• Twitter
  • Google Plus
  • Facebook
  • LinkedIn
  • Email
  • RSS
  • StumbleUpon
This is a guest post by Wijnand Meijer, a Paid Search Strategist at Netsociety, an online media agency based in Amsterdam (the Netherlands) and Brussels (Belgium). He created his first AdWords campaigns in 2006 and is currently helping advertisers and coworkers alike to get their Paid Search to the next level.

 

If you’re managing an account with an overall CPA target and you’re also advertising on brand terms you’ll probably see two types of keywords in your account: brand terms with very low CPA’s (and CPC’s) and non-brand keywords with much higher CPA’s (often 10 to 20 times higher). Both CPA’s are usually not even close to the overall target, nor will they be for most keywords, even if you tried.

However, your boss or client is still mostly interested in the overall target. So if you consider your branded conversions as a given, what should the target CPA be for the non-branded keywords?

Let’s start by saying that this is not an ideal situation. Far from it. George Michie even calls it ‘the cardinal sin of paid search’ and gives 4 good reasons not to mix brand with non-brand results in his excellent article. Please tell your boss or client about these 4 reasons and hopefully you can switch to working with a non-branded target. That would be in everyone’s best interest.

Working with a non-branded target would indeed mean you’ll no longer get credits for the branded conversions. But neither will you be blamed if they drop.

If you think your non-branded campaigns often assist branded conversions, you might be right, but I’d say the burden of proof is on you. Analyze Multi-Channel Funnels in Google Analytics, Search Funnels in AdWords or even request a Campaign Insights report at Google if you’ve been very active on the Google Display Network. Then take credit for the additional branded conversions you find in these reports. Those would be the ones that are preceded by a non-branded click with the subtraction of the conversions that started with a branded click and ended with a non-branded click (this also happens).

So let’s hope this article will become obsolete sooner rather than later. In the meantime, if the overall target is the situation you have to work with, I’d like to show you how to create your non-branded target based on existing conversion data in your account. This will help you evaluate the performance of your non-branded keywords and give you a more realistic target for them.

Especially if branded conversions are significant in your account (or maybe even make up the majority of the conversions) the overall target will not be suitable for your non-branded keywords, so you’ll need a target that will work together with your branded conversions to reach the overall target.

I’ll use Google AdWords in the examples below, but if you get your conversion data from another source, the calculation works just as well, as long as you’re able to distinct brand from non-brand terms. Calculating your non-branded target works best with mature and stable accounts, but if your account isn’t, you’ll just have to recalculate more often to make sure the calculation holds true.

I will provide you with the steps to calculate your non-branded target for both of the following situations:

  1. You have clearly labeled and completely separated your branded campaigns from the other campaigns, and added your brand terms (including misspellings) as negatives to all the other campaigns (for the past year).
  2. You’re not 100% sure all branded search queries were only triggered by branded campaigns (for the past year).

1. The steps for the first (and ideal) situation in which you have completely separated branded from non-branded are as following:

  1. Log in to your AdWords account and select the past 12 months as the date range:
    daterange
    The reason for choosing a full year is to prevent a possible seasonal bias when choosing a shorter period. If your account is running for less than a year, select ‘All time’ in the date range field.
  2. I assume all your branded campaigns contain the word ‘brand’ or another common and unique combination of letters you can filter on. Search for this in the search field:
    brandfilter
  3. Now scroll to the bottom of the page, you will find the aggregate statistics for your brand campaigns in the yellow row:
    brandstatistics
  4. From the image above, we just need 3 metrics (highlighted in red): the number of branded conversions, the number of total conversions and the branded CPA.
  5. Now divide the number of branded conversions by the total conversions, in the example above that would be 23,886 / 34,488 = 69%
  6. Now let’s say our overall target CPA is $40. Given the metrics above, what should be our target for the non-branded keywords?
  7. If our target CPA is $40, then 100 conversions can cost us $4,000. Now we know that out of every 100 conversions, we get 69 conversions for $2.19 apiece. So we get 69 conversions for $151.
    That leaves us with $4,000 – $151 = $3,849 to spend for the remaining 31 non-branded conversions. Therefore their target CPA would be $3,849 / 31 = $124
  8. To make it easier for you, just fill in your overall target CPA, your amount of branded conversions and their CPA and the total number of conversions and this Excel file  will calculate your non-branded target.

2. The steps for the second situation in which you’re not 100% sure to have completely separated branded from non-branded take a bit more time, but at least you will be sure to get the right numbers by separating on the search query level:

  1. Go to the Dimensions tab of your account and select the past 12 months as the date range
  2. Select ‘Search terms’ in the view menu.
  3. Customize the columns and only select Clicks, Conversions and Cost. Remove all the other columns.
  4. Create a filter to show only search terms with at least 1 click:
    dimensions-searchterms
  5. Download this report and open it with Excel (if it gets too big, use a higher number of clicks in the filter).
  6. Delete the first and last row of this report (title and totals)
  7. Name column E (the first empty column) ‘Keyword Type’
  8. Set a filter on the first row
  9. Apply a filter on the first column (Search term) to contain your brand name
  10. Fill in ‘Branded’ in the new Keyword Type column for all these keywords (make sure they all really are branded)
  11. Remove the filter from the Search term column
  12. Select the empty cells from the Keyword Type column
  13. Now you should see all your non branded queries.
  14. Now fill in ‘Non-branded’ in the Keyword Type column for these queries.
  15. If all went well, all your queries are now labeled as branded or non-branded.
  16. Create a Pivot Table based on this data. If you need to learn (more) about Pivot Tables for PPC analysis I can highly recommend reading Master Pivot Tables by Josh Dreller.
  17. In this Pivot Table use Keyword Type as Row Label and the sum of conversions as Values.
  18. Create a Calculated Field called CPA that divides Cost by Conversions
    calculatedfield1
    and add this to the Values field
  19. Create a Calculated Field called Conv. Rate that divides Conversions by Clicks.
    calculatedfield2
    and add this to the Values field
  20. Now you should have an Excel file similar to this one.
    pivottable
    Now you have all the data you need to calculate your non-branded target. Just fill in the numbers in this Excel file and it will calculate it for you.
  21. In the Pivot Table you’ve also calculated your branded and non-branded conversion rate. Though you don’t need these metrics for your non-branded target, they’re quite useful to know for the following reasons:
    • When you add a new non-branded keyword, you can use the average non-branded conversion rate as a benchmark for when to expect the first conversion. If your non-branded conversion rate is 1%, your non-branded keywords need 100 clicks to convert (on average). So you can optimize accordingly.
    • By regularly comparing your branded and non-branded conversion rates you can look for interesting trends in these metrics, for example: if one rises, does the other rise with the same relative amount? If not, why not?

It goes without saying that your calculated non-branded target holds true as long as the proportion of branded conversions and its CPA does not change in your account. But it probably will, so it would be wise to recalculate it regularly (at least once every 3 months I would say).

And if you’re not sure your current CPA (non-branded or overall) is realistic and want to reevaluate it, this guideline explains how a target CPA can be calculated. If you just want to estimate your CPA based on metrics like CPC and Conversion Rate, you can use this calculator.

I hope this guide is helpful for all of you working with an overall CPA target. I also hope this will be a temporary situation for you ;-)

This is a guest post by Wijnand Meijer, a Paid Search Strategist at iProspect|Netsociety, an online media agency based in Amsterdam (the Netherlands) and Brussels (Belgium). He created his first AdWords campaigns in 2006 and is currently helping advertisers and coworkers alike to get their Paid Search to the next level.

 
Opinions expressed in the article are those of the guest author and not necessarily Certified Knowledge. If you would like to write for Certified Knowledge, please let us know.
 

Share and Enjoy

  • Twitter
  • Google Plus
  • Facebook
  • LinkedIn
  • Email
  • RSS
  • StumbleUpon