Marketing Insights
Marin Software's Official Blog

The Journey to Multi-Touch Attribution on Facebook

The Journey to Multi-Touch Attribution on Facebook

By   July 31st, 2017

This is a guest post from Vernon Johnson, Paid Social Account Manager at 3Q Digital.

It’s becoming increasingly important to understand attribution, especially as it relates to each channel and analytics platform. Marketing as a whole is about creating meaningful and lasting connections between people and businesses. But, in order to get a clear picture of how online advertising impacts real business outcomes, we need to understand how it’s tracked.

Essentially, we need to accurately measure the connections that count and drive business impact. Often, platforms measure these channels in silo, which often leads to blind spots and missed opportunities.

How Facebook Measures

When advertisers measure channel performance separately, they end up greatly diminishing overall effectiveness. Not having a view of the customer’s complete journey can stymie business decisions.

facebook

Among channels and platforms, people are the common denominator. Though Facebook is limited in determining attribution across channels, it does a great job of factoring in the actual person’s journey across the web and Facebook ecosystem.

In a world of multi-touch attribution points, Facebook wants to look at more than simply the last click or cookie data. Facebook has even reported that 22% of incremental revenue could be misattributed when using last-click models, and 54% could be misattributed when mobile spend is high.[1]

Does the Click Matter?

One of the pitfalls of the last-click method of attribution pertains to the value of a click versus the intent of the user. When Facebook looked across 478 online global campaigns, they found that clicks aren’t always a good proxy for brand results.[2] In fact, there is no significant correlation between click through rate (CTR) and brand effect metrics.[3]

facebook

For Facebook advertisers, the most effective people are often those less likely to click, and surprisingly, they’re also the least expensive. So, if you’re looking to drive brand awareness, you have to go beyond level of engagement, since a potential customer can notice and be influenced by your content without interacting with it.

A 2012 Facebook and Datalogix ROI study even found that, “more than 90% of offline sales come from people who don’t interact with ads during the campaign.”

facebook

Where Are You in Your Journey?

Understanding and getting a sense for how multi-touch attribution works is one thing—implementing it can be a process. There are essentially five stages that advertisers often find themselves in. Within the roadmap, it’s important to accurately assess where you are and begin to understand what’s needed in the next stage.

facebook

Social Metrics

In the first stage, the business is primarily concerned with “How are my Facebook campaigns performing?” and “What are the demographic interests, purchase behaviors, and intent qualities of my target audience?” This is the very beginning of the journey and also builds the entire foundation for the rest of it.

Are My Ads Working?

The next stage goes beyond simply looking at the metrics and targeting of ads and asks the question, “Are my Facebook ads driving incremental buyers and conversions?” Beyond CTR and CVR, we want to know if our ads and spend are driving incremental growth at the main KPI. This is also the point in the journey where we should be asking, “How are my Facebook ads impacting my brand metrics?”

Optimize

Once we’ve determined what impact Facebook ads are making on the main business objectives, we want to be asking, “How can I optimize my Facebook ad performance?” or, even better, “How do my Facebook ads impact offline or in-store sales?” We want to start capitalizing on what’s working and ditch what isn’t driving bottom-line growth. This is also a great stage to begin split testing, doing a conversion lift test or even a brand lift test.

Media Mix

This is the stage in the journey where we begin looking outside the world of Facebook. It’s where we begin asking, “How do Facebook ads compare to other media channels in driving business objectives across devices?” We want to know exactly where Facebook fits in from the broad perspective. This is where we begin thinking hard about properly tracking attribution between every channel. It may be time to run a Facebook Attribution Checkup or get a Reach Report from Facebook.

Statistical Modeling

In the final stage of the journey, we’re thinking hard about the cross-channel effectiveness across the entire media ecosystem. We’ll be asking, “How do Facebook ads compare to other channels to drive key metrics?” and “How effective is my cross-publisher ad spend at reaching my target audience across channels?” We’re using Facebook data and best practices to inform total media spend.

Conclusion

Multi-touch attribution helps significantly in understanding how marketing campaigns directly correlate to conversions, even when clicks don’t happen. Though the road to multi-touch implementation typically has a few steps, it’s essential that marketers get a clear picture of how lasting connections are built.

Further Reading

Marin Software provides further insights on a few of the topics mentioned here. They also recently published a guide on extending the cross-channel attribution model across search and social channels for even better performance and increased revenue, including a comprehensive reference on Google + Facebook ad formats. Be sure to give them a read.

 

[1] Media figures across 136 Facebook conversion lift studies in all industries except telecomm, May 15-Aug 27, 2015 with at least two weeks of data, positive and statistically significant incremental pixel-based conversion events, only campaigns including FB conversion pixel. Figures not shown by event type: 24-hour click models miss 6% and 24% of lead generation and registrations respectively. “Higher mobile ad spend” refers to campaigns with mobile share of impressions ÷68% (median).

[2] Nielson Brand Effect meta-analysis of 478 online global campaigns that ran between Oct 2014 and April 2015.

[3] Correlation is less than 1%.