In 2017, the global advertising industry is projected to reach $510 billion in spend, and digital advertising is poised to account for $202 billion in market share. With investments at this scale, it’s incredibly important that brands and agencies are accurately managing and measuring attribution effectively across all channels. Social media is one of the fastest growing categories of digital, yet it’s reported that nearly 50% of social referrals occur outside of a social network, subsequently becoming unmeasurable.
For this installment of our monthly webinar series, Katana’s Executive Chairman, Andreas Roell, partnered with Colin Zalewski, the Product Marketing Manager at social media analytics platform, Simply Measured, Inc. to answer all of your questions about how to effectively measure social attribution.
What is social attribution?
Social attribution is the process of assigning credit for business outcomes to social marketing channels and advertising campaigns. While social attribution accounts for specific networks (such as Facebook or Twitter), the full scope of attribution is realized when marketers orient their attribution models to measure how people are using the internet to communicate, how users arrive at a brand’s website and ultimately, the actions taken on a given site.
Unlike other digital channels, social attribution poses several unique challenges. This is namely because audiences are communicating with one another in the same way and at the same time as brands are trying to interject their message alongside user-generated content. As a result, brands have to be cognizant of the ads they post, their messaging and how their efforts are contributing to traffic and conversions. Furthermore, social channels force marketers to be aware of how people are interacting independently (i.e., private messaging) of a brand’s advertising and how these engagements lead to incremental traffic.
Why is social attribution important?
In April of last year, only 25% of digital marketers considered investing in attribution models as a priority, with the majority of companies solely relying on last-click before a purchase to be an accurate indicator. However, the savviest brands understand that attribution is more than just one metric, such as engagement or follower growth. Instead, brands with advanced attribution models dedicate thinking power and resources to making the connections between social and revenue.
Marketers are often focused on soft or vanity key performance indicators, like impressions or engagement, which certainly warrant some degree of social measurement. However, there has been a tectonic shift of concentration towards more actionable KPIs that measure the full spectrum of the marketing funnel across the prospecting, consideration and conversion phases. From our experience, social media marketing shouldn’t operate in a silo, but should instead work in tandem with other digital initiatives to keep accountability at the bottom of the funnel AND push revenue to the right. If marketers can see all the way downstream, then they know what to do at the top of the funnel that will lead to bottom-of-the-funnel outcomes.
User generated content and social sharing has been the catalyst for digital word of mouth, and this has in turn generated sharing activity, energized sales and has even impacted direct response. Social adds value to all overall marketing functions, affecting 1) budget and growth and 2) campaign optimizations and impact. By implementing a social attribution solution, your brand can start taking credit for social activities that drive conversion metrics, illustrate the bigger business impact of social and optimize content effectively.
How do you set up social attribution?
Marketers can approach social attribution by deferring to tracking pixels provided by social networks (such as Facebook) or by leveraging UTM parameters that are associated with more traditional web analytics tracking practices. For example, marketers implement UTM parameters to track user activity for display advertising.
With UTM tracking transcending into social attribution, there is significant value in how social attribution can ‘fit’ into the tracking and measurement in place for other channels. The overlaying of tracking can tell marketers how users came from a social media post or ad to a brand’s website; it can also inform which actions were taken. Unfortunately, marketers are challenged with understanding the complete user path when the social activity happens outside of their control.
What are challenges of social attribution?
Challenge #1: Dark Social
Imagine the following scenario: John is served an ad for soccer jerseys. He knows that his friend Jane likes soccer too, so he private messages her the ad’s URL link in a Facebook message. Jane then clicks the URL, peruses the website and ultimately converts.
It is not Jane’s responsibility to attach a UTM parameter to her name so that a brand can keep track of her actions. This scenario proves that there are still measurement complications that exist when the everyday user gets involved and starts to use social media privately.
This is a prime example of dark social, which is defined as website traffic that originates from a shared link but does not pass a referrer. A referrer is essentially an indicator that comes through when somebody visits a website and tells the web analytics tool where the user came from and where they were last. When a user comes through a link that is shared (and in some cases, it does not pass a referrer), web analytics might not necessarily know how to classify it. As a result, it ends up categorizing that visit and ensuing activity on your website as attributed to direct traffic.
Dark social commonly happens through:
- Private messages
- Side effect of some of the ways that some social mobile apps pass through information when driving visitors
Tracking dark social:
The only true solution for tracking dark social effectively is to do URL ‘uniquification,’ or appending a unique code to a user’s URL.
We then match it with referrals; where, what and if there is a referrer. If there isn’t a referrer, you can go through a deduction process to eliminate traffic sources and ultimately attribute the correct source.
In summary, dark social can essentially be tracked by:
- Put a code on your website the uniquely identifies URLs
- As people come to your website, that URL is assigned to them
- As users share the URL (in private messages, for example), it’s then matched back to them
Challenge #2 Attributing Value to Actions a User Takes on a Website
Assuming that a user lands on a website and completes a lead form, marketers might ask the question, “how do I attribute value to what my brand does on social?” Foundationally, this is the first step for marketers as they begin to think about social attribution. If a website requires any form of user action, whether it be signing up for an account, filling out a lead form or purchasing a product/service, some marketers might not fully understand how to assign a value to each action.
In order to strategically approach social attribution (and learn how to attribute value to actions initiated by social media), we recommend working with a strategic digital partner, such as Katana, that knows your business and what actions your brand is trying to drive people to take on your website.
What are the types of attribution models?
Single touch: All credit is assigned to a single touchpoint or specific source. This type of model is the most common and easiest to use, but also the most inaccurate.
- Last touch: All credit for a conversion is given to the last source of traffic for a session where the conversion happened. This is the most commonly used attribution model and most flawed.
- First touch: All credit for a conversion is given to the first visit a consumer ever has with a specific site.
Multi-touch: All credit is assigned to multiple touch points or sources.
- Equal weight or linear: Credit for a conversion is divided evenly across each touchpoint a user has prior to converting.
- Position-based or starter/player/closer: A slight improvement of the equal weight model, the position-based model gives a fixed amount of credit to the first touchpoint and the last touchpoint of a visitor, and then divides credit amongst the remaining touchpoints. Often, “40% credit is given to the first and last touch points, and the remaining 20% is given to the other touchpoints, but many organizations will use custom values for each of these touchpoints” (source: Simply Measured, Inc.).
- Time decay: This approach recognizes that later touchpoints are often lower in the funnel and are more likely to drive conversions. Therefore, credit is divided across multiple touch points based on how recently they occurred prior to the conversion.
- Custom: A custom approach is utilized by organizations or companies that have deep insight into their customer’s journey. Custom values are designated to each touchpoint depending on what the brand knows about the efficacy of different channels and customer journey stages.
What is the reliability of social attribution data?
Attribution teaches us a lot about people and what they do as brands show them different types of messages throughout the customer journey.
Data reliability a huge challenge for a lot of marketers, but it can realistically be solved through a proper implementation – which again requires lots of resources and expertise, especially if you’re a marketer at an S&B-sized company.
Tips to Get Started with Social Attribution
- Get off the bench and START with social attribution.
- Find a solution that can start looking at attribution beyond just immediate actions, and instead considers how content is being distributed by website visitors among each other.
- Track the dark side of social media.
Without proper attribution in place, social media’s contribution down the marketing funnel isn’t realized, which limits your ability to optimize against data. However, with a thorough understanding of what social attribution options are available and how to properly implement them, your brand will be better equipped to answer your CMO the next time you’re asked, “How many sales came from our last Facebook campaign?”