Dealing With AdWords and Analytics Data Discrepancies
Elizabeth Marsten Apr 26 2011
Today’s post is brought to you by Bethany Bey of PPC Hero!
In a touching speech by President Jack Nicholson, he asks alien invaders “Why can’t we work out our differences? Why can’t we work things out? Little people, why can’t we all just get along?” Although things didn’t end so well for him, this is often how I feel about the relationship between AdWords and Analytics. They both provide excellent information that is extremely valuable to anyone managing a PPC account, so why oh why can’t they just get along.
Last week I was hit head on with the data discrepancies between the two interfaces when I went to implement ad scheduling for an E-Commerce client. I wanted to analyze hourly performance to determine if there were times throughout the day when we could either reduce bids or pause ads to better utilize our budget. Sounds easy enough right?
I first downloaded conversions by hour in AdWords. Below is a screenshot of the results:
The account is spending the majority of its budget at 12am, at a cost per conversion of $465.89. This amount seems high, especially compared with other hours of the day, but I wanted to know the value of these conversions before I made any decisions. I pulled revenue data by hour from analytics to calculate ROI and this is where things got interesting:
The rows highlighted in yellow are hours in which AdWords recorded conversions, but no revenue was generated. The row highlighted in blue is when AdWords recorded no conversions, but the account brought in almost $300.
This discrepancy is due to the difference in how AdWords and Analytics attribute conversions. AdWords assigns conversions using first click attribution whereas Analytics uses last click attribution. For example, lets say a person clicked on one of our PPC ads, viewed our products, went to a competing site to compare prices, then came back to make a purchase by typing our website into the search bar. AdWords would attribute the conversion to the time the PPC ad was clicked. Analytics however would attribute it as a direct goal conversion.
This is where I had to look at the whole picture and start making some decisions. We were spending a lot of money at 12am, but I didn’t want to completely turn off ads at this time. Some of those clicks were turning into conversions at some point throughout the day. I just wasn’t able to tie this information to revenue Analytics.
Ultimately I decided to decrease bids during hours with a high cost per conversion, whether revenue was generated or not. I only turned off ads during times when no conversions or revenue was reported. I also decreased bids for any hours where revenue was less than cost. Bids were left at 100% for all other hours of the day.
Even though AdWords and Analytics data will never match up, it is important to look at both before making account decisions. Using just one or the other only shows you a part of account performance. Maybe one day these two will learn to get along but until then we will have to use the data available and our best judgment when making changes to our PPC accounts.
Bethany is an Account Executive at Hanapin Marketing, a search engine marketing firm focused on generating results through PPC and SEO.
Thank you to Bethany for writing such an awesome post!
VP of Search Marketing
Elizabeth supervises the overall search division at Portent, which includes PPC, SEO, Social Media and Content. Check out her modest brag link bundle if you really want to know more: http://bitly.com/bundles/ebkendo/5 She has also written ebooks, is a ClickZ columnist, a Lynda.com author, a Dummies book author and speaks on PPC across the USA at various conferences including the SMX shows, mozCon and Hero Con. Read More