Google: Click-Through Rates Fell in 2010

Google: Click-Through Rates Fell in 2010

Although banner ads may be experiencing big growth this year, they’re not getting any more effective. The average click-though rate for banners fell in 2010, according to a survey by Google.

Google’s report, commonly regarded as the industry standard, pegged 2010 U.S. click-through rates (CTR) at 0.09% compared to 0.1% in 2009. That means about one person in 1,000 actually clicks on banner ads, despite attempts to make them more inviting.

However, the Google report, which comes out this week, found that the format of a display ad can make a difference.

A 250×250 pixel ad using Flash got the highest CTR of any format — 0.26%. The worst performers were vertical 120×240 banners with Flash and a full (468×60) banner with Flash, which both got rates of 0.05%.

The study also found that most banner ads (54%) served through DoubleClick used Flash, while 28% were image-based ads. Ads using Flash for auto advertising had the best CTRs at 0.13%. Half-page ads also had the best CTRs overall at 0.18%. (See charts below.)

Shaun Herrmann, account supervisor for ad agency Neathawk Dubuque & Packett, says the CTRs quoted in the report are on par with what she’s seeing, although CTRs are generally lower for younger consumers. Herrmann points out, though, that CTRs aren’t the only measure of an ad’s performance. “You can send a message in a banner ad without having someone click through,” she says. “Click-through-rates are just one way to measure performance.”

Media Type

click through

Rich media ads without video were found to have the highest CTRs. Video- and image-based ads were even.

Ad Type by Impressions

google impressions1

Advertisers are really embracing Flash ads. Some 54% of impressions in 2010 used them.

Size Matters

google size

The report found the size of your banner ad has a lot of bearing on its click-through rate. The most clickable format? Square 250×250 pixel ads with Flash.

Source

Related Posts:

 

Comments are closed.