Turns Out, Tired Maxims About Advertising All True

Adding this to things I learned from running an internet company: Sex really does Sell.

So, last time I gave a run down of our advertising options and the relative effectiveness of each in our particular case. That essay ended with a ringing endorsement for Project Wonderful, and I got lots of queries for follow up information. Specifically people were interested in the types of ads we ran and exactly how one ad form compared in results to the other. So this is a follow up guide.

First off I’d like to start with some general guidelines for successful advertising, because it’s not just about paying the big bucks to put your ad on the right site. There are lots of reasons why thousands of pageviews on Website X will not translate to thousands of clicks for your ad. If you stumble over the basics then you’re just throwing your money away.

Obviously having a nice looking ad is key. Use good stock imagery. Sxc.hu can help you if you need to do this on the cheap, but for those who can spring for more an istock.com account is worth its weight in gold.

But this is more than just producing a PRETTY ad. When designing an ad, or when shopping through potential advertising spaces on Project Wonderful you have to keep in mind the overall composition of the ad WITHIN the page you’re placing it on. Putting a black and white ad on a page that is predominantly black and white is much less effective than putting a colorful ad on that very same page. Contrast is key, you want visitor’s eyes to go STRAIGHT to your ad.

That’s something that has to be judged on a case by case basis really. The design process can (and often will) happen from both sides of the issue: sometimes I design an ad and then go find a site for it, sometimes I have a site already in mind and I design the ad specifically for that opportunity. The point is you should never have stock ads that you just plug in everywhere. You can have the prettiest ad ever made and the best placement on the site in question, but if the ad blends into the page you are screwed.

That brings us to another important issue: placement. Is bad placement on a high traffic site worth more than good placement on a low traffic site? Generally, no. Although I have come across some exceptions. Ideally you want “above the fold” placement. That is, you don’t want people to have to scroll down to see your ad. Unfortunately for small time advertisers a lot of the best sites are not going to give you the top top ad spaces, so you will have to consider how far above the fold is okay and if you’re bidding on a below the fold ad space … how likely is it visitors are going to scroll down?

So personally, if the publisher has given the first skyscraper ad space to, say for example, Google and there’s a second space directly below that, I might take it because on most monitors 25% of that second space will be above the fold. If the site is a good fit for the market I want to reach and the content will get the user to scroll down for sure, I can do something with that.

But how do you determine the likelihood that a visitor will scroll down?

Well … did you scroll down when you checked out the site for the first time?

A lot of this is simply intuitive. It’s again about page composition more than anything else. I have no problem with advertising below the fold on a webcomic for example, because users that read webcomics have to scroll either to read the end of the comic or to read the artist’s blog comments below the comic. But I’m not going to advertise below the fold on a site where links to everything you could ever want or need are above the fold. Blogs are very much case by case. Does the blogger update multiple times a day, thereby forcing readers to scroll down to catch up? Or are there days and weeks in between bloggings so that even casual readers can peruse the new stuff without really scrolling down?

The last basic issue is ad size. When you first open up an account with Project Wonderful and try to create an ad they ask you to pick an ad size … what to choose?

Do: Leaderboard, Skyscraper, or Rectangle if the opportunity is right.

Don’t Bother With: Banners, Buttons, the dreaded 125×125 “Square”

In advertising, size does matter. Size really really does matter. Square formats are particularly useless because publishers tend to lump them together in ad quilts. If every ad is well designed and unique than that might work, but generally buying Square advertising space is cramming your ad into a noisy eye-sore.

I liked Project Wonderful so much I decided to expand the ad campaigns. Not all our ads worked (cause we’re not nearly as smart as we pretend to be) so I thought it would be interesting going into detail about the individual campaigns, what worked and some thoughts on why they worked.

First Campaign: Split-Self, Premier.
Started this ad the week Split-Self premiered and it did pretty good. It ran on a variety of different sites, from Tokio Hotel Fiction, to Tales of MU, to webcomics like White Noise and Sore Thumbs and Wondermark. Since there weren’t any options that hit on the exact market we wanted, we focused on sites that came close and offered page compositions that gave this particular ad maximum exposure. By far the best response was from Tokio Hotel Fiction which only gave us a .07% click through but that .07% stayed for an average of 5 mins on the site. The worst of this bid group was Sore Thumbs .04% click through, barely 30 seconds on the site. White Noise was somewhere in the middle, .11% click through 55 second on the site on average.

That’s not to say either of those sites is a bad advertising opportunity. They just weren’t as good a fit as we hoped for Split-Self

Incidentally I liked this ad so much I did another version of the imagery for a banner exchange I’m optimistic about (this is free program so it’s really impossible not to be optimistic about it I guess)

Second Campaign: Fluffy-seme General.
This second campaign (Note the ad was resized to fit this page click here to see it’s full size) did not do well at all. In retrospect I think the ad was too mysterious, too vague, not enticing enough and visually perhaps a bit too busy. Click through rates were not horrible, but we focused on writing sites and many of them did not have enough traffic to really maximize the advertising’s cost effectiveness. Of these only Crystal Hall Forums gave us a okay performance with a .16% Click Through and 1 min on average on the site. The CRFH Forums gave us a 1.36% Click Through … but 1.36% of 220 daily pageviews is … 3. Three Clicks and they were gone so fast analytics can’t even give us an average time.

Goes without saying, we pulled these ads after a week. Why waste your money on something that doesn’t work when Project Wonderful makes it so easy to just do something else?

Third Campaign: Season in the Red Slashy.
Around this time I decided I wanted to run a campaign for Season in the Red. Now Season in the Red has an active fanbase already. Originally I had thought it better to focus on Split-Self because it’s a baby and hasn’t quite found its fans yet … but at the same time we are having problems in the forum keeping activity levels up. And the more I explored the issue the more I realized that SiR fans all kind of ran in the same circles to begin with. So of course they weren’t really hanging out in the forum, they were already hanging out on AIM and other forums all across the internet! So, I thought perhaps if we attract some fresh meat to the group that will help us sustain activity.

Season in the Red is about hockey players, so naturally the first instinct was to see what advertising was available in the sports section.

Ick, nothing good. Either the wrong sport or exceptionally bad placing. Waste of money right away. Didn’t even bother.

A little background information. Season in the Red has no slash in it. Not in the actual story anyway. However since it’s about the personal lives of young hockey players it does attract slash fans. This has never been a problem for me. In fact I LIKE the slash. I would also like to see SiR attract a more general market because I think it can, but what we have right now are the slash fangirls and I’m happy with that.

So when general sports failed, the natural next step was … hey let’s find the slash fangirls! So I pulled together this ad, sexually provocative in the right ways, visually appealing, got it up on June Manga, DMP Books and MangaBullet.

Jesus FREAKING Christmas!

Prior to this I’ve been talking ad effectiveness purely in terms of Click Through and Time on Site because these are the only stats we have for the campaigns. We haven’t talked at all about Conversion Rates (that is visitors who come to site and sign up for an account) because it’s really difficult for us to say. We’d got one or two sign-ups we couldn’t identify as coming through from other efforts, but is this the advertising or is this just a lucky random websurfer?

This campaign made membership EXPLODE. Whenever our bids held the top spot we got a couple new sign-ups. I was pleased and amused and naturally wanted MORE

Campaign Four: Season in the Red, Even Slashier.
I designed a rectangle ad originally because I wanted to advertise on June Manga and MangaBullet and that’s what space they had up for auction. I’m undecided about rectangle boxes, I think they sort of end up in weird places on layouts in general. But, even though June’s space was way way below the fold, I felt these were good opportunities that warranted the experiment and it paid off.

Still there were other opportunities at different ad sizes, so I went for good old Skyscraper format next. This ad performed even better than the first. Many of the Skyscraper opportunities were cheaper, and we were able to get into Smack Jeeves hosted sites. I had never heard of Smack Jeeves, but as soon as I saw the number of Boy-Love comics they run I knew I wanted in. Advertising on the actual service page as been most useful a .3% click through equaling over 250 vists, 5.08 minutes on the site on average. The individual comic pages have been (so far) less successful, but still performing really really well.

It will be couple of weeks before we know for sure whether these new members will become active in the site. Some of them, presumably, are still working their way through Season in the Red’s archive and may not become active until they’ve caught up with the story.

Campaign Five: Split-Self MOAR Vampire.
Moral of this story, change out ads often. Ads lose their effectiveness over time, but once you’ve found an advertising opportunity where people will click through and dig around your site for a couple of minutes … you want to find a way to keep bringing people back. Rather than putting out new ads on new sites, I’m all for new ads on the same sites. Project Wonderful makes it so easy to create new ads and switch them out on current bids. So mix it up. Even the best ads loose effectiveness as people get used to seeing them and begin to ignore them. I loved our first Split-Self ad, but it was time to give it a break and put up some new eye candy.

Quick Note: that image is from sxc.hu … proof that you do not have to pay for good stock, sometimes you can just get lucky.

Can’t give you any stats on this one yet because it just started running. We’ll see …

More thoughts: Where should your ad direct?

As you can see, we ran a bunch of different stuff. Some directed to the main site, some directed to the story pages, some directed to specific content. If you remember my original essay, I said directing to content is better than directing to main pages. So far that still holds true here. Ads that directed to story pages performed better than ads that directed to the main page (obviously if you’re only hosting one story on your site/blog this is not even an issue). There was no noticeable difference in performance when the ad directed to the First Chapter verses the Story Table of Context.

Closing Thoughts

Hope you found this more detailed break down useful Leave me a comment here with any questions or thoughts!

Bookmark It

Hide Sites