Multi-Channel Attribution – A Story of Ice, Fire & Last Click Attribution Death

By Cassie Oumedian | @cass_oumedian | Associate Director of Services at Hanapin Marketing

Whether we like it or not, multi-channel attribution is part of our digital marketing world – just as Game of Thrones has become part of mine. Understanding where each piece falls in the conversion funnel and how much credit should be given at each step, is crucial in maximizing conversions. We all know this, right? But that doesn’t mean it’s any less confusing!


Well today is your lucky day! Below I will break-down the basics of multi-channel attribution and provide some tips for determining which method is right for your account while referencing some of my favorite G.O.T. characters.


What is Multi-Channel Attribution?


Attribution allows advertisers to understand the touchpoints a user makes during the conversion process. For example, a user might follow this path in their conversion process:


Non-Brand (PPC) > Display (PPC) > Social > Brand (PPC)


Understanding the path is important because looking at a last click attribution model would give all credit to the branded PPC campaign and none of the other channels. Last click is bad. Very bad. Understanding the various pieces in the funnel can help prove the importance of channels that don’t get a lot of credit. By reviewing the entire funnel you can see where credit is due and help prove a case for expansion, budget or more resources.


Attribution Platforms


Multi-channel attribution has been available through Google Analytics for sometime now. Google Analytics is great because if your site is tagged correctly, you can see the impact of various channels beyond Google. Fellow blogger Amanda wrote a post awhile back on how to use this tool.


AdWords has also jumped on the attribution bandwagon. While Search Funnels have been available for some time (you can find by accessing “Tools” > “Conversions” > Search Funnels”), you can now access attribution information directly in the main interface.




This is all great! Two tools that provide insight into the conversion funnel right at your finger tips. They even have comparison tools where you can review the various attribution models. We have all this information but what can we do with it? How do you know which model is right for your account? It’s a tough question as almost every purchase journey is different so before determining which model is right for your account, it’s best to understand how each works. You’ll also need to understand your ideal customer and then come up with a plan that fits your account’s needs.


Attribution Models


Example: A customer is looking to purchase Season 4 of Games of Thrones. She clicks on a GDN ad while reading her favorite gossip blog. A week later, she clicks on a Facebook Newsfeed ad with season 5 spoiler alerts. The next day, she clicks a paid search ad, leaves, and then the next day comes back and makes a purchase directly on the site. The purchase cycle would be as follows:


Display (PPC) > Social (Facebook) > Paid Search > Direct


LastCLick Last Click attribution model, unfortunately, is the norm. Most standard reporting gives all credit to the last click, in this case – Direct. The problem is, there were 3 other touchpoints that lead to the conversion that are receiving NO CREDIT for the purchase. If this advertiser made decisions based solely on the last click CPA by channel and cut back on Display, Social and Paid Search, their conversion traffic would tank. Last click is dead, people. Straight up, “Red Wedding” dead. Let’s move on.


First Click  First Click attribution model, the first touchpoint — in this case, Display — would receive 100% of the credit for the sale. I personally think this model is worst than last click. If an advertiser gave 100% credit just to first click you would have no idea what actually got them to convert. I prefer first click about as much as Sansa prefers Joffrey.


Linear Linear attribution model, each touchpoint in the conversion path. In this case Display > > Social > Paid Search > Direct — would share equal credit for the sale. Linear model is better than first or last click, but all in all is mediocre. Vanilla. About as exciting as Dany in season 4 GOT…zzzzz.


TimeDecay Time Decay attribution model, getting better. In the time decay model, the touchpoints closest in time to the sale or conversion get most of the credit. In this particular sale, the Direct and Paid Search would receive the most credit because the customer interacted with them closest to the conversion. The Social Network channel would receive less credit. Since the Display interaction occurred one week earlier, this channel would receive significantly less credit. This beginner approach is good to test, especially for lead gen or conversions that have a longer research cycle. I like to think of time decay attribution like Littlefinger. He’s always focused on moving up and then, BAM! He pushes Lysa Aryn through the moon door to her death.


PB Position Based attribution model – now we’re talking. Position based attribution by default, credits 40% to the first and last interaction, and the remaining 20% credit is distributed evenly to the middle interactions. In this example, the Display and Direct channels would each receive 40% credit, while the Social and Paid Search channels would each receive 10% credit. You can alter the percentages, and I personally, like to adjust the position based slightly to allow more credit to the last click vs. the first click.


This model seems to work well for eCommerce but be sure to do your homework before adjusting and see what makes the most sense for your conversion process. I like to think of Position Based as Tyrion. We see a lot of him in the beginning. He kind of lays low in the middle seasons and then blows our mind in Season 4.


Custom attribution model. This model is for the advanced peeps. Before implementing a custom attribution model, I would encourage you to do your research and test with caution. Kind of like Bran. That guy is into some crazy stuff and needs to understand his power before using them for good. Once you feel comfortable, you can create your own custom attribution models directly in Google Analytics and compare.






Attribution can be complex and confusing. The one thing we do know, however, is that last click is dead so it’s important to understand which other attribution models are out there. If you want to hear more advanced attribution discussions and case studies, join us Thursday 8/28 at 12pm EST for the free PPC Attribution: Bury The Last Click and Move On webinar.