Measure CTR per search term, not site wide

Usually when I hear from customers that their Click Through Rate (CTR) dropped, they equate that to traffic drops. But if you're looking at the site wide CTR, it's likely misleading. I would only recommend looking at CTR for a specific query rather than site wide or a specific URL.

For example, let's say you rank #1 for your main query and get 50% CTR. Then you start to rank more broadly for new queries. Those new queries may rank a bit lower because they are new or less relevant. Your clicks and impressions will improve because you have more queries that you're ranking for, but your average CTR will drop.

Let's dig into this more in detail.

If you rank for a term that has 100 impressions per month and of that, 50 click through to your site. You have a 50% CTR.

Now let's say you rank for a new term that has 900 impressions per month and of that, 50 click through to your site. That new term has a CTR of 5% and your Average CTR is now 10%.

So that extra term has dropped your Average CTR from 50% down to 10% for your site.

That sucks right? Wrong!

You just doubled the traffic to your website, from 50 to 100.

Sure, your Average CTR dropped from the new term, but in the end, Click Through Rates don't pay the bills.

Now you have a large opportunity to improve the CTR for your new term which you didn't have before.

Looking at this example site from Search Console, we can see their Average CTR dropped from 1.8% to 1.3% year over year.

One would think it's time to panic, but as we've just covered, the Average CTR is very misleading.

If we take a closer look, we see:

• an increase in total impressions from 11.4M to 28.5M,
• total clicks increased from 209K to 366K, and
• their average position increased from 14 to 11.9.

In this case, the merchant gained more visibility which led to more clicks (traffic) based on a broader set of search queries.

So as you can see, using Average CTR at the site level isn't a great metric at all, let alone to measure traffic.

Now if the CTR based on a specific search term drops and you see other signals like drops in impressions, then you may investigate potential causes.