Paid Search Metrics Too Good To Be True (and What They Might Be Hiding)

Posted on Tuesday, August 20th, 2013 by Print This Post Print This Post

Categories - Featured, SEM

metric-mentronome

In our recent post Paid Search Metrics With a Bad Reputation (And Why You Should Hear Them Out), we outlined the metrics that might signal a failure within an account when observed in isolation, but that, upon closer inspection, may actually be welcomed indicators of success.  To be fair to paid search metrics across the board, there’s also value in looking at the other side of the coin – metrics that marketers often assume are good, but may actually represent an opportunity for improvement.  Take a deeper look at each of the following before deciding which metrics truly deserve to be highlighted in your next meeting:
  • Impression Growth: More Impressions = More Customers.  Case Closed… right?   Not so fast – while  “growth” may seem like a good thing, a growth in impressions may not be the blessing it appears to be.  Impression growth merely means that your ad is appearing to more people, but it does not mean that your ad is appearing to the “right” people.  Impression growth could signal that your account is using broad match to show ads alongside irrelevant queries.  Unless you are purely executing a brand campaign, if the increase in impressions is not accompanied by a similar increase in conversions, then the extra exposure is not providing value to your campaigns.  Even in campaigns focused on branding, impression growth is only a sign of success if you can determine that the right people are being exposed to the ads.   Mining data around search queries and user engagement will tell the true story of your change in exposure.     
  • Account CPC Decline: Often, advertisers believe that their account has become more efficient when they see a decline in account CPC, a cause for celebration.  More commonly, however, account-level CPC declines signal a change in keyword mix, not efficiency.  An increase in traffic to lower spending keywords or a decline in more expensive terms could actually be a bad thing for the account.  In many verticals, non-brand volume drivers are easily identified and therefore attract a high level of competition.  These terms are often the most expensive as a result, and if you are losing group here you could also be slipping in sales as a result.  To extract a more accurate vision of what CPC decline could mean for your account, evaluate CPCs by looking at exact match volume drivers in both brand and non-brand categories, separately.  Combined with changes in Quality Score, Average Position and other metrics like Impression Share, you will have a much better understanding of the health of your keywords in the competitive market.   
  • Sudden Lead Growth: Spikes in the number of leads that come in each day is only a sign of success if you can determine the quality of the incremental leads.  Sometimes, especially in the financial vertical, sudden spikes in leads can be the result of a type of fraud in which bots fill out lead forms en masse with faulty information. Be suspicious.  If you did not make changes to your account that could be the source of this swing in performance, you need to do some digging.  Check for duplicate leads, pixel issues, and patterns in your lead data to determine if there is an error or malicious intent from an outside source.  If everything comes up clean and leads look similar to your typical quality, congratulations, you’ve found success!
  • Overall Spend Decline: Looking at total spend in isolation can be misleading – did your leads, sales, or other points of conversion remain the same? An overall decline in spending might signal that you may have unintentionally restricted traffic somewhere in your campaign.  Check your budgets to see where you might be limiting your service.  Also remember the distinction between branded and non-branded terms.  Did your spending patterns change similarly in the branded and non-branded buckets? Branded spend is often more easily influenced by the other channels and could indicate that you are not seeing the same level of influence from TV buys, email campaigns, or other types of advertising.  Non-branded spend changes could indicate a change in Impression Share or other external influences that you should try to understand and react to if appropriate.
Like our previous post about Paid Search Metrics with a bad rap, the key takeaways here are that no one metric should be analyzed in isolation, and no major account decisions should be made until the account’s success is appraised on a holistic level. Trends and changes in a single metric are beneficial to watch, but the most actionable and important analysis for the health of your campaigns revolves around how they interact and balance each other to make an account that works best for your goals.  Up next we will look at Paid Search Metrics That Are Popular For No Reason.  

Tags | , , , , , , , ,

Leave a Reply

Follow Us on Twitter

Authors