Many Bites Of The Apple: New Ways To Look At Attribution

"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Adam Solomon, vice president of product at PebblePost.

If a marketer tells you he has mastered summing up total campaign conversion rates across all channels, I have a bridge to sell you.

Like most things in marketing, multitouch attribution isn’t getting easier, yet we need to determine how every channel contributed to the bottom line. It’s a Herculean task, given that attribution requires the tethering of device IDs, email addresses and so many other customer data points to determine who did what along the path to purchase.

But while it’s possible to trace the graph to see how everything has connected, what you cannot see is just as – or perhaps even more – important. That is the loss that has occurred every step of the way.

Tracking Conversions: Not As Easy As We Thought

Traditionally, to determine whether a consumer converted online, marketers have relied on pixels or scripts placed on a “thank you” page that appears only after a customer has transacted.

This worked well enough in the day of the desktop, but in today’s multiscreen world, conversions are much harder to track. Between third-party cookie blocking and ad blocking, it’s a challenge to tell if the shopper about to make a purchase ever even saw an ad.

There’s also a problem relating to conversions that occur offline, for example, by phone or in-store. Digital marketers often don’t have access to the data they need to tie these sales to digital activity.

In order to even think about attribution, marketers need to have some proof that the conversion actually occurred.

Attribution Is Harder, Too

This becomes a bigger problem for marketers trying to determine which marketing touch points contributed to conversions. Having a record of conversion is obviously a necessary first step.

Multitouch or fractional attribution can help marketers begin to see how each touch point contributed to a sale. But how can marketers determine attribution across multiple touch points? How can they be tied together so that for even a single campaign, a true ROAS can be established?

A whole industry has sprouted around attribution. Companies build their business on the science of tethering cookies, email addresses and other data points together to create a journey map to attribution.

The problem is that marketers can’t see what they’re missing. What’s happening to all those known shoppers who didn’t click through or complete a purchase? Marketers are only getting half the story, and it may not even be the half they really need.

What If?

One of challenges facing marketers today in this game of connect-the-attribution-dots is that many times the data and information needed to be combined to get the complete attribution picture is in the possession of different parties, each of whom has permission to use such data for certain purposes with specific privacy-compliance restrictions.

What if there was a way for marketers and other parties in the attribution ecosystem, such as credit card companies, loyalty program providers or marketing technology platforms, to securely match, in a privacy-compliant manner, the data elements required to understand which groups of consumers exposed to a campaign ultimately converted.

It’s possible to do this now safely and securely. Experian, Acxiom, Neustar and other “safe-haven” partners enable this kind of data matching without exposing personally identifiable information to anyone who shouldn’t see it. Marketers can get the match output information they need to better understand the customer journey without sharing customer data in noncompliant ways.

Marketers who are taking advantage of these services are seeing six to 10 times the lift in conversions – not because they’re necessarily getting more conversions, but because they’re able to see them all. This gives them the ability to prove a dramatically improved ROAS. It may also help them better predict total conversions across future campaigns.

But How?

There are emerging solutions that approach the attribution problem from different angles, and marketers who are interested in gaining these deeper insights may want to consider some of them.

Identity management solutions are a starting point. Facebook, Google and Amazon were the early players in this sandbox, but now Oracle, LiveRamp, Neustar and Salesforce have entered the field. Digital marketers should consider seeking a partner in this space, particularly if their company is moving toward omnichannel marketing. This partner will ensure all the data rolls up to a single place – “one ID to rule them all.”

A clean and up-to-date first-party database is also critically important. Both quality and persistence are important here since an organization’s first-party customer data is going to be the most reliable hook into any identity graph.

A marketer’s own clean set of customer data that includes name, address and email will provide the best identifiers to conclude that engagement and conversion occurred – and how each customer got there. Furthermore, an email address or device ID will provide a far better connection than any cookie.

Finally, a reliable analytics solution has always been a marketing imperative, and that’s still true. However, analytics must align with the customer journey, mapping the route customers take from that first or zero moment of truth straight through to purchase. Tracking engagement and click-through rate isn’t going to get you there – not if you really want to see the full picture, holes and all.

For any purchase that requires any amount of consideration, the consumer rarely just sees an ad and buys the product. If conversion is an apple, the customer takes many bites of that apple along the path to purchase. Every bite counts in getting down to the core, and we must consider all of them.

Follow Pebble Post (@PebblePostPDM) and AdExchanger (@adexchanger) on Twitter.


Popular On AdExchanger Right Now:

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>