Measured as a percentage, ‘click-through rate’ (CTR), provides an indication of the relative success of an online advertising campaign.
The number of times an advertisement has been displayed is referred to as the number of impressions.
The more successful the campaign, the higher the CTR. The highest possible click-through rate is 100%, in other words, a campaign where every person who sees the advertisement clicks through to the advertiser’s website.
The click-through rate is calculated by comparing the number of users who have clicked on a banner, against the number of times the advertisement has been displayed.
To calculate the click-through rate of a campaign you’ll need to know:
CTR = (clicks / impressions) × 100
For example, the CTR on a 120-impression campaign, where 12 people have clicked through to your website, is 10%:
(12 / 120 = 0.1) × 100 = 10%
Calculating the click-through rate as a percentage can be more useful than the ‘raw’ number of clicks when you’re monitoring the success of a campaign over time.
Assuming that a banner is shown on a blog, the number of people reading the blog generally increases from month-to-month.
Let’s say you’re experimenting with the position or design of your banner. Comparing the number of clicks for each month won’t help you decide whether your experiments have made the banner more effective. For that, you’ll need a percentage value.
|April||Added black border||250||20||8%|
|May||Moved to bottom right of the page||400||50||12.5%|
|June||Moved to top left of the page||720||60||8.3%|
|July||Moved (back) to bottom right of the page||1020||130||12.7%|
More sophisticated campaign management systems provide tools that allow you to experiment with banner ad placement, size, design, copy, etc. without having to ‘wait-and-see’ what impact the changes have made.
In ‘A-B testing’, multiple versions of the same webpage are displayed simultaneously. Each visitor to a website is shown a different variation of the page: the first visitor is shown variation [A], the second variation [B], the third variation [C] and so on, with the cycle looping for the n variations.
This form of testing can also be carried out in usability studies, to help fine tune an online product or service.