Search Insider Summit 2012 panel on attribution and integrated digital marketing

Marketers consider attribution to be the holy grail of digital marketing. Some liken it to an oasis that constantly eludes them despite well laid-out plans, shiny technologies and catchy taglines to rally everyone around the cause. Regardless of your metaphor of choice, this much is clear – the world of marketing recognizes the need for attribution, knows that we have made progress and faces challenges ahead.


Our CEO, Jim Yu, was part of a thought-provoking panel at the Search Insider Summit 2012 in Florida in late April. The discussion, involving some of the best minds in marketing, describes obstacles to attribution and suggestions on managing these obstacles. We strongly recommend watching the video of this very engaging discussion.


If you are new to the world of attribution, you can read the quick primer below. If you know it all, you may want to skip to the bottom and see the video.


Attribution = which channels get credit and how much?

As users, we consult multiple information sources across channels and devices before we make a purchase online. Let me walk you through a real example of how I recently bought my beloved Brooks B17 bicycle saddle/seat. Note that what you see below is a super-simplified version of the steps I took to buy the Brooks seat:

  • Searched on smartphone – I used the default Google Search on my Android to search for ‘touring saddles’.
  • Clicked first organic result – The first result was a discussion forum site where the consensus was that that ‘Brooks B17’ was best for bicycle touring. I decide to buy it.
  • Called vendor and made purchase – I also see an REI display ad on the forum site advertising a deal on the Brooks B17. I call the toll-free number in the ad and buy the saddle for $140 by phone.


In a world of perfect attribution, REI would know exactly how much influence the 3 channels (Google Search on Android, display ad on bike forum site and phone) exerted on the sale. REI would have answers to questions like ‘What prompted the purchase?’, ‘Was it directly related to the ad or was it a combination of factors that drove the purchase?’ and ‘How much dollar credit does each channel get?’. Such insight enables marketers to measure the ROI they are getting from each channel, understand what campaigns are paying off and what aren’t, and know where to spend. That’s attribution in a nutshell!


What blocks attribution?

The video below actually talks about an exhaustive and arguably discouraging list of issues blocking attribution. Here is a sneak preview of these issues, which can be summed up by ‘obstacles to truly integrated digital marketing‘:

  • Technological – Tracking user activity across multiple channels on different types of devices is challenging. Many marketers use more than one technology to track these, and consolidating them to get a single view of what drove conversions is not easy.
  • Data issues – An interesting perspective that came up in the discussion is that we have too much data available to us due to which we may be over analyzing everything to death.
  • Structural and organizational – Despite the fact that user behavior has evolved and spans many channels (search, paid, display, phone, email), budgets and teams are siloed by channel. Also, marketers might not have the incentive to understand attribution outside their areas of responsibility.

There are best practices in attribution

Don’t let the hurdles above discourage you – the distinguished gentlemen in the video below have several great ideas to overcome them. We do have a strong point of view on this topic, which Jim Yu does talk about during the discussion, and best practices on getting there. All this is based on real results our customers achieved by tracking attribution in a pragmatic manner. We’ll save that for another day. For now, enjoy the panel discussion!




If you found this post helpful, please feel free to share using the buttons below. Thank you!

Take our digital marketing quiz to test your knowledge.