Understanding keywords is fundamental to SEO. There’s little point ranking at number 1 for a keyword that hardly anybody searches for. We can’t know how often a keyword is searched for, and so in most cases we use the (soon to be retired) Keyword Volume Tool for Adwords.
If you’re reading this and you’re an SEO, skip to lower on the page where I talk about the correct settings in Adwords to get accurate keyword volume, as you have to be careful to get the settings right. For non SEO’s, I’ll talk about why you need to use Adwords to get accurate data.
The keywords you think your customers use are not always the ones they do use
I’ll tell you a quick story. I built a website about 3 years ago targeting a specific keyword. According to Google’s keyword tool, the keyword had 1200 searches a month in its plural form and 900 a month in its singular form. That’s not a huge amount of searches, but with an average CPC of around £30 for these 2 keywords it should have been insanely profitable. Assuming I could get to number 1, get a 20% CTR, and get a value from my traffic of at least the Adwords cost I should have made the following amount of money:
2100 (searches) * 20% (CTR) * £30 (value of clicks) = £12,600 per month.
That’s nice money if you have the skills to rank a site in a (clearly) competitive niche, and you can monetise it effectively.
So what actually happened?
I got the site to number 1 in Google for the singular and plural forms of its keyword. So I should have been quids in, right? ……… WRONG
I had made the fundamental mistake of relying on what Google told me the search volume was. These keywords had almost no searches at all. I should have made a lovely £12,000+ per month, and in reality all I got was 1-2 clicks per month!
This is (unfortunately in my case) an extreme example, but my mistake was not creating a campaign in Adwords, and actually testing to see if it was worth me building a site around those keywords. This not only applies to micro sites but to SEO campaigns in general. I think as SEO’s we’re all too quick to rely on Google’s tools when we know that they’re not always as accurate as we’d like to believe they are. We should ALWAYS test keyword volume for clients if we’re planning on the type of campaign that goes after specific keywords.
One of my oldest and best friends is a very talented voice over artist based in London (yes I did mean to give him a nice anchor text citation – he’s really good at what he does – and deserves to rank well). We wondered if more people search for “voice over artist” or “voice over artists”. Google says that about double the amount of people search for the singular and that about half search for the plural. This is what Google’s tool says:
My friend (and I) thought that Google’s tool was wrong and that the plural would have the most searches. So this time we ran a campaign in Adwords. The results showed that Google was actually right. The singular voice over artist keyword had double the searches of the plural. Well done Google!
After 1 month:
Voice Over Artist – 576 searches (impressions) – Google estimated 390
Voice Over Artists – 241 searches (impressions) Google estimated 210
So we can see that in this case Google was pretty accurate with the ratios, although (for this month at least) there were more searches than expected.
Hopefully you can see the value of testing this before paying for a campaign going after specific keywords. This is even more the case if you decide to go about building a micro-site around a specific keyword set.
So, how do I test accurately in Adwords?
It’s pretty simple to test in Adwords, but there are lots of pitfalls to avoid that will give you inaccurate data if you miss them. I’ll show you here how to conclusively test real search volume using Adwords.
This may well be useful for SEO’s who are excellent at what they do – but they don’t get to play around with Adwords very much.
Firstly, hit the new campaign button, being sure to select “Search Network Only”
We don’t want that pesky display network messing up our data!
Make sure you select “All features”
Without “All Features” we won’t be able to set Adwords up with the correct settings to get the data we need.
Be sure to unselect “Search Partners”
We only want to know how often a keyword is searched in Google, not in every site that Google has a relationship with!
Select your desired location in “Location Options”
We only want to know how often our keyword is searched in our country (in most but not all cases). If you want to know how often a keyword is searched in multiple countries I’d suggest setting up a duplicate campaign for each country – although there are other ways to do it.
Hit the link for “Location options Advanced” and select “People in my targeted location”.
The more astute amongst you will realise that as we will only be using exact match keywords this particular setting change doesn’t really matter, but it’s always best to give the correct settings when possible.
Hit the link for “Delivery options advanced, and select “Accelerated: Show ads more quickly until budget is reached”.
This is one of the most important settings as we don’t want Google deciding when our ad shows. We want it showing 100% of the time so we can accurately know how often a keyword is searched.
Notice that Google gives a warning saying “You may miss traffic later in the day if you choose accelerated delivery. Standard delivery is recommended for most advertisers”?
This can safely be ignored, and in fact is often bad advice, but that’s a whole other blog post. I call that setting the “idiot tax”. If you are so limited by budget that you need to NOT show sometimes, then you are bidding too much for your keywords and need to reduce your bids. For this example we need to show all the time, so go with my recommended settings.
We’re now going to move on to “Advanced Settings”
In schedule – make sure that your ads are running all the time and it says “Show ads all days and hours”.
If Ads are only running at certain times of the day, the test won’t be accurate.
In Keyword matching options, select “Do not include close variants”.
This option is VITAL, so don’t miss it. We DON’T want to show for related keywords, only our chosen exact match keywords. This option will negate us using exact match keywords and will ruin our test, if we get it wrong.
So select “do not include” and click “save and continue”.
Now you’ll have to create an ad.
You have 2 options here. You could write a bad unrelated advert as you might not care about getting clicks – as we just want the impression data. I’d suggest that doing that is a bad idea. Write a great ad – you’ll get clicks and hopefully some sales. You can learn more than just about keywords from running Adwords!
Input your keywords below the ad, as I’ve shown
You MUST make sure that you use exact match keywords. This is done using square (not curly) brackets around the keyword that you want to test [done like this]. We don’t want to show for any variations on the keywords we’re testing.
We’re almost done now, we just need to make sure that we’re bidding enough that we’re always going to show for our chosen keywords. So create the adgroup and click on the keywords tab.
In the keywords tab, select columns > customize columns, as show below.
Click attributes and then add “Est. top page bid”
Knowing what we need to bid for each keyword to make sure we always show is important. It’s no good bidding enough to show ‘most’ of the time. So if we bid based on the estimate for a top page bid we should be in good shape.
Now we can see the estimated top bid for each keyword as shown below.
Adjust your keyword level bids to be the same as the estimated one. Again, the more astute would realise that we could automate this using automated rules, but this article is focused on keeping it as simple as possible.
Step 14. (bonus step)
You’re already done and steps 1-13 will give you the info you need. All you need to do is leave it a few weeks and you’ll see exactly how often your keywords are searched.
It’s important to also know “how” they are searched. It’s great info to know how many searches are from a computer or a tablet or a mobile. Here’s how you do it.
In the campaigns tab (you can however do this right down to the keyword level in the keywords tab) select “Segment” and then “Device”.
Now you’ll get the following data (I’ve blacked out the name of the clients’ campaign)
You can now see not only how many searches your campaigns or keywords get, but what devices are used. For example, are your keywords top heavy with mobile searches? Then maybe you need to look carefully at how your website converts on a mobile and how you can improve the mobile experience.
So there you go – that’s the complete guide to accurately finding out how many searches individual keywords have.
See you next time.
Latest posts by harveypearce (see all)
- Adwords Standard Vs All Features – Be Sure to Optimise Wisely - October 5, 2013
- Outsourcing SEO abroad – It May Just Cost You Your Business - July 2, 2013
- 5 Beginners Adwords Mistakes to Avoid - June 23, 2013