There was a recent Google AdWords Hangout, “Improve AdWords with Google Analytics Data,” that discussed three reports in Google Analytics and how they can be used to improve your AdWords PPC performance. Two of the concepts were fairly basic and straightforward, though still missed some key information I thought would really help Advertisers out. These two concepts looked at the Google Analytics Devices and Landing Page Reports.
The third concept (the first one covered in the hangout) focused on the Location Report. I felt this has the most potential for advertisers to learn from and improve AdWords performance but again some key information was missing.
The hangout was also covered on the Search Engine Land post “Improve Your AdWords Performance With These 3 Google Analytics Reports.” The author expounded a bit on the how the analytics reports could help, but again, I felt I could add some insight to help improve your PPC performance.
Google Analytics Location Report to Improve AdWords Performance
First login to both your Google Analytics and Google AdWords accounts (for the same properties of course). Set the same date range for both Analytics and AdWords. I would suggest choosing a longer date range, like 3-6 months, though it may also be a good idea to compare date ranges later, like 1 month to 3 months or 6 months, also keeping in mind any seasonality in your products or campaigns. You also need to have your Analytics and AdWords accounts linked and either Goals or eCommerce conversions setup in Analytics (or both!).
Navigate to the Location Report in your Google Analytics account:
Audiences > Geo > Location
Click on the number one country/territory by sessions (Analytics will default sort by sessions) or the country/territory you would like to focus on for your specific AdWords campaign.
Finding Your Regions
Now you will see the Regions (or States) listed, again sorted by sessions. You can also drill down into cities by clicking on a region/state. For now we will focus on Regions but you can apply these same steps to cities if you choose or are running a localized campaign.
Here is where both of the aforementioned hangout and post fall short. The hangout instructs you to just focus on your top three Regions (or Cities if you choose) of course by Sessions only, a common suggestion by Google (wonder why?). Which is ok if you only want to improve your AdWords performance in the Regions where your site gets the most sessions. But we are interested in true performance here–results, conversions. The SEW post touched on determining your highest converting regions but didn’t provide enough details for you to make the correct decisions.
We need to focus on both sessions and conversions (or revenue if you choose). Simply sorting by conversion rate won’t do. A 100% conversion rate isn’t very useful if you only have one session. And of course simply optimizing on sessions is a bad idea if your region with the highest sessions has zero conversions. Though that is useful information to have and if true I would put in some further analysis to find out why.
Filtering Down with Useful Data
First you want to put on an Advanced Filter to only include your Regions with highest traffic. That threshold of sessions is going to depend on the traffic of your site, so expand the report to Show All Rows and get a sense of the range between all regions. Your looking to narrow it down to maybe 10-20 regions so notate how many sessions there are for the regions in that range.
Now click on Advanced Filter, type in Sessions into the dimension field, and choose Site Usage > Sessions. Then type in your threshold number for sessions in the last field so it looks like this:
Include > Sessions > Greater Than > (Your Threshold Number).
Next click to sort by Conversion Rate (either eCommerce or your desired goal). Now pay attention to your top regions by Conversion Rate, BUT also keep an eye on Revenue as well.
A region with a slightly lower conversion rate with double the amount of revenue is going to provide you with much bigger results.
(Export the Analytics Location Report as a csv file. We’ll need that for later.)
Making Sense in AdWords
Now head over to your AdWords Account > Campaigns.
Click on your desired Campaign > Settings tab > Locations tab.
Click on the red +Locations button and add your top 3-5 performing regions from the GA report and Save.
Then click in the Bid Adjustment column and increase your bids for each region–start with maybe a 10-15% increase. You can also try decreasing bids for under performing regions.
Now here’s where things can be a bit confusing. The data we’ve pulled from Analytics is for all sessions, not specifically for AdWords PPC, or even more specifically each AdWords Campaign, which is where the location settings in AdWords is controlled–at the campaign level, not for the account overall.
So you can think about this a few different ways. If all users to my website from these regions are out performing then increasing PPC bids (marketing dollars) for these regions should put our Ads and product/service in front users from those regions and they should perform better as well. Which has a strong case and is worth some testing.
BUT you can also dig further into your data and see what regions are the top performers for your Paid Traffic and also for each AdWords campaign.
Segmenting Data for Paid Traffic and Campaigns
Let’s start by looking at the Paid Traffic in Google Analytics, specifically Google AdWords Traffic. You will need to start the process over by removing the Advanced Filter.
If you don’t already have a custom segment for AdWords traffic you can make one with the conditions in the image below. Or you can use this Google CPC custom segment I’ve already created. (NOTE: There is a system segment included with Google Analytics for Paid Traffic, but that includes ALL paid traffic, so if you are using Bing for PPC it will include that data as well. If you are only advertising with AdWords, the custom Google CPC segment will match the Paid Traffic system segment.)
Now that you have your custom Google CPC segment:
Click Add Segment > custom segment view > the checkbox next to the Google CPC segment > Apply
Then delete the All Sessions segment view from the report. Follow the same process of sorting by sessions, finding your threshold and setting and adding the Advanced Filter as explained above. Then sort again by Conversion Rate and look at the top regions and remember to also pay attention to their revenue as well.
Do the top performing Regions match the ones from the All Sessions view? If so you can feel even more confident you’re on the right track. If there’s some new surprises you may want to consider adjusting those bids as well or at least taking note.
(Export this Location Report as a csv file as well and follow these directions)
Compare Segmented Data in Excel
Open up both location report csv files exported. Choose the All Sessions segment file first. Insert a new column after Region, name it Segment and insert the name “All Sessions” for all rows. Then File > Save As an excel file.
Open up the Google CPC segment file and repeat the process of inserting a new Segment column but insert the name “Google CPC” for all rows. Now highlight and copy the entire table and paste it into the newly saved excel file containing the All Sessions segment so you have both segments in one document.
Keep the All Sessions text as black, but make the text of the Google CPC segment a different color to differentiate it. Then do a Custom Sort–first by Conversion Rate, then add a second level by Revenue, both ordered Largest to Smallest.
Now you can start to compare what your Top Performing Regions are by each segment. Highlight each of the Top Performing Regions (3-5) for each segment and fill the row with a color (like a light grey).
This will give you a good visual representation to compare the top performing regions for each segment and will look similar to the screenshot below.
Segmenting Comparisons by Campaigns
Now here’s where your comparisons go even deeper. Remember when I said the location settings in AdWords is controlled at the campaign level? So again, doesn’t it make sense to segment your sessions even further down to the campaign level and compare the results to what we have already found?
If you don’t already have custom segments setup in Google Analytics for each of your AdWords campaigns now is the time. Unfortunately I cannot share these custom segments with you as each account will have their own unique AdWords campaigns. But becuase your GA and AdWords accounts are linked the process is rather painless.
Head back to the Location Report in GA and click Add Segment. Click the red +NEW SEGMENT button > Traffic Sources tab > Filter Sessions button > Campaign > Exactly Matches > and when you click your cursor in the field GA will populate a list of all the campaigns you’ve created (this also includes any manual campaigns you’ve created for email marketing, events, or other advertising). Choose one of your AdWords campaigns, name the segment appropriately (like CPC – Campaign Name) and Save. Repeat the process for each of your AdWords campaigns.
Now that you have your custom Google CPC Campaign segments:
Click Add Segment > custom segment view > click the checkbox next to the Google CPC Campaign segment > Apply
Again (I know, I know) follow the same process:
- Sorting by sessions
- Finding your threshold number
- Adding the Advanced Filter
- Exporting as a cvs file
- Inserting and naming a segment column
- Pasting into your main excel file
- Coloring the new text
- Custom sorting by Conversion Rate, then Revenue, ordered Largest to Smallest
- Highlighting rows of top performing regions and
- Comparing to your other segments
Repeat this process for any AdWords campaign you wish to target, compare, test and improve performance–increasing/decreasing your bids accordingly.
Don’t forget to monitor these new regions for performance, giving them an ample test period.
Interested to hear your questions, thoughts and how your performance has improved using these processes in the comments.