What to Measure? Only 16% of the Web Is Clicking Display Ads
But ComScore, Starcom Study Shows Banners Are Still Effective -- Especially When Paired With Paid Search
NEW YORK (AdAge.com) -- The number of people online who click display ads has dropped 50% in less than two years, and only 8% of internet users account for 85% of all clicks, according to the most recent "Natural Born Clickers" study from ComScore and media agency Starcom. As the pool of people who click on banner ads rapidly decreases, it begs the question: Is the long-used click-through rate now officially useless?
Clickers only represent 16% of U.S. internet users, according to ComScore data from March. The study initially found that 32% clicked on display advertising in July 2007. If that first study, released last year, crystallized skepticism that click-through rates weren't the be-all end-all success metric for display, this most recent report might just be the last nail in the digital coffin.
What's more, the 8% of internet users that compose a majority of clicks is also down by half from the last study, which found 16% are responsible for 80% of clicks. The 2008 study found half of all clicks come from lower-income young adults, so prizing clicks ignores the vast majority of internet users, especially the types of users many marketers want to reach. This year, the study focused more on alternative measurement, suggesting that a low number of clicks doesn't necessarily mean banners don't work, but that marketers are looking at the wrong success metrics.
From client studies, ComScore found that display ads, regardless of clicks, generate significant lift in brand-site visitation, trademark search (searching for, say, Toyota or Prius) and both online and offline sales among those exposed to the ads. Within one week, consumers exposed to a display ad were 65% more likely to visit the advertiser's site than users who never saw the ad. Even at four weeks, people exposed to displays ads are 45% more likely to visit the brand's site.
"The click has always been of dubious value," said Joshua Spanier, director of communication strategy at Goodby, Silverstein & Partners. "But clicks are easy to understand and easy to measure. We still know that display advertising has unequivocal value; your search performance improves as well. Together, search and display are much stronger than apart."
Banners lead to search
ComScore found that online users exposed to a particular brand's display ads conduct more searches on that brand's name than those in the control group. Still, the percentage who search is relatively small. When exposed to both paid search and display ads, consumers were found to be nearly twice as likely to make an online purchase on a retailer's site, which is greater than the sum of each ad tactic's individual effects.
Steve Kerho, senior VP-analytics, media and marketing optimization for digital agency Organic, has also seen a strong correlation between banner ads, higher interaction times on brand sites or branded microsites and better-performing paid search.
"Clicks are only one measure," Mr. Kerho said. "We look at how much an ad costs, how many people saw it and, more importantly, if two or three days later that person comes to our client's website. That ad is an impression. Go back and think of TV. When we put ads there, it's about carving out space in someone's mind. Banner ads create awareness and familiarity, too."
Mr. Kehro describes Organic's measurement model as considering banner ads and paid search in relation to site visits. Each site activity -- be it 10 minutes spent on a brand site or a visit to the store locator -- is assigned a dollar value. A site visitor's activities are then tracked back to see if there's been exposure to display ads or purchased keywords, and the agency can determine the value of those digital investments.
John Lowell, Starcom USA senior VP-director of research and analytics, points out that certain marketers may not have associated action with their display creative. But with more creative that doesn't demand clicks and few clickers online, it's time to look for new metrics online.
"For the most part, clients get it," said Mr. Lowell. "The problem is, what's the alterative? Clicks are easy to measure, so it's a good default. It does put the onus on 'What are we going to do instead?' like looking directly at sales. We look at clicks in a world of isolation. Consumers aren't only exposed to online ads. Digital has only been measured in the past on its own; increasingly you need to look at everything they are exposed to."
Tips for living in a post-click world:
1. Goodby's Mr. Spaniers points out that it's important for marketers to ask their agencies how display ads fit into the larger marketing plan. "With the media mix, how do you set up overall metrics?" he said. "It's unrealistic to look at one element of one media for success. Since online media is so trackable, it's is held to a higher standard. The reality is no one looks at print advertising like that. You can't measure print that way or TV because they don't have built-in capacity. It's a little unfair that display becomes the whipping boy."2. In the digital space, look at how elements like display and search coalesce, Mr. Spanier said. How do people end up at a website to make a purchase? Look at sales of a product on-site with consumers' exposure to ads. "There is no way display advertising is wasting money," he said, "even if people don't click."
3. Clicks are a direct-response measurement. For display campaigns, look at brand-awareness studies, purchase-intent lifts and engagement rates. "Users might work with an ad, but not click on it," said Paul Gunning, CEO of Tribal DDB Worldwide.
4. Look at display view-through rates. "CTRs have a value, there is still a hand being raised, but view-through is more important," Mr. Gunning said. "How many people actually end up where you wanted them to? How much time are they spending?"
5. Look at the creative. While consumer attitudes online have undoubtedly changed, think about how messaging does or doesn't demand clicks. If you want to measure success against click-throughs, include strong calls for action. Or, if you have a rich-media experience, other metrics such as engagement time are probably more telling. And that may be the case as more banners include innovations such as embedded video and rich media, said Atmosphere BBDO's Warren Griffiths, group account director for Visa, Emirates Airlines, Hertz and Starwood. "More banners have become the destination. I don't believe that would account for a 50% drop in clicks, but this may have had some impact."
RSS Feed
RSS (What is RSS?)
Not all that much. For the most part, Internet users are actively looking for content and it's this very active mode (as opposed to the TV viewer's passive mode) that makes them very adept at automatically NOT looking at the display ads.
On a scale of 1-100, I give a TV ad a 90 based on commercial ratings data, TV's sight/sound/motion and the passivity of the TV viewing experience.
On the same scale, I give an Internet banner a 1.
I'm not saying there aren't times when they should be used. But the CPM's one pays should reflect the extraordinarily low impact. It's the same reason that CPM's for 30 sheet billboards are very low - there is relatively low impact.
The more the Internet evolves, the more it becomes apparent that it is a great vehicle for content delivery, a great platform for commerce, but by and large, a very inferior vehicle for advertising.
Display ads are at the truly inferior end of the continuum. At the othet end of the Internet advertising impact continuum are Search and Pre-rolls. And Search is of course very dependent on product category.
Give me TV and radio any day.
Having said that one cannot rule out the persuasive power and pervasiveness of the medium. We need to understand on which of stages of the consumer - brand relationship journey; does Internet has a pre-eminent role and look at metrics that capture the same. Is it building awareness; or intent; or establishing dialogue; or actual purchase; advocacy..?
It works for certain categories well through the entire gamut (Amazon; ebay...) and for some lends a supporting role (where involvement of the consumer is high and purchase decision cycle longer).
Though Internet may not kill TV / Radio yet; but its time to acknowledge the potency of this teenager on the block.
Arun Vemuri
Scatterbrained Scribblings
http://ivak99.wordpress.com
Propagated by code-junkies at technology companies - it has practically no basis for understanding human physiology, consumer behaviour or advertising psychology.
It is time to find a metric that attempts to measure exposure, not 'a' response. What did the user see, then what did they do - did they click, did they search, did they walk into a shop and buy - but lets at least find a way to measure 'did they see'
The demise of click, the rise of Dwell
http://bit.ly/Kp0xL
http://twitter.com/vickie_smith
All display media should be priced and sold on a cost per engagement basis......the metrics are there - let's use them.
But what's an engagement I hear you say? It doesn't really matter - as long as it is agreed up front - it could change on a campaign by campaign basis: time spent, an install, a share, a video viewed etc..
Our format - the Open IMU - is priced this way. It's low risk to the publisher and the advertiser and generates far more cash. This puts the onus on creative to ah....be creative.
Display media should deliver a rich content experience at the first point of engagement...draw users in and then encourage repeat interaction.
Display media needs to be a window into content and a vehicle for transporting content everywhere. Remember 'web portals' that failed concept of 9 years ago. Display media are the new portals.
thisisopen.com
twitter.com/OPENtwit
Additionally, does it vary by type of content? Obviously someone using Rhapsody will have the speakers on. But what about the person reading a news story or blog? etc., etc.
Bring back Impression Time as a key metric as it is more in keeping with the way branding and advertising work together to get the message across. Longer Impression times have been clearly linked to increased buyer awareness and activity, so let's use it.
The newer social media have been difficult to measure using click stream metrics, but as soon as you make them Impression Time based, it all fits into place.
Long Live the Impression Minute!
No one has talked about what REALLY makes the internet valuable. Instead of casting a giant net over all mankind with traditional media, it allows super-specific targeting to your exact client based on location, behavior, frequency, etc. It's always about integration and frequency. If I see a 30-sheet, then hear a radio spot, then see a banner on my favorite site, that product will be top of mind since it followed me everywhere.
As far as metrics, I believe click-thru is good to measure how effective your creative and/or call to action is; the time spent and the pages visited after the click through tells you how compelling the presentation is of your product, and potentially leading to a sale. Ultimately we want them to buy a product, yes?
What is a Web content provider to do? If advertising online has little value, why the heck should anyone even try posting content?
Let's all go back to the days of print media when advertisers would buy on strong metrics. You know, like page size, circulation and whether or not to make the ad color?
That worked, right?
Clicks result in visits that are valauble.
Brand marketers need to view their sites as experiences that drive immediate and latent purchases. The real question is what is value of non transaction based visits.
But we work closely with our advertisers to present a sponsored environment, with high SOV and relevant visitor profiles. We allow advertisers to integrate useful content into our pages, and combine email, display, and other elements. These aren't $1 CPM banners. It's a partnership between web site and brand designed to provide visitors with helpful and relevant content, whether that's coming from ad banners or editorial. And it works.
But click-through rate? That's a benefit of being helpful, relevant, and/or entertaining. It doesn't happen by magic, and it doesn't happen by buying remnant inventory at $1 CPM from whatever random web site couldn't sell it otherwise.
One interesting metric I have not seen measured are the number of clicks that continue once a display ad campaign has ended and it dissapears from the website displaying the ad. It can be months before those click rates stop coming in. Those clicks never make it to the original advertiser's site because the link is no longer there. Generally, some other advertiser on the website is discovered to replace it.
And before anyone comments that newspaper and magazines are advertiser supported also, please shelf that thought for a moment because people are used to paying for them.
If content sites ask for minimum payments the way newspapers and magazines do, say, a subscription fee, are people on the willing to pay for that?
The web is free, the consumers out there believe. But it is NOT FREE as those of us who make a living on it know. We have to pay for server space, IT people, etc.
Pure play companies know that since the beginning and today, sites owned by media companies that has print are finding out that they need to generate income or else they will go on the wayside.
So, how are we going to sell to the consumers a paid site? Or better yet, how are we going to sell to advertisers when the click thru rates do not work anymore.
Or advertisers just see the low click thru rate as a way to save money. No click, no payment. In short, no advertisers, no subscribers, who will foot the bill?
For example, we have some clients who measure success via conversion rates, some by average order value, others by overall sales, etc. There are any number of success metrics that can tie back to banner ads.
One thing is for sure: if it's something worth tracking, it's worth testing. No need to wring your hands over click-through rates (if that's what constitutes successful banner ads for your organization) that are decreasing when you can avoid that scenario altogether! So plan your creative accordingly, test your hypotheses, and watch your success metrics increase.
Happy testing.
Some of them, for example Warren Lee, even publish here claiming that Ad Networks are doing well. Maybe they do, but for whom? Certainly not for the creators and publishers who want to monetize their content.
This means that when people are looking for products to buy, they see these full page spreads and because they look so good, the click through rate to the company's website is high.
In addition, by featuring insightful and funny commercials on Zovue.com, it keeps internet users engaged and the click through rate is higher than a standard banner ad that does nothing to grab the attention of the user.
Unless advertisers can find a way to make click banners more attention grabbing and useful, their appeal will continue to plummet.
Timothy Kelley
CEO The Baby CD
www.TheBabyCD.com
Yes, if you have products to sell online, you should display them online. The Internet is a truly amazing way to enact commerce on so many levels.
But if you are just advertising your brand, then generally Internet display ads are of almost no impact on consumers. Stop wasting you money and your company's marketing energies on this. Sorry.
You stated a fact that seeing a banner ad makes one 65% more likely to visit a website...what is the baseline???????? I mean if the baseline is 1% big whoops. If the baseline is 20% now we are talking.
As for me sitting here with my Firefox 3.5 with No Scripts and Ad Blocker Plus...what is a banner ad?
We've gone into detail on the merits and drawbacks of these metrics, and we'd love to hear your thoughts:
http://www.adify.com/kill-the-cpm-let-s-take-a-closer-look/
a bit more... http://www.connectedadvertising.com/blog/?p=103