Keyword Elite AdSense Arbitrage Experiment Post-Mortem

As I mentioned before, I wanted to do a post-mortem of the Keyword Elite AdSense Arbitrage experiment we've been looking at over the last while.

The Results So Far

Let's start by looking at the results so far this year. First, the AdWords side. For the three days from January 1 to January 3 there were 39 clicks on my ads out of 24273 impressions. That's a clickthrough rate (CTR) of 0.16%. The ads cost me $1.78 in total.

On the AdSense side I can't be as detailed with my stats, so I'll just mention the important info: there were 19 clicks that made me $0.76.

So over that three-day period I lost $1.02. Not a large loss, but still a loss. Interestingly enough, though, it looks like half the traffic that I sent to the page via AdWords went and clicked an AdSense ad, which is not a bad rate. With a max cost-per-click (CPC) of $0.05 on the ads, all I need is to make at least $0.11 per click on the AdSense side and I would be making money. As it stands, I'm only making $0.04 per click via AdSense, which explains why I'm losing money.

Let's say we want to stay with this niche. My ads were shown almost 25000 times, after all, so the niche is definitely busy. How could I fix things?

Problem 1: Low AdWords CTR

Assuming I can fix the per-click earnings on the AdSense side, I need to get more traffic to the page. A CTR of 0.16% on the AdWords side is pretty low.

That's the average CTR, of course. Currently I have 6 different ad variations running. They've all been shown about 4000 times. The best ad has a CTR of 0.27%, not great but almost double the average CTR. The lowest has a CTR of only 0.07%. So the obvious thing to do is to cut loose the ads that aren't working as well, perhaps only keeping the best-performing ad.

But I should be able to do better than 0.27%. Obviously, my ad text is not compelling enough to attract visitors. I'm not surprised, really, because I didn't spend a lot of time writing them. Writing good ad copy is hard. Don't let anyone tell you otherwise. Some people have a natural flair for it, but most of us have to work at it. I need to sit down and re-read AdWords Miracle and come up with some better ads. If I could get even a 1% CTR, for example, that would give me at least 80 clicks per day. I say “at least” because the more clicks I get, the lower my AdWords costs will be, and therefore the more clicks I'll be able to afford… it's a positive feedback cycle that AdWords advertisers try to take advantage of.

Problem 2: Low AdSense Per-Click Earnings

I'm not doing too badly in converting the AdWords clicks into AdSense earnings: every two AdWords clicks yields me an AdSense click, which is basically a 50% conversion rate. This is good.

What's bad is that my per-click earnings are very bad, at only 4 cents per click on average. As I said above, I need to get that to 11 cents per click to make any money, and ideally more than that to make a really good return on investment (ROI).

Here's where tracking is important, and where I'm flying blind because I didn't do the right thing. What I should have done was assign a separate custom channel to each ad and link unit on the page. As it stands, the page I'm sending the traffic to has one link unit in the top right corner, a large rectangle at the top of the page and a large banner at the bottom — I'm following the template laid out in the report. If each of those had its own channel or I used click tracking software like AdSpyTracker then I'd know which of those three was performing.

Why is this important? Because the ads displayed on a page are ordered by value, with the highest-paying ads showing up in the first ad unit (where first refers to the first in the HTML, not necessarily first on the page), starting at the top or left of the ad unit and then going down from there. There are 8 ads on the page because there are two ad units with 4 ads each. It may make more sense to only have one ad unit displaying 4 ads (or less). It depends on where the visitors are clicking. I go into this in more detail in Chapter 14 of Uncommon AdSense, Display the Right Number of Ads, but with AdSense sometimes less is more, if you get my meaning.

Then there's the link unit. I like horizontal link units, but I've never been a big fan of vertical link units. I was surprised to see it mentioned in the report. Link units require two clicks to make you money. So really your visitor has to click three times: once on the ad to get to your page, once on the link unit, and then again on one of the ads shown by the link unit. Each click reduces your chance of getting paid. It can probably be dropped — again, proper tracking will tell you this. (Note, though, that link units can't really be tracked by click trackers, so you need to use custom channels to properly track them.)

Of course, maybe advertisers just aren't paying enough in this particular niche to make us more than a few cents per click. So I really should do some bid gap analysis on the AdWords side and see what kind of payouts might be happening on the AdSense side. But we'll leave that for tomorrow.

Depending on what we find out, though, it may be that we can salvage this arbitrage campaign with some better ad text and some fine-tuning of the ad layouts.

Sponsored Link: For a detailed look at AdSense arbitrage, see Michael Plante's book AdSense Arbitrage and Leveraging.

Eric Giguere wrote Make Easy Money with Google and Uncommon AdSense. His goal is to get his AdSense blog into Matt Cutts' blogroll.

Comments