Bigger Isn't Always Better

Bigger is better is the zeitgeist when it comes to display advertising. Efforts by large Web publishers to experiment with larger display formats are in full swing as they strive to grab more brand advertising dollars from TV and traditional media.

But does size really matter? In an attempt to answer that question, Web measurement firm Compete recently conducted a study that compared Kohl's ads of different standard sizes that ran April 24 on the home pages of MSN and AOL. MSN ran the familiar 250 x 300 medium rectangle, and AOL ran a 300 x 600 half-page skyscraper.

Both ran on the right side of the page and featured the same creative showing a reduced-price blender and storewide touting of savings and free shipping. As might be expected, the more attention-grabbing skyscraper ad drew 30% more click-throughs per session (1.5% vs. 1.2%).

But the smaller unit was more effective in driving people toward sales -- with 8.3% of those who clicked through to Kohls.com filling up a shopping cart, compared to 6.6% for the larger ad -- a 26% better conversion rate. Consumers who clicked on the smaller ad also averaged more time on the landing page (8.9 minutes versus 7.5 minutes).

Kohls Banner Ad Comparison chart

"In other words, the larger ad drew the attention of more would-be shoppers, but the smaller ad found a greater concentration of motivated buyers," wrote Compete analyst Alex Patriquin in a post on the company's blog.

Women and consumers with higher incomes were more responsive to both ad units, and each led to people doing same-day online searches for "Kohl's" at the same rate. Compared to a control group of people who had not seen the ads, however, the smaller unit actually delivered higher lift in search activity for the brand.

The humble rectangle got a boost from a separate, more extensive study last month by social media ad company Lotame that focused on engagement. It found that 300 x 250 ads averaged 13 seconds of "viewing" exposure per user served, compared to only 5.4 seconds for leaderboards and 1.9 seconds for skyscrapers.

Compete's takeaway from its ad test was that marketers should look beyond a single metric such as a conversion to grasp the full effect of ad campaigns, and that ads with lower conversion rates can still help build brand awareness.

But this looks more like another upset victory for the rank-and-file rectangle, delivering a better conversion rate and holding its own in brand lift metrics. Maybe it's time for publishers to start thinking inside the box again.

Tags: banners, display
Recommend (22)
7 comments about "Bigger Isn't Always Better".
  1. Jim Burnette from FreeAllMusic.com , May 18, 2009 at 10:11 a.m.

    Nice study, but how about clarifying the different user base on AOL vs. MSN. How about traffic levels, lead stories, etc. on April 24th. I hear the online community complain about research all the time. The research metrics provided for this story are lacking to say the least.

    Jim Burnette
    jburnette@optonline.net

  2. Jon Levy from Hype Circle , May 18, 2009 at 12:35 p.m.

    All other factors being equal (per Jim Burnette's point), this puts some weight on David Koretz's article, posted on Online Publishing Insider May 7, "There Is No Excess of Inventory, There Is A Shortage Of Intent".

    While a huge banner gets you noticed, it does not qualify the intent. If the goal is to create brand awareness, perhaps bigger is better. If the only goal is to produce an e-commerce transaction, then ROI might be better served by investing in targeting.

  3. David Kopp from Healthline , May 18, 2009 at 2:26 p.m.

    I have to agree with Jim B - audience and context "quality" is a big factor that is un-controlled in this test.

  4. Phil Guest from PeerIndex , May 18, 2009 at 3:42 p.m.

    Is this the best our industry can offer, if is doesn't work make it a bigger!

    All this effort to shore up a diminishing ad format when more and more services are offering engagement platforms based on social media.

    Time for some fresh thinking me thinks.

  5. Joshua Rex from This is Open , May 19, 2009 at 4:59 a.m.

    What a lazy study - verging on dangerous. And what does it tell us - f*uck all really.

    The inherent problem with display is rooted in 2 issues:

    1) The formats: 'print ads online' - simply not built for the medium

    2) The pricing models: 'print pricing online' - circulation morphed into impressions.

    I have to almost laugh when pundits suggest the solution is to make the ads bigger. Of course bigger isn't better. Aside to this, let's try and remain somewhat pragmatic. Web publishers would need to completely reinvent how their sites and structured and bodies such as the IAB would have to get with it and put their weight behind new formats. Whilst I totally welcome this, and it will happen eventually - it aint happening any time soon. They still put a 40k limit of most rich media!

    I think one of the possible solutions, as pointed out in the closing lines of the article, lies with 'thinking inside the box'.

    Here at Open we have created a number of new formats - within existing formats. The biggest demand is for our IMU format. Served as an MPU - that's where the similarities begin and end. The IMU format delivers multiple utilities, goes full screen, is socially portable, optimises natural and is completely dynamic.

    I think advertisers need to stop using display inventory in an effort to drive traffic - click through completely suck. As an industry we reward campaigns that drive a 1% CTR.

    Display media should be seen as an opportunity to engage with an audience in an ongoing manner - not a slim chance to drive traffic. People have multiple digital locations and your content - wrapped in utility - should be made available everywhere appropriate.

    Publishers need to lead the charge in the development of new formats. This includes new pricing models. Forget about diminishing CPMs. That's a down spiral and with new technologies emerging daily that allow better targeting at a fraction of the cost, it's a battle you'll never win. Change the game and offer performance-based pricing models to your advertisers. You know it makes sense.

    Interesting times ahead as online advertising comes of age.

    http://thisisopen.com/blog

  6. Alex Patriquin from Compete , May 19, 2009 at 1:14 p.m.

    Jim - user bases at AOL and MSN are nearly identical. We showed no significant different between exposed and control homepage visitors in terms of demographics and other factors like search activity. Both ads ran on the same day and were placed next to rotating headlines. This is actually not the full extent of our study - we only included a slice on our blog.

    To obtain a free copy of the full study, please email me at apatriquin@compete.com or visit compete.com/adimpact.

    Jon - great point! At Compete, we measure the intent of millions of consumers by measuring their exposure to online advertising and subsequent behaviors, from conversion to search to research at 3rd party sites.

    Phil/Joshua - I think we agree on this point. We're not saying bigger is better every time... or even in this Khol's case. We've developed Compete Ad Impact so advertisers can triangulate metrics with campaign goals.

    Check out Compete's blog series for more on this point - blog.compete.com/adimpact

  7. Ben Johnson from Interactive Internet Websites, Inc. , May 19, 2009 at 6:56 p.m.

    With data from two different user groups, I don't see how you can draw meaningful conclusions about the conversion rates. It could be that MSN users were more inclined to purchase for reasons that are unrelated to the size of the ad that attracted them.