Marketing attribution can be a complex, costly, and overwhelming concept for many marketers, data analysts, business owners, and executives to wrap their heads around. It has become an increasingly important topic in today’s digital landscape as it provides invaluable insights into how an organization's campaigns influence conversions. 

However, understanding the ins and outs of marketing attribution is easier said than done. It requires deep analysis and often involves multiple teams working together to deploy this strategy efficiently. In this blog post, we'll discuss what marketing attribution is, how it can be used successfully, and provide specific advice on avoiding common mistakes that lead to failure when implementing an effective marketing attribution strategy.

What is ROAS (return-on-ad-spend)?

To better understand why companies use marketing attribution tools or build them in-house, we need to understand the return on ad spend, or ROAS for short. 

ROAS is, simply put, the amount of money that is returned for every dollar spent on a campaign.

ROAS is the simplest way to determine if campaigns are successful financially. Traditionally, a good ROAS is a 3:1 return on investment though some might strive for a 10:1 return depending on the market a company is in. Check out this article for a more in-depth description of how to calculate ROAS.

Generally, marketers responsible for running a company’s campaigns will calculate their ROAS, but they are hardly the only ones taking a deep look into the metric. Everyone from CEOs to Business Operations and even venture capitalists will keep a watchful eye on a company's ROAS as it’s a critical indicator of its overall health and success. 

How marketing attribution influences ROAS (return-on-ad-spend)

Now that we’ve highlighted the importance of ROAS and how it relates to a company’s success, it's even more important to understand how ROAS is impacted, sometimes negatively, by marketing attribution models and tools. 

Marketing attribution, in this context, is the identification of a set of user actions that contribute to a desired outcome and then the assignment of a value to each of these events. In its simplest form, it can look like this: You check your email inbox when you see an email for the latest and greatest golf clubs. You click into the email, see there is a 10% off discount code, and rush to the site to purchase. 

In this scenario, the email was the attributable action that led you to purchase the golf clubs, and you would get 100% of the credit for the conversion. Anyone that's familiar with running marketing attribution models, knows this is seldom how a typical consumer journey happens, as there are often multiple channels or touch points involved before conversion and that’s where attribution becomes a challenge. 

Changes such as the increasing number of marketing channels, the impending depreciation of third-party cookies, and rising privacy concerns of consumers can create a challenge when attributing conversion influence to specific channels muddying an accurate ROAS calculation.

How common marketing attribution strategies work

Marketing attribution can be viewed as the art of piecing together a customer’s journey to understand their decisions on their path to conversion fully. The “art” of this puzzle is knowing that some channels hold a much greater influence than others, and attributing a specific portion of the conversion to each channel requires more than one standard model. 

While there are many models out there, we’ll focus on some basic ones that attempt to cover a broader spectrum of conversions. 

First-touch attribution

First-touch attribution is a model that awards 100% of the conversion credit to the first “touch” or marketing channel a consumer interacts with before converting. An example would be an interaction with an Instagram post five days before conversion. The consumer might have read an email or clicked a banner ad after seeing the Instagram post and then converted, but a first-touch model will still apply 100% of the attribution toward the Instagram post. 

Last-touch attribution

Last-touch attribution follows the same practice, except it awards 100% of the conversion credit to the last “touch.” If we look at the same example above, the banner ad will receive the conversion credit instead of the first or intermediary touch points. 

Multi-touch attribution

Multi-touch attribution modeling reviews all touchpoints across the customer journey and assigns a portion of the credit to each one so that marketers can identify the impact of each channel or campaign and how much it contributed to a sale. Multi-touch attribution models are much more complex than first or last touch because of a set of rules the model will follow, which determines the amount of conversion credit each touchpoint receives. 

• The lookback window example: “A lookback window is a period you’re willing to consider when assigning credit to certain touchpoints. For instance, if you have a 6-week lookback window, you will consider assigning credit to all touchpoints within this period. Anything before this period doesn’t receive any recognition. The period assigned to your lookback window is often based on how long your sales cycle is.” (Source)

While finding which attribution model best suits your business and its structure may require a bit of thought and research, the real challenge lies within how marketing attribution models track and identify users across channels. 

Downsides and limitations of common marketing attribution models 

Marketing attribution models are only as good as the data they are pulling from. Which, in turn, affects how accurate your ROAS reporting is. Attribution modeling could be better, and gaps in user journeys and duplicate users can all affect the outcome of your return on ad spend. 

With marketing attribution models primarily relying on cookies and local storage to track users across channels, there are several instances in which the user can break this.

Cookies & local storage

Cookies and local storage fail to recognize a returning visitor when users clear them or are blocked, use incognito browsing, or visit through a webview browser via Instagram or Snapchat. In those instances, a single user could interact with multiple channels within the same consumer journey but will be identified multiple times. Outside of data collection, the model’s logic can cause issues. 

Lack of conversion journey insight

Last-touch attribution needs to acknowledge insight before the immediate action before conversion. The same can be said for first-touch models, just vice versa. Today’s modern consumer is exposed to multiple channels through paid channels like social and search in addition to the amount of research on a particular product or service they are doing. 

Build vs. buy

Some companies or departments might create their own marketing attribution model instead of paying the typically hefty subscriptions that come with a SaaS-based product or service. This can present several challenges not only from a cost standpoint but also from an internal point. 

While potentially saving money from another inclusion to an already costly tech stack, time spent creating and upkeep of an marketing attribution model might pull valuable resources away from other areas of the business. Additionally, when a model does come to fruition, deficiencies in certain campaigns like direct mail, for example, might cause that team to ignore the model or insist it’s not working correctly. 

While the build versus buy question will always exist in the world of SaaS (software-as-a-service), and elsewhere too, for that matter, attribution’s main woes will always derive from the quality and consistency of data fed into the models. As concerns grow about cookie deprecation and blanketing privacy laws impacting the ability to attribute credit to channels accurately, there are ways to get ahead of the curve. 

Future-proof your marketing attribution with device identification 

Accurate device identification can remedy many of the woes of attribution and return-on-ad-spend (ROAS) reporting caused by cookies, new browsers, and users attempting to conceal their identity with VPNs. Sometimes, a user might visit a website or view an ad and return to that website and convert in incognito mode, another browser, or with a VPN. With an accurate device ID, you can associate logged-in users with anonymous sessions to ensure accurate identification of returning unique visitors. 

Even when users visit a webview link from within an app, accurate device identification can attribute social media app webviews, like Twitter, Snapchat, or Instagram, with browser sessions on the same device. 

Device identification wins where cookies and other methods of third-party data cant and can easily be incorporated into user analytics, reporting, and attribution models. 

For more on Fingerprint's device identification and its application for marketing attribution, check out our website. If you’re interested in speaking to a sales representative, click here.