So yeah, the title of this post is a mouthful (or an eyeful, assuming you’re not moving your lips as you read this). I’ll start with the title, then. Taking it piece by piece, buzzword by buzzword:
ROI means “Return on Investment.” It’s one of those concepts that is fairly simple on its face, but can sometimes dissolve into blurry meaninglessness the harder you stare at it. As it relates to advertising, ROI means: how much money did you get back in the form of sales, versus the amount you spent promoting the product? Like I said, it seems simple. In a direct sales mindset, it is simple. If your product nets you $1 in profit, and you spent $1000 on an ad, and sold 2000 products as a direct result of the ad, then you had a 2x ROI. But it’s only in the world of direct sales advertising that ROI is simple. Most advertising on television, for example, or in magazines and newspapers, is not measurable by direct sales. You can’t click a Wii commercial to buy one. Those kinds of ads are called “branding” campaigns, and they have very different goals. The idea isn’t that you’ll buy a product as a specific result of viewing a specific ad, but that over time you’ll grow more and more familiar with the product, and be more likely to buy it, versus its competitors, thanks to the warm fuzzy branding messages you’ve received. Large advertising agencies spend millions of dollars “proving” ROI for branding campaigns — using focus groups, decades-long sales trends, demographic data, polls, etc. We won’t be doing that kind of analysis here, in part because I’m not qualified to do it, in part because we don’t have that kind of budget, and in part because ROI for a branding campaign is only measurable at the highest levels of saturation (Coca-Cola, McDonald’s, etc), and a product designed to appeal to hundreds of people, like, say, a t-shirt by a webcomic creator, is just too small to make any kind of noticeable branding impact in a poll or a focus group. Our experiment is strictly being conducted to determine a direct sales ROI: how much did we spend, how much did we make, and how do those two compare to one another.
Oh, did I mention this is about t-shirts? This is about t-shirts. Specifically, it’s about one particular t-shirt, the God Is Cute t-shirt by James Kochalka. I have chosen that shirt for this experiment because a). it hasn’t sold as well as James’ other shirts in the ComicSpace store, but b). it outsold all of our shirts when we exhibited at MoCCA, and c). it has a potential audience that exists outside of the standard “James Kochalka fan” audience — specifically, people who believe in God, and think He’s, you know, cute. As opposed to the people who think he’s all about fire and brimstone or whatever. Reaching out to this audience is key to ComicSpace’s online sales strategy, because, let’s face it, anybody can sell a t-shirt to the fans of the webcomic it came from. You, yourself, as a comics creator, can sell stuff to your own fans, all by yourself. You don’t need help. Unlike some of the other online stores that serve the webcomics space, ComicSpace is committed to finding audiences outside the already-existing fan audience for any individual creator’s merchandise. Part of that is identifying designs and concepts that might be able to “hook” other audiences, and then reaching out to them.
Here is the experiment: I have taken out ads for this shirt on three popular advertising platforms: Project Wonderful, Facebook, and Google AdSense. These ads will stay up for one month (through July 22, since I launched the campaigns yesterday). Every day between now and then, I’ll let you know how each ad platform is doing, in terms of generating clicks, and (more importantly) in terms of generating sales. If any. That’s the “liveblogging” part of this, the constant updates, for those who might be interested. If I tweak a campaign, change keywords, or target it to different websites, I’ll mention that, too, and share the results. I’ll also let you know how much we’ve spent, so far, on each platform. At the end of the month, I’ll wrap this all up with my conclusions. If any.
So, to get started, here are the parameters as the exist today:
Google AdSense
I set a budget of $10/day, bidding on the keywords “webcomic,” “comic” “kochalka” and the keyphrase “God is”.
So far, there have been no clicks (no impressions, either, which is weird — maybe I’ve done something wrong here) and the cost is zero. ROI is neutral.
Project Wonderful
I set this up to run across the entire PW network, with conservative bidding strategy and a cap of $500/month.
So far, there have been a few clicks. PW itself tracks 72 clicks. Our internal stats track 91. We have spent $19.46. There have been no sales. ROI is currently negative to infinity (because you can’t divide by zero).
FaceBook
I set this up to display only when the Facebook member has specified an interest in “comics,” “webcomics” or the phrase “God is Love”. According to Facebook, that’s 145,000 people.
So far, there have been a few clicks. Facebook itself tracks 25 clicks, while our internal stats track 31. We have spent $13.95. There have been no sales. ROI is currently negative to infinity (see above).
It’s likely that the difference between our tracking and PW’s and Facebook’s relates to their attempts to cancel out “click fraud.” We don’t have any such filters on our own click-tracking. Since I can’t imagine that anybody has any reason to commit click fraud, I actually trust our numbers more than theirs at the moment. I’ll keep watching this and see how it plays out, though.