Re: Respectable Click-through Rate?

From: Michael Martinez <Michael_at_xenite.org>
Date: Tue 26 Sep 2000 19:47:07 -0500

KEITH LACY <keithl_at_vistatec.ie> WROTE:
> People really do seem to get hung up on the idea of
> click-through rates. Yes - they have a certain value
> but they're not the be all and end all. If you had a
> shop on a main street you wouldn't get too excited just
> because a bunch of people walked passed your window or
> maybe a few of them decided to come in and have a quick
> look around and left without buying anything, that
> doesn't pay the bills. Michael Martinez suggested that
> banners may be more beneficial as branding tools than
> previously thought, proving this could be difficult.
> It is not easy, through surveying, to parse out the
> individual effects of one branding tool when a company
> is currently employing a number of various branding
> tools in its marketing communications mix.

I agree. And, I apologize, but I feel compelled to
share some experiences and observations. Get your
coffee and cake, because this is a long one.

Undoubtedly the value of branding with banners is one
area that requires more study. My gut feeling is that
branding is effective only if people see the same
banner over and over again. I suspect there is a
threshold for cost-efficiency which is hard to define,
nearly impossible to identify for a given campaign.

If I may digress for a moment, there is a television
commercial right now which I find to be very
disturbing. My 9-year-old niece loves it. Well, the
music, anyway. It's the anti-Meth commercial where the
girl scrubs the bathroom floor with the toothbrush.
The song is very catchy and is geared toward children.
And I believe the commercial designer intended for
parents and guardians to immediately feel compelled to
talk about the dangers of drug abuse.

At least, that is the reaction in my family.

Can this effect be achieved on the Internet with
banners? I think we're still trying to answer that
question. Take the "Punch the monkey" campaign. What
the heck is that banner selling? It's selling
something. Most of the people I've discussed PTM with
are aware of it, hate it, and wish they could achieve
its purported success (I don't believe the campaign
has been formally documented by the advertiser, though
it's earned a lot of commentary from observers).

I dread the day I load a Web page and everything on my
PC stops working as the multimedia banner downloads a
sound file, a video file, and pops up a teaser window
to keep me occupied. I actually looked at an
interactive comic site the other day which did let me
play a game while the comic downloaded in the
background. What will it take to make a richmedia
banner do that?

The banner ads I do click on usually tell me exactly
what the site is about. "We have the widgets you're
looking for!" And usually they are running on a Web
page that has failed to give me what I'm looking for.
The search services know that keeping the banners tied
to the content is important. Their theory is that I'm
more likely to click on a "widgetworld" banner if I
type in a search for widgets.

Suppose I search for widgets every day, and I see that
banner every day. Eventually, I'm going to click on
the banner just to find out what it's about. Right?

Stealth content doesn't really work, in my opinion.
Relevant content does. So I don't believe there would
be a good transition from television advertising to
Internet advertising. I think the meth campaign
wouldn't work like on television because we're not
going to surf with the kids to sites that are likely
to run anti-drug banners.

I've seen banners lately with just one word on them.
What are they selling? What do they want me to
remember? I have no idea. They don't seem relevant
so I ignore them. But I do remember going to a medical
Web site one day a month or longer ago and filling out
a little form at the bottom of the page which asked me
about my favorite foods. It was a calorie counter.
This was a rich-media banner ad. I only realized that
just before I clicked on the button to run the search.
I clicked anyway, just out of curiosity.

That form wasn't branding. I'm not sure if even named
the Web site (probably not -- it was designed to look
like content on a Web page). But if I had been in the
market for the nutritional products that site was
selling, I think I would have been a pre-qualified
click-through.

It's been said often enough. A 10% CTR doesn't mean
anything if the conversion rate is 1%, whereas a 1%
CTR is great if the conversion rate is 100%. So,
okay, the industry average is down to around .31% right
now. I wonder what it would be without the PTM
campaign?

Trends are another aspect of the issue. You cannot
really brand with a trend, but you can have instant
success if you nail it immediately. When word broke
last Spring that Ted Raimi didn't want to play Joxer on
Xena: Warrior Princess any more, I changed the ad for
my Xena directory to use a simple 'Windows
application' background on which I imprinted the URL of
the directory and put in the "text box" the question:
"Is this Joxer's last season?"

Joxer is the third most popular character on the show,
according to a former producer on the series who has
seen the marketing research. I believe him. My CTRs
on the Microsoft Banner Exchange soared for several
weeks with that ad. Springtime is a pretty rotten time
of the year to run a television-related banner
campaign. We're almost to the end of the first-run
episodes and most people have found their favorite Web
sites.

Over the summer it was announced that NASA had found
water on Mars. Well, I run a science fiction domain,
so I thought it might be good to carry science news
feeds from Moreover.Com. No one cared about the news
feeds. I put more effort into the page than it was
worth, and that consisted mostly of cutting and
pasting Javascript from an email into a templated HTML
document, and then uploading it to my domain. I broke
a sweat by adding links to the page on my site map and
main news page.

But when the water-on-Mars story broke I immediately
set up a banner campaign with a simple two-tone
background (divided by a curve) that said, "NASA finds
water on Mars. When do we leave? Click here." That
campaign burst through the roof with a 2% CTR.

> Click-throughs are very limited in what they can tell
> you. They do not tell you much about whether what was
> presented following click-through was read and even
> less about whether it was understood or resulted in the
> desired brand attitude. Clearly there is a need for a
> better standard of measurement for the effectiveness of
> online advertising.

To some extent I agree with you, but I think there is a
very great IF which divides campaigns into two or more
categories.

When I first started running banners last year, I
signed up with several banner exchanges but eventualy
settled on the Microsoft Banner Exchange (for reasons
I don't need to go into). I created between 25 and 30
accounts. I thought, "This is great! I'll get loads
of traffic to all these Web sites and build up their
popularity!"

Right.

I ended up with dismal results. It was depressing.
I'd check the statistics every day and would see some
sites had 10 exposures, some sites had maybe 100
exposures. The CTRs were almost in the negative range.

Around August 1999 I began studying banner campaigning.
But I also looked at the large-scale commercial
campaigns and asked myself, what are these people
doing that I'm not? The answer (at the time) was that
many were pushing domain names. They were BRANDING
domain names on their banners. The tutorials I read
mostly agreed that branding the URL on the banner is
the very least you can do.

So I started doing that, and the CTRs limped up. I
decided I had taken on too many campaigns, so I
consolidated nearly all my accounts into three major
areas. Suddenly I had three accounts which each earned
several hundred exposures a day.

Then I looked at WHERE the banners were showing. The
banner exchange advised: "We reach 35,000,000 Web
sites! Show your banners everywhere!" Poppycock.
What good is a science fiction banner on a motorcycle
hobbyist's Web page? I decided to restrict my
exposures to relevant Web sites.

My CTRs averaged 1.2% for a couple of months.

As we rolled into December and January my CTRs dropped,
but I noticed that more and more people were starting
to visit the main index page for the domain.
Eventually, through improvements in search engine
placement and reciprocal linkage I was able to get the
three accounts up to earning about 1,000 exposures a
day. The CTRs levelled off around .8%. I had to
experiment with new banner designs and eventually
decided that asking relevant questions, hitting on
trends, worked the best. At the very least, changing
a banner to ask a new question improved the CTRs for
about a week.

I'll admit, I got tired of changing banners. I hate
designing banners, so I've let them slip. My traffic
has levelled off for now but it's more than twice what
I had a year ago. I'm satisfied that all my efforts at
increasing traffic worked and I need to redesign the
domain and add some improvements now.

But all this new traffic has enabled me to create my
own internal banner network, and since early May I've
been able to design and test my own banner campaigns
and see immediate results. I think it's important that
a domain with a lot of content promote itself. The
CTRs tend to be 2-3 times higher for in-domain
advertising than for external advertising. If people
like what they see on your page, and if they see you
have another page (remember, I put the domain name on
the banners), they are more likely to click through.

Of course, this is what worked (and didn't work) for me
over the past year. Your mileage may vary. And I'm
sure that if I had known when I began what I know now,
I could have learned even more than I have.

Science Fiction and Fantasy info_at_xenite.org
New and expanded discussion forums! Over 20 topics!
http://www.xenite.org/forums/
XENITE.org





Received on Tue Sep 26 2000 - 19:47:07 CDT