Is CTR a Useless Metric for Bloggers? The Pros and Cons

I’ve been seeing a lot of discussion lately about the click-through rate metric and how it may or may not be as useful as everyone believes. I think there’s an important discussion to be had here, but it’s important to acknowledge that CTR is a very contextual metric, and it’s important to use it in the right ways if you’re going to use it. So, let’s dig in and talk about SEO, concepts, KPIs, the danger of optimizing for the wrong metrics, and how to use CTR properly.

What is CTR and Why Do So Many People Love It?

First, let’s start at the beginning: what even is CTR, and why is it so frequently used throughout marketing and SEO?

CTR is the click-through rate, and it’s a simple percentage of people who see your content – a link in another post, a paid ad, a search result – who click through to it.

If I show 100 people this link and 50 of them click it, that’s a 50% CTR. Simple, right?

Sure! That’s a big part of why CTR is a commonly used and commonly cited metric in marketing. It’s very easy to record and harvest. You can see quite easily how many people view your paid ads, how many impressions your content gets on Google’s search results pages, and so on.

CTR is also a compelling metric to try to optimize for because it has direct results. If I make a change to the link above – like, for example, making the anchor text into something meaningful so you have literally any idea what the destination is about – I would be able to see a direct and attributable increase in the CTR for that link.

All of this seems fine and dandy, right? So, what’s the problem?

There are actually quite a few issues with CTR, so let’s talk about them.

The Two Types of CTR

The first issue is that there are, in a sense, two types of CTR.

The first kind of CTR is narrow CTR. Narrow CTR is the CTR of one specific entity in one specific situation. One single PPC ad run for a selection of keywords, with a single destination and a single target audience, will have a narrow CTR. You can make changes to the ad, see them reflected in the CTR of that ad, and make estimates on the performance of the ad campaign based on that metric.

There are problems with CTR in this context, but I’ll get to those in the next section.

The second kind of CTR is broad CTR. This is the kind of CTR you generally see reflected in Google Analytics, Search Console, and other reporting platforms. It’s an overall CTR that encompasses your entire site.

This kind of CTR covers all of the impressions your site is getting and all of the clicks it gets from those impressions. Seems fine, right?

Ask yourself this: how many keywords can be used to find your content?

I don’t mean how many keywords you rank for. I mean, how many, out of every possible combination of words and letters and numbers and characters and search parameters, can your content show up in the search results for, anywhere, at all?

The answer is a lot. Any page on your site can be found in the search results for thousands, tens of thousands of keywords, if not more. Most of those are only going to be on page eight, page nine, and page ten. Nowhere near where people are actually visiting and clicking, not unless they’re immensely desperate.

Desperate or, more likely, harvesting data. Think about the kinds of users that show up on the latter pages of Google’s search. You have people harvesting data for case studies. You have search scrapers and data harvesters accumulating topic data (like Topicfinder, in that sense.) You have rank-checking bots that are searching for a million different keywords an hour and recording the pages that show up in the results all the way from page one to page five or ten.

The problem is, Google isn’t filtering any of these, not really. Some they will, but a lot of it they rely on things like their anti-bot captchas to stop, not data filtering in their analytics platform. On top of that, there’s no way for you to filter the bot traffic from your search console metrics, even if you wanted to. Google won’t even give you the data to try to do it yourself, which is why their CTR metrics are aggregates.

90% or more of most website impressions – not traffic, but impressions in search results – are coming from, essentially, nothing. It artificially suppresses CTR. So many of the impressions any site gets are not just not clicking; they won’t ever have a possibility of clicking, and all they do is drop your CTR lower and lower.

When people talk about CTR being a bad metric, they’re generally talking about one of two things. The first is that CTR, as a broad, aggregate metric, is effectively meaningless. The second is that CTR can have a lot of more specific issues, even narrow CTR.

The Problems with CTR as a Metric

Now, let’s talk about some of the problems CTR has as a metric for use in marketing, optimization, and iterative testing. There are quite a few of them, so I’m only covering the most important ones here.

CTR needs to be taken into context with raw numbers

Here’s an example.

Compare these two campaigns, A and B. They both run for 30 days, and they both have the same budget of $1,000.

• Campaign A has a CTR of 1.4%.
• Campaign B has a CTR of 3.9%.

Seems like campaign B is the winner until you look at the rest of the picture.

Campaign A has 10,000 clicks, while Campaign B has 700 clicks. Campaign A has 712,000 impressions, while Campaign B only has 18,000.

From pure CTR, B has the advantage, but from raw numbers, A is immensely better. And really, most sites would prefer 10,000 clicks over 700 clicks, all else being equal.

Since CTR is a percentage metric, it’s only part of the whole picture. Without knowing what the scale of the numbers involved is, you can’t tell if a CTR is good or bad. A 50% CTR sounds amazing, but when it means you had two people see your link and one clicked it, that’s nothing.

CTR doesn’t consider the quality of a click

One thing I talk a lot about on my blogs and with my clients is user search intent, conversion rate optimization, and all of the related metrics.

Why did a user click the link? Did they think you answered their question, only to realize you didn’t and leave? Did they mean to click a result below yours but got distracted and clicked the wrong one? Do they habitually tap their fingers and accidentally clicked when they bumped their phone? Was your link too close to another link or another element they wanted to click, and they clicked the wrong thing?

There are all sorts of ways to get a higher CTR that don’t involve more satisfied users or better traffic. A high CTR but a near-100% bounce rate isn’t exactly a good thing, right?

That’s why so much of my content and my marketing efforts are focused not on CTR or raw clicks but on the quality of the clicks you get and the conversion of those clicks into some tangible action, like sign-ups or registrations.

CTR is easily abused in both directions

CTR is not a reliable metric for marketers who want to use it to prove their efforts.

Let’s say I want to improve the CTR of a site. What do I do? Easy, there are thousands of ways to buy clicks in varying degrees of legitimacy. Cloaking links, bot clicks, Fiverr clicks, whatever. If each click I pay for is one impression and one click, the CTR goes up because there are no excess impressions.

The opposite can also be true. What if I want to make your site’s CTR look worse? All I need to do is get people to view search results pages or paid ad displays without clicking on anything. Paying thousands of people to just look at a page for a second or two is absolutely trivial.

I even mentioned above how this is broadly done completely unintentionally. I wouldn’t even need to pay some scammer to drive up clickless impressions; I’d just need to run a few rank trackers on keywords they’re unlikely to have harvested recently, scrapers from a rotating proxy list, or anything else of the sort.

Some beneficial actions can hurt CTR

A lot of actions you might want to take to improve your site as a whole can hurt your CTR, both in the short term and in the long term.

For example, aggregate CTR is hurt each time you publish a new piece of content. A new piece of content gets indexed but doesn’t necessarily rank well. Whether it’s Google’s sandboxing, a bunch of new keywords that only land on page 3+, or some other reason, it fluffs up the number of impressions coming from those rank trackers and other sources. Eventually, that content can rank better, and the CTR can climb again, but each time you publish new content, it dips back down.

Another example is writing content with a broader keyword appeal. Not every keyword you pick is going to be a winner, and Google will also make plenty of assumptions about synonyms and related keywords that may not be accurate. All of those impressions might not be aligned with your actual goals, so CTR can drop because of it, even if the overall broader exposure is a good thing.

When CTR is a Valid Metric to Use

CTR can be a valuable metric in some cases. Specifically, two cases.

The first is for narrowly defined paid ad campaigns. These are cases where you generally aren’t getting as much in the way of bot-based and other wasteful impressions (largely because ad platforms like Google will more actively filter it to avoid complaints from people paying for impressions that aren’t worthy of anything). They’re cases where you can see narrow and direct cause-and-effect between the changes you make and the impact on CTR, as well.

You can also much more easily map clicks to conversions, as well as get more data about overall audience pools and other data that puts CTR into context. It’s still not a terribly meaningful metric in a vacuum, but you won’t have it solely in a vacuum in these instances.

The second is as a comparative metric. Most of us are all laboring in the same conditions out here, so when you compare your CTR with a similar site’s metrics, it can be useful to see what the differences are.

It’s also useful as a comparative metric over time. If you have a bunch of content at a specific CTR, and out of the blue, something changes, and it shoots up – or down – you can investigate and see what’s happened. More importantly, it can prompt you to either avoid or replicate the effects elsewhere.

Other Metrics to Use Instead of (or Alongside) CTR

So, if you want useful metrics to use instead of CTR – or to put context into CTR – what can you use?

Here are a few options.

• Return on ad spend. If you’re trying to analyze the value of paid ads, nothing is better than the returns you get from them. This can be harder to quantify if you’re paying for visibility or other non-sale results, but you can still add a value number to something like a newsletter sign-up or even just an impression based on past data and calculate ROAS instead.
• Raw numbers. Tracking simple things like impressions is important to put CTR into context, and tracking the clicks you do get and figuring out their overall quality is also important.
• Time on site. The longer users spend on your site, the better off you generally are. As long as your site isn’t made up of one-page wonders (which can be unavoidable in some cases, like with food blogs where a recipe is the goal and no other pages are enticing), it’s useful to encourage users to stick around.

Overall, I think I broadly disagree with people who say CTR is a worthless metric, but I’m in full agreement that it’s not nearly as valuable as a lot of people act like it is. What do you think?

Leave a Comment