result, has the reason for your search query been solved?
Yeah, it obviously has. This is a fact-based query and you don’t need more answers than what Google already gives you.
Now let’s pretend that you want to make chicken soup. So you type in “chicken soup recipes” in Google.
Now there’s a featured snippet here from Google, but can you tell me if the reason for your query has been solved?
Even if all of the directions were here, you’d probably want to compare other recipes. Maybe you’ll click on one of these results and be like...darn... I don’t have celery.
So you’ll look for other recipes. Or maybe you’re some chicken soup connoisseur and only eat 5-star broth. So you’ll look at reviews.
Now, if we look at this comparison again, you’ll also notice the “RR” metric which stands for “return rate.”
Return rate is a relative number that gives you an idea of how often people perform the same search over the course of a month.
This doesn’t mean that people search for “time in new york” X% or X times more often than chicken soup recipes.
We know it sounds confusing, but what you can take away from this is that people search for “time in new york” more often. Another example of a query that people search for in Google again and again is “facebook”, which has the highest return rate.
You’ll see that it has a return rate of 3.7 in the United States and obviously this doesn’t mean that people only search for this 3 or 4 times per month on Google. But it helps you understand how other searches compare to this “baseline.”
Also in the clicks section, you can judge the commercial value of the keyword phrase. So here you can see that it has a cost per click of a $1.20, but what’s more interesting is the percentage of clicks that go to paid vs. organic.
You can see that paid search only gets about 2% of the clicks, while organic takes the rest.
And if you compare this with the keyword “car insurance”, you’ll see something completely different.
The CPC is $40 per click and 38% of the clicks go to advertisers. And it makes sense.
If we Google the search query, then you’ll see that the entire fold of the Google Search Results is plastered with ads.
One last example we want to show you is something like the keyword phrase “medical school.”
You can see that the entire fold is plastered with ads, but check out the data in Ahrefs.
Only 2% of clicks go to paid ads!
The thing that this boils down to again, is keyword intent.
If you’re looking into “medical schools”, you may be looking for a list of schools, school ratings, or you want to know more information like how many years is medical school.
There are so many possibilities with a query like this that it shows that paying for this keyword term may not produce the results you’d expect.
Here you can find the same cool graphs that will show you the separation between paid and organic clicks. So for certain queries, you might find that advertisers are more active during seasonal months.
Now, with the Adwords CPC figures, it’s worth mentioning that we don’t update it that frequently. So treat these as approximate values and if you need the freshest data, then go to the source: Adwords.
CPC can change pretty much every hour and they’ll still be estimations since Adwords runs as an auction.
But the main takeaway from all of these examples is to focus on search queries that will generate clicks for you.
Because even if there is high search volume, but no one ever clicks to your site, then what would be the point of targeting that keyword? Right?
Traffic Potential and Long-tail Keywords
Alright, the next part of this article is super cool. So originally, when we were doing our research for this blog post, the keyword phrase that we entered was “how to BBQ steak.”
And if you look at the search volume, it’s only 350 searches per month in the United States.
Looking at the global volume, you can see that the country with the most searches is Canada! So we guess barbecuing is a Canadian thing which we would have never known.
So if you look down here, you’ll see that the parent topic is different. It’s actually the one that we’ve been showing you this whole time, “how to grill steak.”
The parent topic tries to determine if you can rank for your target keyword while targeting a more general topic on your page instead.
So the way we do this is by looking at the number 1 ranking page for your keyword and then look for the most popular keyword that brings that page the most clicks.
So if we had a site on barbecuing, then we would know that we can target the phrase “how to grill steak” to reach a larger global audience.
Now we want you to pay close attention here. You can see that this parent topic has 11,000 monthly searches, but the traffic potential is 77,000?
How is that even possible?
Let’s go back to the original keywords explorer results and we’re going to quickly look at metrics on the top 10 Google rankings for this keyword phrase. But first, we’re going to click the “update” button to get some fresh search results because the ones you’ll see by default are cached from this date that’s displayed next to the update button.
Now, if we scroll down to the bottom of the page you’ll see the top 10 SERPs and a bunch of cool metrics which we’ll go deep into in another tutorial.
The first two results are from a featured snippet and some related questions.
But look at the next one from foodnetwork.com. This one page is generating over 76,000 monthly search visitors from all of the search queries that it ranks for!
And if you look to the column next to it, you’ll see that it’s because they rank for over 4,000 search queries.
This tells us that this topic has a lot of similar long tail queries for which Google is showing this search result.
So if you wanted to get similar results, you could create a comprehensive resource, get some quality backlinks and rank for all sorts of relevant queries.
Isn’t that awesome? You can see the total search traffic of all of the top ranking pages so you can start targeting topics and not just single keywords.
A crazy thing you can do with the Top 10 SERP results feature
There’s also something crazy you can do with the top 10 SERP results feature. So we’re going to open up another tab with keywords explorer and type in something ridiculous.
we’ll type in: “what is spiderman’s web thingy.” Then we’ll run the search.
And as expected, the search volume is “not available” since it’s not exactly a natural query that many people would search for.
But if we scroll down to the top 10 SERPs, you can see that there are a bunch of different top keywords we could target and potentially rank for the target query, “what is spiderman’s web thingy.”
The “spiderman web” has 4,400 monthly searches with traffic potential of 802.
And “spiderman web” from Wikipedia has potential of around 3,000.
Then there’s “spiderman web shooter” which has a search volume of 5,200 with traffic potential of around a 1,000.
So in this case, we’d probably choose “spiderman’s web.” But since search volumes are a bit arbitrary, we’d have to do more research before settling on the topic.
The key takeaway here is to focus on traffic potential of a topic rather than targeting a single keyword based on search volume alone, which we feel like we’ve really drilled down on.
And as you and we continue to go through this series, you’re going to find that shifting your focus to this traffic centered keyword research process, paired with some crafty tactics will pay dividends.
So we hope you enjoyed this keyword research tutorial.
By this point, you should be a master at analyzing search volumes of keywords, determining keyword intent based on clicks (and some common sense), and most importantly, you should see the value in using search traffic potential over search volume.
In the next article in this series, we’re going to show you how to find thousands of keywords with a bunch of cool hacks.