Tag: Adometry

The Multi-touch, Multi-device Attribution Dilemma (and some companies that can help)

By Sue Brady

IAB/TagMan Study, July 2012
IAB/TagMan Study, July 2012

Over half of the folks surveyed by Atlas Institute  (Cookie Retention. Is the sky falling on cookies?) say they delete their cookies at least monthly.  This data is not unlike data shared over the last few years from Nielsen and Jupiter, and is just another factor that needs to be considered as you think about attribution. Attribution refers to how you give credit to your various marketing efforts. It’s complicated and messy, but necessary to understand. Attribution tracking, or understanding the customer journey to conversion, is key to optimizing your site performance.

So why the buzz about multi-touch attribution modeling, and what does it even mean?
Just a few years ago, marketers favored First or Last Click  attribution, meaning the first or  final behavior that lead to the conversion is the one that received all of the credit. That means that if someone searched for your product and clicked on your ad one day but didn’t buy, the next day went directly to your website on their desktop, and on the third day went to your website via a mobile device, you’d attribute 100% of that person’s ROI to either that first ad or to the mobile device. The new norm however considers multiple sources along the customer journey as a way to gain much more real and useful information.

There are three primary ‘forms’ of multi-touch models: linear, position-based, and time-decay, and there are variations on each of these. A linear model assigns an equal attribution percent to each touch. Though better than a single-touch model, linear models are still somewhat arbitrary in that they assume each touch is as important as every other touch.  Position-based models spread the attribution across the touch points, but not evenly. Typically these are more heavily weighted to first and last touch, with the remaining percent assigned to the ‘in-between’ touches. This seems to be used most with marketing efforts designed to generate a lot of leads into the funnel.  A time-decay model assigns a larger % for attribution as the consumer goes through the touch points. This type of model is often used when a brand has a special promotion running with a quick close. Slingshot SEO published a nice checklist to help you decide which type of model would work best for you. It’s on page 13 of this study.

Why is multi-touch attribution important?

  • It helps you understand your consumer’s buying process
  • It helps you distribute your marketing spend to produce the greatest ROI
  • It can help justify your marketing budget
  • It creates much more accurate ‘cost per orders’

Multi-touch models aren’t perfect. Things get messy with cross-device tracking that can undermine the integrity of many of these systems. And it’s hard to tie offline advertising into your online attribution. But even messy, there is no question that multi-touch attribution is far more valuable than last-touch or first-touch attribution.

Given the relative newness of the field, there aren’t a huge number of companies in the attribution business, but there are some and the offerings continue to grow. Casey Carey, the CMO of Adometry provided me with this list. I’ve added a brief blurb from each company’s website:

  • Adometry: They bring media types together to provide insights to guide and improve overall performance and incremental ROI of cross-channel campaigns.
  • TagMan: Their clients can manage and unify tag based technologies to produce one independent stream of clean marketing data from all channels.
  • Visual IQ: They are a cross-channel attribution software company looking to improve marketing performance.
  • DC Storm : They do multi-channel measurement, attribution and optimization.
  • Google Analytics: Released earlier this year, the GA tool allows you to choose your attribution preference (first touch, last touch, time decay).
  • Convertro: Their claim is that they provide advertisers with actionable spend recommendations so that marketing spend can be allocated in the most profitable way possible.
  • ebay enterprise, aka ClearSaleing: They claim their solution measures and calibrates across paid search, comparison shopping engines, display media, email communications, social media, natural traffic and more, and delivers recommendations for improvement both within and across channels.
  • C3 Metrics: They offer the ability to capture and make sense of billions of advertising touches.
  • Kenshoo: Their uniqueness comes from applying mathematical modeling with machine learning and algorithms.

Cross-device attribution
Here’s where things get even trickier.  Tracking multiple touches when a user is on one device is hard enough, but tracking across devices requires more attention.  Tapad has a robust product that does cross-device attribution and cross-device ad targeting.  They work with the majority of the Fortune 500 brands. bluecava also does multi-device tracking and targeting.  Whatever tool you choose, keep consumer privacy in mind because that certainly comes into play. This topic is already generating interest in Washington with the FTC.

I spoke with Chris Brinkworth at TagMan  about multi-touch and device tracking. Their company conducted a study with the IAB (Internet Advertising Bureau in the UK) on this topic and he told me that “the majority (55%) of customers who make the journey to purchasing a product have had at least 2 marketing touches before doing so.”  Where TagMan  comes into play is, they drop tags in various places (in an email, on a landing page etc.) that will identify the same customer each time he turns up. They use 1st party cookies to do this (in many cases 1st party cookies are better than 3rd party cookies because 3rd party can be readily blocked by the user). Using tags to identify a consumer based on various data points such as IP address, email opens and cookie data allows you to follow that consumer. Additionally, Microsoft and Google both recently announced their plans to move away from 3rd party cookies and towards technology that will help with cross-device tracking.

And Chris reminded me, you can’t forget about the halo effect, especially as related to offline advertising.  Here’s his example: say Macy’s is running a TV advertising campaign for sofas. TV viewers might be prompted after seeing the ads to search online for other sofa deals. As a result, JC Penney might see a spike in their search traffic. If they aren’t aware of Macys’ TV campaign, they may falsely believe that they’ve done something in search that’s having a positive impact. Companies like Optimal Social can help to tie these activities together.

In summary, attributing marketing success across multiple consumer touches, including cross-device, is key to understanding the ROI being produced from your marketing budget. Understanding the consumer’s journey will help you properly allocate your marketing budget as you move forward to accomplish your goals.

Advertisements