we've been using utm tracking for links in the email. but when we compare the results between marketo's linked clicked vs users in google with the match utm. It's about 50% lower in google.
Any idea why this might be the case?
First, realize that by comparing link clicks (logged by Marketo's tracking server, whether or not the subsequent page is loaded) with GA pageviews, you're not comparing apples and oranges -- more like apples and armadillos.
Link clicks will be registered by mail scanners pre-following your links; GA is typically excluded (deliberately) from this security pipeline. You can read more about the scanner behavior throughout the Community.
For a more valid (though still apples-and-oranges) comparison, look at GA pageviews and Munchkin pageviews.
Thanks for your response. I don't expect the number to be exact, but currently it's about 50% difference. I realize that link click tracking and GA user with utm will never be the same due to one is pre page load and one is after but I think those two should be relatively similar (within 10% delta).
Can you also explain why GA pageview is a more accurate measure? We want to identify # of unique users that visited, and pageview will double count users based on number of pages they saw in a session no?
Your concept of 10% delta is not reasonable, though. You can make future assumptions based on current observations, but there's nothing to suggest that 10% should have been expected here. What you see is an accurate accounting of how these extremely different technologies measure these starkly different types of traffic.
The GA pageview counts for a given UTM parameter are roughly congruent to the Munchkin 'Visit Web Page' events matching that same pattern. (The difference being each product's web spider detection.)
At the end of the day, when we measure impact, only one of these numbers can be reported. With 50% delta it can make a huge difference.
I'm still not sure why even though these two technology while different should be expected to produce results this different. at the end of the day, they are both trying to give campaign runners and idea of level of engagement, and when the numbers are this different, only one of them can be closer to reality.
You should be suspicious when technologies that are so drastically different in what they're measuring do appear to give similar results! The default must be that they will give different results, as they are measuring different things.
Again, the Clicked Email activity means any client app (could be an automated scanner app or a human-operated app) made an HTTP request for the tracked link as printed in the email.
A GA pageview is similar to Munchkin Visited Web Page except that it uses a more sophisticated database/heuristics to see if the hit was from a search engine or other crawler/spider, so overall will exclude more hits.
Can you see how the first one is so different from the second two?
Yes, the common scenario is that mail scanners pre-follow your links. Such scanners try to omit web analytics calls; they cannot (by definition) omit the original click. I mentioned that up top.
We have run into the same issue of the two being different. We use the UTM codes to bring in different values from the campaign into our DB.
I would suggest using the Marketo Munchkin code mainly because you are able to track their path from sent email(s) to opens, clicks, web sites visited, etc.
But I agree with Sanford about GA vs Munchkin. It also all depends on what you are looking to do and how you want to report on the path of the lead.
do you happen to find a big delta between links clicked and munchkin page visited? what could be some common reasons one doesn't lead to the other?