Tracking Bogus Google AdWords Conversions

I recently bought a few AdWords ads for Microsoft adCenter's affiliate program. These links were direct affiliate links that headed directly to Microsoft - the searcher never touched my site.

Compare the following 3 AdWords ad campaigns

The first campaign is a strategic one, where I do not mind losing money if it increases usage, which may lead to more links (and better organic rankings) over time. But that second campaign, with a similar number of clicks, never even touches my site and still has a baseline conversion rate and conversion cost similar to the strategic ad group.

Truth be told those conversion sample sizes are so small that it is hard to draw concrete evidence from them, but if I was telling myself that some of the ad sales caused by that strategic ad campaign help subsidize at least some of the cost, then I might be operating under a false pretense. Some of those conversions may have happened anyway.

That third ad campaign consists largely of brand related keywords and a few other somewhat related terms. Notice how the conversion rate is higher. Microsoft recently published research that brand related search terms tend to convert better than twice as well as non-brand terms.

Why is this important? As Google controls an increasing large piece of the online advertiser pie, if you use their analytics, many of those conversions THEY track are falsely tied to their ads. They would have happened even if you were not buying AdWords ads. As far as brand related conversions go, conversions tied to brand phrases typically are not incremental, which means those would have happened even if you were not buying AdWords ads.

You can use direct conversions as a proxy for the value of advertisements, but if you have a large ad campaign and a well known brand, you are likely buying brand exposure more than direct conversions, even if you control your spend on using a CPA metric.