Why is cross-channel attribution so elusive to marketers?

Gone are the good old days when marketing dealt with one or two channels to reach potential customers. Thanks to the digital evolution most are looking at many, many more. And for every channel they use, marketers need to understand performance - is it performing as expected? Is it contributing to conversions?

Performance metrics are usually only based on a single marketing channel. But what happens when a customer is converted not by a single channel alone? Understanding cross-channel attribution is something every marketer needs to do, but many can’t.

First, a definition so we’re all on the same page.

In marketing, Attribution is the process of identifying a set of user actions (“events” or "touchpoints") that contribute in some manner to a desired outcome, and then assigning a value to each of these events. Marketing attribution provides a level of understanding of what combination of events in what particular order influence individuals to engage in a desired behavior, typically referred to as a conversion.

IDG Connect did a recent study on 250 marketers looking to understand the challenges they face getting unified performance data across all marketing channels and understanding attribution. (Note the study was commissioned by ConversionLogic, a marketing attribution service provider).

From the study, Seeing the Forest for the Trees: Unified Analytics for Modern Marketing (free registration required):

In the past, marketers often were unable to quantify the exact impact of their programs on the company’s success, except perhaps when running direct marketing on a single channel and lead generation campaigns. Now marketers are expected to continuously improve and optimize their programs across all channels, know the effects of any one channel on the rest and report results that indicate how each activity is supporting business objectives.

The challenge of unified performance data

Here’s a look at all the different types of performance metrics many marketers deal with today:

Image from IDG Connect Study, Seeing the Forest for the Trees: Unified Analytics for Modern Marketing

The IDG Connect study found that 39% of marketers will use six or more channels in the next two years. This is a mix of both online and offline, which makes the challenge of attribution even harder.

What needs to happen is all these data must be combined and analyzed to give the marketer an understanding of which channels are providing the best results, where they fit in the conversion process (at what point are they typically used and how), and what is the best combination to increase conversions.

According to the study, 65% of marketers are required to report on customer satisfaction and customer insights, while 59% must report on sales opportunity wins/losses. To do that right, the marketer must look across each channel and understand its contribution. Then they need to optimize marketing strategies and spend on channels appropriately.

But the data. Marketing is still stuck with data challenges. Accurate data collection and centralization continues to be the biggest problem marketers have. Think about all the different applications that provide data: marketing analytics, web analytics, marketing automation, sales automation, CRM, ERP, BI and more. Many marketers are not even sure the data they do have is accurate.

As more channels come online, marketers need to find a way to measure performance across channels and the impact of each channel in the buyer’s journey to acquisition.

One solution to cross-channel attribution

There’s no one “solution”, but there are technologies in the market that can help. ConversionLogic is one of these, as are Google, MarketShare, AOL/Convertro and Visual IQ, among others.

ConversionLogic’s founders come from Visual IQ. I spoke with Alison Lohse, co-founder and COO of ConversionLogic, to better understand how they are helping marketers with cross-channel attribution.

She talked about how marketers are using multiple metrics against multiple channels making a holistic view pretty much impossible. They need a centralized view of measurement performance. Part of the challenge, she said, is finding their way through all of the data they have in front of them, and coming up with a solution that no single vendor they work with is providing.

Currently, most marketers either use a single vendor solution, or they use an off-the-shelf channel attribution solution that provides attribution models such as Last Click, or follows a rules-based process for attribution. Neither of these is the best option, nor do they provide cross-channel attribution. Almost half of the marketers in the study said they know they need a solution that will help them understand cross-channel attribution.

Getting buy in for a solution such as ConversionLogic (CL) is hard. The study shows costs, lack of skills, centralized data collection, and accurate data are some reasons marketers are slow to purchase.

ConversionLogic focuses exclusively on unified cross-channel attribution. In business for two years, they are creating what Lohse calls second generation cross-channel attribution. She said they made choices about the infrastructure early on that ensured the data was available in real-time so customers could get the insights back as quickly as possible.

They wanted their solution to be faster, adaptable and transparent.This is the approach they took in their attribution model, using their proprietary Ensemble Method. The Ensemble Method is predictive; there are no rules to set up. Instead, it looks at multiple algorithms as opposed to just a single one, ensuring that the best algorithm is selected when the data goes through for analysis:

Image from ConversionLogic.com

CL gets data from client’s existing systems (ad server, etc.), through pixels where they capture data themselves, or through vendor system partnerships. They also capture conversion behavior from different places, such as a mobile app or the website. Even data from offline sources such as TV and print are brought in. All of this data combined provides a unified view of the customer.

The configuration for every client is unique. The plumbing is there for many integrations, and where it isn’t, they do custom tweaking to get the data.

One of the biggest requirements for marketers is having technology that is easy to implement and easy to use. CL, said Lohse, can be used by all types of people in the organization to get the top level insights needed in a portfolio view, or drill down logically for more detailed information in a channel.

The majority of CL’s clients are B2C (it’s about at 70/30 split), with clients in e-commerce, telecom, financial services, education and retail gaming. For B2B, the dataset can be a problem. Lohse said that a lot of the data a B2B company has might not the the right data. Many of the channels B2B use, such as events, may not be trackable, or may not have the view of the customer needed to do the analysis.

Final thoughts - cross-channel vs. omnichannel attribution

Attribution is important whether the company is B2B or B2C. But there is a different approach to attribution if you are looking at cross-channel (more B2C centric) compared to omnichannel (more B2B centric), says Lohse.

According to Lohse, omnichannel marketing takes in more business-centric data than marketing data. Business-centric data can be more difficult to measure, whereas marketing data looks at things such as acquisition media - media that gets customers to the experience that drives behavior.

Omnichannel is a focus on the entire customer journey and how each stage of the journey builds the customer experience. It’s about more than just the acquisition, more than where the marketing budget needs to go to win new customers.

Overall, it seems we come back to the same challenge over and over regardless of what a marketer, service and support team, or sales team needs to do. It’s all about the data. How to capture it, sift through it to find the good stuff, and analyze it to get the best insights.