The Value of Google Analytics Data Driven Attribution/
December 3, 2015
Google Analytics provides several ways for you to track attribution, or help to assign credit for your conversions and purchases to the different channels that helped contribute.
You’ve probably heard of the most popular attribution models, like “First Click” or “Last Click.” These are fairly simple to grasp, as we give full credit of the conversion to whichever channel was the first or last method that brought them to our website.
If you imagine the last time you made a purchase, it likely involved more than just one visit to a website. Think about the research or the browsing period of your purchase leading up to your decision to buy. Shouldn’t those channels get some of the credit?
That’s why there are also some mixed models that adjust how much credit we give to different channels depending on where they fell in your customer journey. These are models like: linear, time decay, or position based. These models all rely on the same principle, let’s give a fixed amount of credit to a specific point of journey.
Attribution In The Real World
People have a knack for custom attribution modeling in our every day lives. We use it intuitively, and we don’t even know it.
Think about it. Let’s say you want to take your significant other out to dinner, and you’re looking for dinner reservations, and you ask anywhere from 1 to 10 friends. People give you different recommendations, but you don’t necessarily go with the first or last recommendation, nor should you give credit to everyone who gives you advice.
Some people give terrible advice, or give you advice that turns you AWAY from a certain recommendation. Why should they automatically get credit for their recommendation of that Belgian restaurant, just because they were part of your conversion path.
If we usually are going to have multiple touchpoints, it stands to reason that each should somehow share credit, though the exact breakdown for your site may be completely different from other sites.
Take some quick stats for example:
- In the Tech industry – 61% of tech purchases take place on that first day, only 53% of revenue comes from single-day purchases.
- In Consumer Packaged Goods – 82% of purchases happen quickly, likely smaller and simpler purchases
- In Education / Goverernment – 41% of revenue comes from multi-day purchases, but 60% of revenue comes from multi-step purchases
You should especially avoid the risk to just look at the last person you talked to, though that might be the easiest option. Maybe that last person was actually terrible advice. Maybe it was meaningless compared to something earlier. Why should we only attribute success to the last person we asked about our restaurant choice?
Source: Source – WAAFiles.org
People also regularly weigh by channels, adding some gut instinct to their logic. Perhaps they think Adwords is more important, or Social Media. Usually this is nothing more than a guessing game.
We think we get more revenue from Facebook. We get more engagement from our Newsletter. Whatever. 9 times out of 10 it seems like people build their attribution models to fit their preconceptions, and when they line up, they’re happy.
Resist the urge to listen to your gut, and let’s get to the good stuff!
What’s Data Driven Attribution?
Google Analytics Premium offers Data Driven Attribution as an additional model you can use, in addition to things like Last Click Attribution. The key here is that rather than using static proportions for their model, they use algorithms instead. That’s right – the massive Google computers start crunching your own data to determine how important each touchpoint was to your website.
Google Analytics & Adometry recently released this a detailed guide to Data-Driven Attribution, your return on investment, and how to make the leap.
“the genius of data-driven attribution is that it makes all that other data better, more relevant and actionable to improve the bottom line.”
Maybe you have a model that assigns a high value to Facebook and implies that it assists many conversions in a time decay model. If instead those users are actually less likely to convert, and what you may be seeing is them converting through something else that makes up for it, then the algorithm can determine that, and lower the value of Facebook accordingly.
It assigns value not just on time decay or channel, but on running sophisticated calculations based on any related conversions and how the one source affects others. That AdWords campaign might actually be hurting your efforts, and you don’t even know it.
It’s your data and your custom model.
Where’s the Value?
Attribution is all about looking at past performance to help influence your decisions going forward. Data-Driven Attribution gives us a much more customized picture of your website, our customers, and the journey they take to conversion.
Using Data-Driven Attribution helps you do the three most important things that Attribution provides.
- It lets you actually justify your digital spend. You’ll find those hidden gems that you had no idea were working so well, and you can pump more money into them, or starve the bad campaigns too.
- It lets you create a more effective mix of your media. You can better re-target certain users if you discover certain methods work better than others.
- Lastly you can get a better understanding of your multi-channel funnel, and ultimately plan better campaigns.
- Provide influence for your budget and strategy to help make smarter decisions going forward
Data-driven attribution comes with any Google Analytics Premium subscription, but it does require some smart setup to get the best results.
Good data is key – you need have to conversions and ecommerce tracking properly into your Google Analytics. Smart measurement will pay dividends in the long run!
You also need to make sure your customer journey is as accurate as possible. If your customers travel across multiple sites you own before completing a purchase, you need to make sure you connecting these sessions together.
Similarly, solutions like User ID for Universal Analytics can help to bridge the gap when a customer’s journey crosses onto different devices, from their cell phone to their computer, or from one computer to another. Keep the journey intact to get a more complete picture of the steps that contribute to your final goals.
Get your Data Driven Attribution running, and get a clearer non-gut based picture of your data.