We’ve run Google AdWords campaigns for years. Currently, by far, our largest online advertising campaign is our Print-Bingo.com campaign on Google AdWords. We do run ads on the Microsoft+Yahoo platform, but that hasn’t been worth the effort to work on optimizing, it’s hard to see any real ROI on these.
A couple months ago, Google offered AdWords users an incentive to add the conversion tracking JavaScript snippet to our campaigns. With conversion tracking, Google will track which advertising clicks lead to actual sales, in our case, we are counting upgrades to Premium access to Print-Bingo.com. In theory, Google will optimize our bids, both up and down, to match based on geography, time of day, search phrases, and potentially more, so that our cost per acquisition is optimized. We still pay for every click in CPA; however, Google tries to get us the best ROI for our clicks. The alternative on Google, is the traditional cost per click advertising, where we set a maximum bid for a click and find some other way to make sure that we are getting a positive return on our ad dollars.
Since it was on the to-do list anyway, we took up the offer to add CPA tracking. We’re glad that we did… probably.
We’re only a couple weeks into the process, and we’ve seen Google start to adjust the bidding. Search ads do much better than the content network – we knew this already, but we underestimated the difference in ROI. Our average cost per click is going up; however, our total spend is about even. It seems that higher spots in the advertising have a better ROI. And our conversions are at least even, if not up moderately.
Note that we do have a couple months of tracking data in the system before we turned on the CPA bidding. So we are trying to compare apples-to-apples when we count the number of conversions over a period of time. It is a bit worrisome, we’re putting quite a lot of trust that Google won’t try to cheat us; but what can you do… ROI is ROI.
There is a hole in this setup. Print-Bingo.com generates revenue from advertising, and we benefit greatly from word of mouth referrals by both paid and unpaid users. Since we are only tracking Premium users, we’re not giving a value to the free-level users that visit our site from ad clicks.
In the end, we’re tentatively switching to CPA bidding. We might even increase the advertising budget if the numbers hold up. We might add a second, parallel, campaign, aimed at getting cheap clicks from the content network… maybe.