Why click-through rate shouldn’t be in search advertising reports

Yesterday I discussed what to say when someone who doesn’t really understand search advertising starts querying the relevance of keywords. Once you’ve put them straight on that aspect, focus inevitably moves on from keywords to the adverts – which to be fair, is where it should be. But once again, the first thing many people jump on is very misleading.

Here’s conversation number two.

Marketing manager: These are the ads we’re running at the moment.
Sales manager (or MD, or anyone else): The click-through rate is much higher on the ads for blue widgets than it is for red widgets.
Marketing manager: Well, that’s because…
Sales manager: …and look! The click-through rate is falling! That’s not very good, is it?

The principle you need to get over here is that click-through rate is not a measurement of success. It’s a piece of data which is only relevant when comparing two adverts which are identical in all external respects: the searches for which they’re showing, the position on the page they’re appearing, and the times when they’re running.

Otherwise, you’re comparing oranges with apples. Obviously, an advert for blue widgets, appearing at the top of the page for searches on ‘blue widgets’, is going to get a better click-through rate than an advert for blue widgets, appearing at the bottom of the page for searches on ‘red widgets’. But so what? This is all simply meaningless out of context.

The problem is compounded by some advertising agencies, who insist on reporting to clients on click-through rates (and trends in click-through rates) as if these numbers meant something. These people either don’t understand what they’re doing, or are just preying on their clients’ ignorance.

The easiest way to explain this to your colleagues is to point out that an advert will get a different click-through rate, depending on where it appears on the page. Everyone can understand that.

Then you can explain that it’s pointless to compare the click-through rates for the same adverts run against different searches, because the adverts will inevitably appear in a different position. You can explain that that it’s pointless to compare the click-through rates for the same advert from month to month, because it will also appear in a different position (simply because the competition will change).

When they ask what on earth the click-through rate metric is for, you can explain that it allows you to compare different adverts, run simultaneously for the same search. It’s called A-B testing. This is how we steadily improve the effectiveness of our adverts, based on data rather than intuition.

And that’s click-through rate: data for helping with development, not for reporting on results.

Discussion

  1. Andrew Walker

    Great points Chris,

    The steer for PPC and social media advertising should be towards conversions. At the end of the day, that’s what it’s all about.

Leave a Reply

Your email address will not be published. Required fields are marked *