• http://speedypin.com SpeedyPin

    @Benny This is a great article amongst a few others I’ve read regarding CPA and ROI optimization for PPC.

    I agree with your thoughts on CPA optimization for PPC campaigns. But, I really want to focus on ROI optimization theory and share an issue that I’ve faced a few times when optimizing for ROI at the keyword level. Hopefully, you and/or other readers can shed some light upon this issue.

    With regard to ROI optimization at the keyword level, I did indeed achieve our target CPA; I paused keywords with a negative ROI. In doing so, however, I somehow simultaneously over-optimized the campaign, which obliterated conversions. I know this sounds weird, but it is exactly what happened.

    My theory is that I had correctly paused keywords with a negative ROI, but some of those keywords were acting as a Helper or Assist. These now paused keywords could not do their job, which was to lead potential customers to further investigate our products by clicking on ads generated by more relevant keywords within the same ad group. As a result, the campaign experienced a significant decrease in conversions and the CPA actually increased to an amount slightly higher than where it was just prior to my optimization. Good stuff!

    Note: The aforementioned campaign had approximately 100 active ad groups at the time. Each ad group represented a different country, which is also likely affected by many unique and uncontrollable outside factors. Further, this campaign is being run in an environment where the number of customers has decreased by over 60% in the past 3 – 4 years and the competition has increased.

    I still believe that ROI optimization is smart and necessary, and at the same time frustrating. One spends several hours doing what the data says is correct and the results tell you that you merely wasted several hours ($$$$$) of your time. Ouch!

    In the interim I have moved back to CPA optimization for this campaign. CPA is acceptable and conversions have held. No significant increases in either.

    Again, any input you can share would be greatly appreciated.

    Thanks for the very timely and wonderful post.

  • http://www.esearchvision.com Benny Blum

    @SpeedyPin I feel your pain re: over-optimization and the issue that you touch on is very relevant. Fortunately, AdWords released conversion funnel reports last week (http://tinyurl.com/ykgflo6). These reports effectively identify terms that are supporting conversions but aren’t generating the last click in the sales funnel. Look for these terms, isolate them and optimize the creative. Isolating terms will also help to serve as a reminder to keep bids up.



  • ashton walsh


    PPC Services

    1. If you see your ads are not performing well, the CTR is
    low and you are not getting many clicks, do not just increase the CPC (cost per
    click), you might end up spending a lot of money without attracting quality

     2. The quality score
    for Goole search is completely separate from the quality score for the search
    network. If your ads are doing well on Goole search page it does not mean your
    ads on the display network are also doing well nor can they benefit from the
    text ads’ good performance on Google’s search pages.

    3. Always being on the top as ranked 1-3 does not guarantee
    a high conversion rate. Make sure your ads rotate around the first page and
    even the second page. Not everyone looks at the top of the search query results

    4. Make sure your ads and keywords in one campaign are not
    competing against each other as a result of keyword duplication.