Google Steps It Up With Data Driven Attribution

More news from Mountain View on Tuesday, where Internet powerhouse Google released the much-anticipated Data Driven Attribution (DDA) feature for its Premium users. The release of Google’s DDA approach comes as no surprise to the analytics and measurement community. The world of attribution measurement is constantly evolving and new attribution approaches, new players, and new tools regularly enter the market, enabling marketers to select the right attribution tool for their business needs. It was only a matter of time before Google released a persuasive, more advanced measurement offering.

First, the Data Driven Attribution feature is only available for Google Analytics Premium users. It has several notable features worth highlighting:

  • Google DDA’s approach is statistically driven methodology. Google’s DDA approach is a huge improvement over its rules-based Attribution Modeling tool (which is available for FREE for Google Analytics users). The DDA approach uses probability modeling to best estimate the values of each interaction. The approach itself is transparent, understandable, and Google is extremely open about how it calculates the value parameters.
  • Google’s DDA feature can include non-Google data. Yes, you read correctly. The DDA feature leverages the Google Universal Analytics capabilities. Universal analytics capabilities allow marketers to upload a richer set of data, including offline data, mobile app data, and richer data through more accurate tracking. This enables marketers to track converters and non-converters, from online to offline —this is huge.
  • Google’s DDA feature integrates seamlessly with existing Google products. If you’re a Google shop, it’s easy for you to integrate your existing data sources into the DDA feature. Google does a great job with creating usable interfaces, and the ease in which they can transfer data from one Google product to another makes this feature extremely appealing to Premium users who do not want to invest in another measurement tool.
  • Google’s DDA’s feature is now a serious option for marketers wanting a more advanced attribution measurement approach. Google’s advanced approach demonstrates that it is serious about measuring more effectively. It’s a serious option for marketers seeking a more advanced attribution approach. To be clear, I certainly think there is room for a wide breadth of attribution offerings, but it’s going to force the attribution measurement experts to be more innovative with their approach, and their insights, which for users, is a good thing.

But it’s not all puppies and rainbows for Google; it’s going to face some major obstacles. It needs to:

  • Convince organizations to send it non-Google data for a more accurate attribution measurement. This is the biggest challenge for Google. I anticipate that the DDA feature within GA Premium will expand its capabilities and depth. In order to do so, Google needs to convince organizations to send it CRM data, offline data, and data from other sources into Google DDA to gain a more robust view of the customer and conversion paths. This will be a huge hurdle for Google to overcome, as consumers are increasingly concerned about their privacy and businesses don’t want to be perceived as Big Brother-ish.
  • Develop more robust services and support for the Data Drive Attribution approach and feature. I’ve been shouting that attribution is simply not a plug-and-play feature within a tool (or a standalone technology for that matter). Attribution is complicated; it has major implications on change management within an orgnizaiton, how firms measure marketing effectiveness, how consumers are valued, and how marketers purchase media and build marketing strategies. Premium users need support to fully benefit from the DDA features and insights. They need help to understand results and provide actionable recommendations. I’m not saying that attribution is a full-service offering, but I’m not saying it’s a complete solution either. Any type of measurement insights and optimization engine needs to have some support services and analytics expertise. And while GA Premium has a service and support team, we have yet to understand the extent of that service and support in the context of attribution and consumer analysis and optimization. 
  • Assure DDA users that Google can be trusted to sell media and provide rich, unbiased analysis. Ahh…yes, the elephant in the room. Google sells media to marketers.  So, this begs the question: how can Google also provide an analysis of media performance and complete, true ROAS without being biased and win the trust of their Premium users? I think this will be another great hurdle for Google to overcome.

Further, I don't see Google's development of DDA as a threat to the existing attribution competitive base. On the contrary, we are seeing more and more service and technology providers developing advanced attribution solutions and offerings. Google is a bit late to this advanced attribution party, but it is making a splash with its approach and ability to intake different types of data (and we need to wait and see on its capability to do that). I do think there is room for everyone, because every marketer has different needs. But Google's in a great position to be a powerhouse in this very competitive market.  

It’s an interesting, but expected move for Google to come to market with an advanced attribution approach. But who wins and who loses?

Winner: The Google Premium user. Premium users will reap the immediate benefits of the Data Driven Attribution feature. It’s available to all Premium users NOW, so I expect many expert users to test this new measurement approach. And if you’re a Premium user who’s using this feature, please share your experience with me!

Loser: The Google Analytics user. While GA users can still leverage the Attribution Modeling Tool through the free, GA base product, marketers seeking an attribution approach with a little more power may want to invest in Google Premium for the DDA functionality . . .or go elsewhere. Not having access to the Premium version may turn off non-Premium users, leading them to turn to the competition for a robust competitive attribution offering. And the attribution vendor market is a fierce group of analytics and marketing experts that understand consumer purchase behavior, marketing performance, and the complex data and media buying ecosystem.  

Google made the right move in developing a more robust advanced attribution approach. It’s positioning the DDA approach as a way to better understand customer purchase behavior and touchpoints, which is a uniquely different value proposition than the go-to-market attribution message of generating efficiency by making better marketing and media budget allocations. However, it faces many challenges in a world where consumer data and privacy awareness is at the forefront of everyone's mind. 

It’s a very exciting time for us measurement and analytics geeks. I hope to see more measurement and insights innovation coming from the market in the next year.

Comments

Thank you Tina for such a

Thank you Tina for such a well written article. I agree with you that so far DDA is the newest and the best thing in Attribution Modeling industry. Now you don’t have to be statisticians, to create and implement an attribution model for your visitor’s paths. I think DDA provides a very big advantage for Google Analytics over their competitors. There are different features in other analytical tools, but not one of them come even close. Also, another cool thing is the Model Explorer feature lets marketers dig into attribution model if they want to understand how credits were distributed between touch points and as you mentioned, even lets them manipulate the distributed percentage if by some reason they disagree. Another cool thing is that you can upload offline data, but again as you mentioned, there some privacy issues, but still, it’s cool to have this option. Regarding a price, well many people complained that Premium version is very expensive, I actually agreed with them before DDA came out, now I think it’s totally worth it to spend that much money. Also, let’s not forget that to hire someone to create an attribution model will take not only money, but the time too, and if you need to revise it or make some changes to attribution model, you have to go through the same process again, when GA does it for you.

Ecommsprak.com