Fair enough! May be I just need to sit down and review some CPI trends in more detail…

Kudos on putting the idea out there for discussion – and for defending it.

Appreciate the feedback. I wrote this blog entry for two reasons. One was to take a look at the PPC process as a whole. That is from initial search query —–> clicking an ad —–> going to a landing page —-> completing the conversion. I understand that this is a “big metric” picture and can be influenced by many variables, but so can CTR, conversion rate, CPA, etc. when you look at these metrics at the campaign level. For example, a campaign might have a CPA of $10, but when the branded terms are removed the CPA rises to $30. Just like these metrics, CPI need to analyzed at the keyword, ad group, and text ad levels.

My second reason for this entry had to do with a recent client meeting to go over the monthly report. I went into detail about the trends we were seeing and what the metrics were telling us, but the client specifically wanted to know the percentage of searches that were turning into conversions. This client query got me interested in exploring CPI in more depth.

As I say, it may be time to consider the metric of CPI. CTR and conversion rate are both crucial metrics, but when you look at them as two separate processes each only tells half the story. Those stories being getting the click and then getting the conversion. CPI ties them together and evaluates the entire process. Time will tell if CPI turns out to be a truly actionable PPC metric.

Interesting thought process. Though I would argue that CPI is not a very actionable metric. I feel that it is too “big picture” – too all-inclusive. Impressions as a pay-per-click metric are determined by a very wide spectrum of variables: Max CPC, negative keywords, search queries, competition, seasonal/daily fluctuation, etc. Pairing this seemingly “wild” metric with conversions doesn’t really tell me much. Yes, you can gain understanding of how your entire PPC-targeted audience funnels down into customers/conversions, but I’m interested in what you propose we do with that information? Maybe go crazy with search query reporting and add tons of negative keywords to drop your impression stats… (and therefore increase CPI for the sake of increasing CPI)

The way I see it, you would still have to break your statistics down into the run-of-the-mill CTR and conversion metrics to make any meaningful changes to your campaigns.

Don’t mean to burst your bubble – just throwing in my 2 cents! ; )