Search volume is a metric that refers to how many people are searching for a particular keyword for a set time frame, more often, per month. Google’s Keyword Planner provides this information and is often the go-to tool since it is a free tool that can be used for keyword research and has the data for search volume.
Before this test, there was a story released that the previously free Keyword Planner data from AdWords has been scaled back to only provide “accurate” data for accounts with active campaigns running. There was a whole debacle over it which prompted us to test the accuracy of the Keyword Planner data, as we had always thought it grossly underestimated search volume and in some cases, had no connection with reality at all.
The accuracy of the search volume metrics in Keyword planner is important not only for SEOs but also for business owners, site webmasters, content writers, etc., because we rely on this data for deciding on which keywords to target and use in our content, and which to optimize for in order to get that rank that would bring in visitors to our site.
SEOs also include keyword research sheets that estimates on what kind of traffic increases clients can expect based on achieving specific ranking goals. They may even go as far as comparing estimated search volume and CPCs from Adwords against estimated organic search traffic and in some cases, recommend that the client is better off just using Adwords.
So Imagine if all these projections we are basing off from Adwords are inaccurate?
In this this test, we see if the Keyword Planner is really underestimating search volume, if there is a pattern for it or if it is just random, and to see if the alert in Keyword Planner about having an active campaign in order to get accurate search data was correct.
The only way to accurately test this is by using Exact Match only. We tested keywords ranking in search volume from around 2,400 searches per month down to 1 search per month for our test.
What we found blew our minds. Contrary to what we thought, Keyword Planner actually overestimated the average monthly search volume that we extrapolated from 12 months of our own Adwords data.
To review this data, we divided it into 5 categories to look for patterns that might emerge. Categories ranged from 3000 – 2000 searches per month down to less than 100 per month. Results showed that Keyword Planner overestimated our own actual search volume data usually by between about 15 – 30% in each bracket.
In the bracket with the second most data, 500 – 200 searches per month, the Keyword Planner estimated an average of 18% more searches than our own data actually showed.
In the category of 100 searches a month or less it is clear that your own Adwords testing is going to be more useful than anything provided by the Keyword Planner. Adwords just can’t accurately estimate on low volume searches. In this category, our own data registered 50% more searches than the Keyword Planner.
This means long tail could be worth a LOT more than we thought
On the whole, we think this should be the main takeaway from this experiment, the overall difference between our own data and the Keyword Planner was only about 10%, which is generally considered an acceptable tolerance.
When estimating data for clients, we have started giving estimates that are +/- 20%. For example, if the keyword planner gives an estimate of 1000, we tell clients the actual search volume is probably somewhere between 800-1200.
When the planner gives estimates of under 200 we advise clients that there really isn’t any way to know if those estimates are accurate without running a test for at least 3-4 months.
In this video, Clint discusses this test and his opinion on Adwords’ Keyword Planner.
This is test number 54 – Search Volume: Keyword Planner Vs Reality
This is actually a pretty old test and it’s born out of something that I said and proved a long time ago and it’s that the search volume versus of what’s reported here, regardless of the tools – Keyword Planner, Ahrefs, SEMrush, Majestic, whatever tool provides Search Volume data is not based on reality. It’s based on averages, right? So this is the average that we get for this keyword, probably phrase matches kind of added up, etc. and oftentimes, it’ll do one or two things. One ‘ll say there’s absolutely no traffic there and you’re wasting your time, and the other one will say – oh, there’s way more traffic than what actually is there.
The biggest example that I’ve ever given is the keyword “ductless heat pumps”. Ductless heat pumps at the time that I did this test was 14,000 keyword search volume a month and the client I had was actually pretty excited about that because she obviously sold ductless heat pumps. It was a husband and wife team that owned this business in Olympia, Washington. They service the Thurston County, pretty much, they didn’t want to go up in Pierce. Well they did a little bit of Pierrce County, but they didn’t want to go into King and all that. If you know the area of King’s counties, there’s more people there but it’s farther away from them and farther out of their service area.
So they wanted to run a search campaign targeting ductless heat pumps, obviously, I could rank for the term. But what I wanted to know was, was it worthwhile because it has 14,000 search volume, as you can imagine the competition from an SEO perspective is pretty high, you’re going to invest a lot of money into this, and do we really want to go down that road?
So set up a PPC campaign and I got their service area, this is the only place that we wanted to install these machines and I set that as the target area in the ads campaign. And as a result, ductless heat pumps got zero impressions. We actually set the bid up to be the number one ranking ad, we’re willing to pay that much for the click and met that much. Plus, the profit on one of these units is significant. So if we pay 50 to 75, maybe even $100 a click and it converts, we do that four or five times a month, we’re fine. The ads are paying for themselves.
Well, that exact term, zero impressions, zero clicks, over the course of a month. And so you look at it, 14,000 search volume, it’s exciting, we’re going to rank for it, I want to be ranking for it. That’s the word I’m looking for. But no one in that search area is actually doing it, no one is looking for it, especially in your service area. So you’re investing all this money to rank for this, what I call now dream keywords, and it turns out that no one that you’re targeting even cares about it. So it’s kind of a know your customer and know how they’ve identified that problem and how do you put your solution in front of it?
In this case, it was ductless heating and cooling units that was actually getting all the search volume and that’s where we got all the clicks and all the conversions. What’s also really cool is, it was way easier to rank for that because nobody had that data and that information.
So this test kind of back that up, they took one year of keyword data, 12 different accounts, did search impressions and figured it out. As a whole, Keyword Planner actually overestimates and sometimes it could underestimate. So you heard my story about the overestimating and the under estimating is there are just some markets that Google doesn’t want to share that information. And in particular, if you’re in the rehabilitation market, you’ll know that for a fact 99% of the keywords in their search tools, I’ll tell you that there’s no traffic there. But when you get into ads, if you’re lucky enough to qualify for it to be able to run ads, or more specifically, using something like Search Console and reverse engineering your traffic with a search console Google Analytics combinations, you can find that all those search volumes that say zero in the tools actually have a lot of traffic.
So what is the more or less? One, search volume is a great filter and at the end of the day, that’s really all it is. If you see high search volumes, you understand the competition’s going to be higher. If it’s lower, you have the potential of less competition and easier conversions because you’re solving that problem for the customer in an easy and fast way, right?
Two, you should be doing some AdWords campaigns, you don’t have to do a whole lot of these. You don’t have to spend a whole lot of money 100 bucks, 150 bucks, depending on the market, you only got to run it for a week, find out what the impressions are for the terms, multiply by four and then you get your average for the month. And you’re going to be just fine. That obviously won’t work if you’re doing seasonal stuff. If you’re doing gardening and you serve people up in the northeast, you’re probably you’re going to have less search volume in the winter, then you are in a summer. But you already know that.
So you run one week during the winter, one week during summer, figure out your averages, and then you can figure out where you want to target your marketing dollar during that period of time. That’s the only exception to this, right.
Use us search volume as a as a tool. Figure it out, identify your keyword, your keyword clusters using search volume, and then run a cheap PPC campaign, even if you don’t get any clicks, it doesn’t matter, you’re after the impressions number. And once you have that impressions number, you have actual true search volume or even a truer search volume, trusting that Google’s not lying on the impressions count and then you can turn that into a good SEO campaign that is backed by data.
Run a couple of these campaigns by yourself plus you get to learn a bit about Google Ads at the same time and it kind of saves you up. Don’t go in there thinking – Oh I am here to collect traffic, I’m going to win. All you want is the impressions stuff and you can learn about keyword selection etc, setting these up. You won’t even have to pay. You can do this, it’s cheap data and it can save you a whole lot of time and the customer’s money going after keywords that are just aren’t there for you.