Not long ago, I wrote a blog post about advertising. I’d like to expand on what I said before.
Last time I talked about not using your Twitter account to spam advertisements, since a) it only advertises to the limited number of followers you have and b) it can annoy people and lead them to unfollow you. Instead, I suggested using online advertisements, and mentioned that I was using an ad service called Project Wonderful. My Project Wonderful ads for “Radiance” had, at that time, received 53,000 views and 36 clicks in the course of a week.
Most of what I want to talk about today is the way I’ve been experimenting with these ads. I know very little about marketing and advertising, though I’m learning a lot more as time goes on. I’ve done a lot of independent study into advertising, reading articles online. I’m also doing Graduate Assistant work with Rowan University’s College of Business. Working with the College of Business is (indirectly) teaching me a lot about marketing, which is knowledge I’m trying to apply to my own independent business of self-publishing.
Since I don’t know much about advertising, a lot of what I’m doing is trial-and-error. Trial-and-error is actually how I learn most things; I’ve self-taught myself a lot about fixing computer problems and building a new computer from scratch just by diving in when a problem arose and figuring it out as I went along. Likewise, I’m figuring out advertising just by getting into it and seeing what happens. I want to explain my results here since I decided to analyze the numbers, and I figured as long as I’m doing so, I should share the information in case anyone would find it useful.
So far, I’ve been using four total ads. I made each one myself, using the cover for “Radiance” and making some very simple modifications. Project Wonderful gives specifications on what size the ads can be, and each website I advertise on decides which sizes they display. For example, one site might use horizontal banner ads across the top and bottom of the page, while another uses skyscraper ads on the right or left sidebars. Here are the four ads, along with the stats on how they’ve been performing:
Since the beginning of October, the full length “Leaderboard” ad has accumulated 50,027 views (13,182 unique users) and generated 37 clicks (36 unique users).
The smaller “Banner” ad has accumulated 149,252 views (19,121 unique users) and generated 104 clicks (86 unique users).
The “Rectangle” ad has accumulated 23,233 views (8,007 unique users) and generated 14 clicks (13 unique users).
Project Wonderful also allows three other ad sizes (a half-sized banner, a 125×125 square, and a 117×30 small button ad). I don’t yet have any of those simply because I haven’t had time to create them yet, so I’ve stuck with the ads seen here.
I’ve got a bit more to discuss about how I’ve used the ads, but first I’d like to just look at the raw numbers. Obviously, the “Banner” and “Skyscraper” ads have been the most successful. I don’t think that has anything to do with size; after all, these are all big, clearly visible ads (as opposed to the 117×30 button ad that is smaller than your twitter avatar). Based on my experiences with Project Wonderful, I think the main reason those two ads have been the most successful is because more sites allow those sizes to be posted. The “Rectangle” ad, for example, is a bit of an awkward size; it doesn’t fit comfortably around the edges of most websites, so depending on your site design, there might not be a good place for it. “Banners” and “Skyscrapers,” on the other hand, fit easily on a sidebar or below the page.
Also, as a general analysis, I’d like to look at the ratio between views, unique user views, clicks, and unique clicks. I find this information important when considering what I’ve learned about how advertising works (which is so far very little, of course). First you need to build awareness so that people know your product exists. Then, a certain percentage of those aware of your product will consider making a purchase, and a certain percentage of those will actually make one (for a more thorough explanation, here’s a website that explains this concept, and also goes beyond it in much greater detail, including a discussion of competition with other brands). While this concept is far more complicated than this (with complications I’ve barely started researching so far), the most basic part of it is fairly straightforward.
Grand total, my ads have been viewed 358,696 times, with 73,658 unique users. So a little over 70,000 people are (in theory) aware that “Radiance” exists, and each one may have seen the ad about 5 times (assuming the same visitor had visited the page 5 times over the course of a few days). Of course, these numbers may not be as accurate as they seem at first glance, since one person might visit a page from different IP addresses. For my purposes, however, these numbers give me a good benchmark.
People actually clicked on the ad 277 times, and those clicks represent 251 unique users. That means that out of 73,658 people who visited a page with my ad, 251 of them actually saw the Amazon sales page after they followed the ad. That’s only 1 in 293 people.
Out of all those clicks, there have been 3 purchases made in October (though I don’t know for sure if the purchases were made by people who found the ebook by these ads or from Twitter).
Breaking this down and rounding the numbers off, I find that for every 120,000 or so times my ad has been displayed, 24,000 individual people have seen it, about 80 have clicked on it, and 1 has purchased it. Based on that, I can take a rough guess that if I want to sell 1,000 copies of “Radiance” (which would make me $350 in royalties), I’d need to have the ad seen about 120 million times (or about 24,000,000 individual people).
Of course, there’s a few things I should explain about these ads. I’ve been mostly posting free ads, which generally only go on very low-traffic sites (mostly sites with less than 1000 page views per day). I’ve also posted a few penny ads (costing 1 cent per day per site they’re displayed on). It’s very likely that the numbers here would be very different if I were posting ads on, say, MS Paint Adventures, which gets over 2 million views per day at a price of about $50 per day. I’m not willing to pay $50 a day for an ad for a 99 cent ebook, since I’d almost definitely take a loss on the advertising costs.
There might be some other factors that influence the numbers I got. For one, my ads aren’t particularly fancy; the artwork for the ebook cover was designed by a professional, but I made the ads by cropping that picture and piecing things together in Paint. I don’t know enough about graphic design to know if my ads are particularly effective in their design. For another, my “Radiance” ebook is only a short story; when I get “Manifestation” published, the full-length novel might show very different numbers. Finally, these ads aren’t targeted to any particular groups. Since I post the ads for free, I just put them on the sites with the most traffic (many of which are webcomics). I might have more success if I could post ads on sites where the visitors are more likely to be ebook readers, but I don’t yet know enough to be able to do that effectively.
So that’s about it. I’ll probably revisit this topic again in the future to see if the ratios between views > uniques > clicks > purchases changes at all. In the mean time, I’ll keep trying to find more effective ways to promote “Radiance,” and if I find anything that is especially effective, I’ll be sure to share the results.