Forum Moderators: skibum
AdKnowledge Q1 2000 OAR [adknowledge.com]
Press Release (contains link to report in PDF)
"...only 14.3 percent of the
time is the ad with the highest click rate in a campaign also the ad with the
highest conversion rate."
Report: Little Correlation Between CTR and Conversion Rates [internetnews.com]
InternetNews Aug 17, 1999
Though these are older articles, they echo some of my ramblings about advertisers not "getting it." They are looking too hard for results/performance based advertising.
Another excerpt from the first article:
"...advertisers that focus only on clicks or even post-click conversions may miss vitally important effects of their advertising campaigns"
and from the second:
"This new data proves that Web advertisers who are just looking at click rates are missing the bigger picture."
I wish more mainstream focus could be put on this now, as for advertisers to come to grips with these points would surely help the rebound we are hoping for.
For 33% of all conversions to come from branding type ads, with NO clicks, speaks bounds about this type of advertising.
[internetnews.com...]
Though this is a closed study for one client, and actual impression/click numbers are not provided, the results tell.
"...Web surfers who either visited the client's site, signed up for e-mail messages, or made a purchase, most did so after seeing an online banner ad."
This study found the client had a 10% increase in total conversions. 80% of them came from ad views that were not clicked!!
CPM branding works. Just because you can't easily track it doesn't mean it can't perform.
Added - Another article (from last November) based on Engage's first quarter report from last year.
[internetnews.com...]
Branding results seem to typically take longer. 38% of conversions were generated from 8 to 30 days after viewing an ad, vs. 61% making a conversion within 30 minutes after a click.
"...online media has spurred online advertisers to demand more hard effectiveness data. That has generally meant cost-per-click or cost-per-acquisition deals, but this latest study gives supporters of the CPM model more ammunition."
These reports don't really go into detail there, but I am assuming that the "impression" left the .com name in the user's mind. Maybe they used a SE and clicked the listing that included the name that they had already seen from branding, or simply typed it into the address bar. The branding should cause them to choose the advertiser over other companies in comparison situations.
I have wondered how they tracked this kind of thing, and I think the only effective way is using cookies stored when the branding ad is viewed.
It seems to me that this kind of effect can only be true when the advertiser also has LOTS of other vehicles to put their URL in front of the prospect. And that means a rather BIG company with a very easy to remember domain name -- or at least someone with a somewhat dominant presence in their market niche.
I've been on the web for six years, and I can't remember ever going to a site by remembering their domain name from a banner ad. Heck, many banner ads don't even give the domain.
For instance, I never tried to "punch the monkey", and I have no idea right now who ran the ad -- it made a strong impression, but it didn't brand anyone for me. Now if I ran into a blurb somewhere that said "We are the punch-the-monkey people", that might get me to visit.
I see what you mean about needing an easily remembered domain, and advertisers simply shouting their company name. I think a lot of this action is based on the subconscious. IMO, they don't really think "Hey, I remember seeing a banner ad 350 times, this company must really be the right one for me." It is probably more of one that the company is subconsciously "hi-lighted" in comparison to other companies, or the company is subconsciously given more credibility due to previous impressions.
The punch-the-monkey is more of a CPC results banner. Though it is probably predominantly sold as CPM, its message is one of "CLICK HERE NOW." Not one of branding or a company message.
This illustrates the abuse (from a publisher's perspective) of CPC/CPM campaigns that advertisers are notorious for: They sell "trick" and highly-clickable banners as CPM, and push branding type banners for CPC, trying to leave the advertiser with less revenue for their results. More bang-for-their-buck, if you will.
I should be the case study for this. I rarely -actually, almost never- click the banner directly. I put the cursor on the banner, watch the status bar for the redirect url, and mentally log the eventual target. Some banners list the domain on their face as part of the branding -I like that, saves me the mouseover.
I developed this habit because A) I dislike being distracted from the primary page and B) I know that I'm disrupting the cookie trail if I key in the url directly.
Yep, you're pretty much describing what I think has become the average surfing mode now.
As for what this means for advertisers and publishers, I think we're going to have to bottom out on the current dotcom-bashing that's in vogue before we'll be allowed to find the middle ground on the real "value" of banners.
How do you get the domain name from such a banner in order to type it in later?
The target url, or some recognizable portion of it, is often in the mid-section of the string, after all the redirect mumbo-jumbo. NS4 will pick up the url of the banner with the right button (copy link drop-down), I sometimes drop long ones in a text work area to decipher them in bulk later or drop them in the location bar of a second open browser and view/trim the url.
added:
I'm a confessed status bar junkie. You can learn so much about who-is-doing-what-how by watching the page load there. It's how I first found Flycast and CJ.com, for instance.
That's what I never thought to try. Thanks.
However, back to the original study that started this thread, I don't see this as a common way for the average surfer to "eventually" find their way to the sponsor site of a banner ad.
Some other factor must be at work -- the back of the mind impression that Drastic mentioned, perhaps. This is the MO for branding campaigns, anyway -- build up a critical mass of exposure to the brand, always consistently presented. Then, at some point, the prospect converts.
I'd love to know more about the specifics of the research reported on in the article above; so much depends on the details.
I'm suspicious of branding campaigns. They seem like such a potential black hole for funds. By the time the results are seen as less than effective, the money is spent! More direct marketing methods allow for much better control.
In the bricks world, a branding mindset almost killed Levi-Strauss. I think that branding is mostly good for the advertising house -- it sells more spots -- and that's a lot of why it gets pushed.
* include the name and/or URL whenever possible, possibly in a fixed corner location, preferrably in every frame (you never know which frame someone might see, and you reach people you otherwise wouldn't).
* make sure the first frame can stand on its own (allowing people surfing with animations off to still get enough information to act on).
My humble opinion is that many animated banners get the last frame correct, but have trouble with the first and middle ones.
p.s. Tedster, you might find Nick Usborne's article [clickz.com] on banner response/awareness a interesting read.
