Cross-Channel Attribution and How to Cut Through the Noise

Research by Econsultancy/Google shows that 76 percent of all marketers currently use marketing attribution or will use it in the next 12 months, yet only 17 percent say they are looking at the performance of all their digital channels together. While attribution models were created to help advertisers determine which sources are driving value and to avoid double counting conversions, there are so many competing models — first-touch, last-click non-direct, linear, time-decayed — that it can leave advertisers feeling dazed and confused, especially when it comes to cross-channel.

With so many possible consumer touchpoints, it’s more important than ever for advertisers to assign campaign credit where it is due and measure the full customer journey, from awareness to conversation. The first step to tackling this challenge is understanding the difference between the competing attribution models and the respective benefits and drawbacks of each—as an advertiser, your goal is to use that knowledge to refine your overall marketing strategy.

At a basic level, there are two types of attribution models. One uses rules to credit conversions; first-touch, last-click and time decay are all examples of a rules-based approach to attribution. The other, called a statistical model, is a more advanced approach that uses algorithms to assign credit to each channel that influenced to the conversion. With the mounting pressure on advertisers to demonstrate a return on ad spend (ROAS), the attribution model you choose should provide a complete view of your marketing performance at the campaign level to facilitate data-driven decisions.

Also Read: Six Ways CMOs Can Keep Pace with Technology Innovation

The Challenge with Rule-Based Models

While rule-based attribution models such as last-click are both convenient and quick to implement, they ignore most of the steps a customer takes on their journey to conversion. Basing your ad spend decisions on these results will likely impact the success of a campaign—after all, a number of studies show that it can take up to 13 individual marketing touches through different channels to generate a qualified sales lead.

However, last-click attribution only credits the last marketing message or ad the user clicked on that resulted in the action you wanted, such as a sale. In football terms, you can consider last-click attribution as the equivalent of giving 100 percent credit to the player who scores the touchdown. It’s an incomplete picture of the path to success, ignoring the quarterback who threw the ball or the linemen who blocked the defense, and so on.

Although many advertisers still use it today, last-click attribution discards much of the creativity and credit from the build-up to scoring a “touchdown”. Advertisers will never get an accurate ROAS if their models ditch crucial data about previous touchpoints that may have led to a conversion.

Also Read: 6 Effective Ways To Drive ROI On Your Landing Pages

The Benefits of Statistical Models

A statistical model is a more elaborate way to build and track conversions. It’s more accurate because it gives credit to each marketing touchpoint a user interacted with throughout the customer journey.

Placing this into the context of a modern marketing campaign: a thought leadership webinar may influence a prospect to take a discovery call, but we can’t discount the benchmark report they downloaded earlier in the year, or the newsletter that contained a relevant customer success story, or the email campaign that generated a website visit. Statistical attribution is the only way to track all those meaningful interactions that happen along the customer journey to conversion.

Algorithms that have been set up behind the scenes define the credit allocation, resulting in a more dynamic set of results that allow marketers to learn from historical data. You can make better marketing decisions based on this more complete picture of the customer journey. For example, marketers may learn that the expensive ad placed on YouTube did not garner as many conversions as the same ad placed within Facebook’s News Feed. By having the data to back up which platforms were more effective at customer conversion, advertising budgets can be better aligned, ensuring future success.

While the benefits of statistical models are evident, the one most common drawback is that all marketing campaigns (search, social, email, etc.) must be tagged with the same tracking parameters, otherwise a marketing team may be looking at incomplete data points or double-counting conversions on multiple channels. Ensuring alignment across all marketing functions will allow for a more cohesive and successful ad campaign.

Also Read: Artificial Intelligence: The Next Frontier Of Programmatic Buying

Maximizing Campaign Effectiveness

Attribution is not just about measurement; it also impacts how marketers can drive success in the most efficient way possible. Last-click and other rule-based models are typically short-sighted and no longer suffice for mapping and measuring the full customer journey. Marketers are more likely to get the most accurate information if they deploy a statistical attribution model based on algorithms that take the entire customer journey into account—allowing them to give ad campaign credit where it is due, more accurately calculate ROAs and allocate their budget accordingly.

Also Read: Top 5 Things to Know About Your B2B Prospects