Friday, 17 July 2009

What is the average click through rate for online banner ads?

Well it depends...

Those last three words are perhaps one of the most unhelpful answers you can get.

The reason for this post is that I've just been asked this question by a colleague in sales.

Now usually I’d fob off my sales colleague and say that they shouldn't be selling on click through rate (CTR), they should sell on who will see the advert, the demographic of our web users, the power of our brand - after all an ad that doesn't get clicked on doesn't mean that it hasn't been seen - BUT this time I thought, 'Why not? There must be some figures out there'. So here goes:
2%2% - Off the top of my head I’d say that a 2% CTR would be particularly successful. I’ve seen ‘averages’ suggested at conferences, seminars and by friends of anything from 0.1% to 1% but not usually any more.
0.1% to 0.37%0.1 - 0.37% - According to this post that quotes a study by MarkingSherpa with CTRs by banner size - I'm not sure of the B2B relevance of this though.
2.8%2.8% - This post quotes the same MarketingSherpa study however states a "conversion rate" of 2.8%. So what this means is with a CTR of 1%, you'd need 3571 people clicking till you made a sale (or whatever your conversion is).
So with this in mind, let's go back to, "Well it depends..."

What else matters for banner ads?
  • Branding – banner ads are like TV ads – there as much about branding as they are about click through so that’s an unseen metric that should be sold to any customers.
    • Association - By advertising with you they are receiving association with your website's brand and the prominence that provides. i.e. Seeing a brand advert on the ft.com will probably be enough to convince you the advertiser is not a dodgy spammer. The same advert on a blog such as this one, may not have the same effect.
    • Demographics - The demographic of your site users are also important to this branding exercise. If you have statistics on your visitors you should sell this to the advertiser. This is how facebook advertising works.
    • Page views should be as important a metric to quote as click-throughs. Page views are the only way to measure ad views. i.e. An advert has had 300 clicks, but if it's appeared on a page with 30,000 page views it's likely been seen and not clicked many times.
Top of Times Online home page with banner ad

Home page of the Times Online

  • Targeting– showing an ad for a video game company on the video game section of a technology website will get a higher click through than putting it in an irrelevant section.
  • Call to action – asking the user to do something increases click through – most banner ads seem to just consist of a logo and click here with no reason why you should bother. i.e. Win a prize, get a discount, find something that no other website has and that you need etc.
  • Ads designed for site – This is really just targeting again. The more personalized you can make your advert the better. It's even worth getting a screenshot of the website the advert is going on so you can mock up what it will look like when it's live.
  • Testing – running multiple ad variations in different slots and optimizing based on their performance is a sure fire way to increase click through. The more you test, the better results you can get.
    Google Adword example
    A text ad from Google Adwords

  • Positioning and size – an ad in the middle of the page gets higher click through than one at the bottom. At the top of the page, bigger ads perform better usually. Again, you can test this.
  • Text vs. image – there’s a load of discussion online about which is better – text can be better when subtly put in with your main content. Again test!
Related posts
Stop pop-ups popping up, online advertising needs to evolve (Nov 2007)

3 comments:

web design India said...

thanks u r information

reactorr said...

wikipedia suggests about .23%
sounds about right these days

school ball photographer said...

Thank you. I have been looking for this info all over the web four the last 15 minutes. Very informative. Thank you