Right side up logo



An Overview of Marketing Attribution


December 14, 2022


December 20, 2022

Right Side Up (RSU) recently hosted a panel discussion webinar on the evolution of marketing attribution strategy. It was hosted by Ryan Jones, head of project strategy at RSU, and featured panelists and growth marketing leaders Rian Long and Chelsea Cramer, who have a combined 35+ years of working with companies of all sizes across a wide range of industries and marketing programs. Watch the full webinar for more attribution insights from our experts.

It’s no secret that creating a solid marketing strategy is crucial for the success of most businesses. But in order to know whether your marketing is working (and how it’s working), you must also implement a strong marketing attribution strategy. Attribution is the process of measuring how well our marketing efforts are working based on a set of assumptions. It’s essentially a way of understanding and mathematically quantifying the inputs that drive valuable outcomes for a business.

In a perfect world, we’d be able to track and understand the contribution of every customer touchpoint. We’d know which TV ads a customer saw at a friend’s house, how (and with whom) that customer talked about the brand from the ad, and the relative importance of these channels in influencing that person to buy that brand’s product. In reality, that’s rarely the case. It’s impossible to know every touchpoint and how each one changes the customer’s perception of value and buying behavior. And many interactions can’t be easily tracked through a click or a pixel. Because of that, we have to make assumptions (while accounting for potential biases) about what goes on in the mind of a customer, and those assumptions form the foundation of attribution strategies. 

Marketing Attribution at Different Stages of Growth

Like most aspects of marketing, attribution is going to look quite different for an early-stage company compared to mature or late-stage company. Budgets, staff resources, and data volume constraints limit how sophisticated a company can get with its attribution strategy. 

Attribution at an early-stage company

For early-stage companies, typically those with $50K per month or less in paid media spend, tools like Google Analytics provide enough data to make helpful generalizations. Google Analytics is free, easy to set up for most simple products, and offers the flexibility to change between different attribution models (more on those types later). A good place to start is using Google Tag Manager to set up tracking and share data with your marketing stack. You should also take advantage of server-side tracking options, as they’re available.

Attribution at a late-stage company

For later-stage companies with more complex media mixes, it’s important to understand incrementality on a channel level. You should be thinking about how to bring tracking in house (i.e. from Google Analytics to tracking first party data on your own server), as well as leaning into media mix modeling for budget allocation. We’ll dive deeper into incrementality testing in a bit, but for now,  it’s important to note that you should only jump into that stage when you have the necessary internal resources and capacity to do so without sidetracking your entire attribution strategy.

For later-stage companies with more complex media mixes, it's important to understand incrementality on a channel level.

Late-stage attribution challenges

A common attribution challenge that many mature companies face is how to track offline spend, e.g. direct mail and podcast advertising. This often requires taking a look at mail files or other offline, non-digital signals. For these types of channels that are more difficult to measure, it can be helpful to use tools that help you pull it all together—usually an external customer data platform (CDP) vendor with experience handling complex attribution.

It can sometimes be tricky for marketing teams to get company-wide buy-in on upgrading attribution technology, especially when the rest of your colleagues have only a basic understanding of this space. To help them see the importance of this type of investment, address initial skepticisms to build trust, focus on the break even, and always bring it back to knowing your audience.

Understanding Marketing Attribution Methods and Incrementality Testing

There’s no one magic attribution method that will tell you everything you need to know about your customers. Different methods will be helpful in different situations, and most often, you’ll want a mix to give you the clearest understanding. Click-based attribution is typically broken out into a few main categories:

  • First-click gives all attribution credit to the first touchpoint that a customer interacted with
  • Last-click assigns most credit to the last channel a customer touched before converting
  • Linear is used to proportionately distribute credit when a customer has multiple touchpoints
  • Time decay gives most of the credit to the last touch, but steps down over time to earlier customer touchpoints
  • U-shaped splits the credit between the first and last touch, and sometimes assigns a small fraction of credit to middle touches.
Click-based attribution categories include first-click, last-click, linear, time decay, and U-shaped.

To build upon the findings of your attribution data, you can use this information to begin incrementality testing. Incrementality tests demonstrate the true added value of a tactic instead of just which click generated the last touch. Essentially, it allows you to leverage holdout groups to establish a statistically significant ‘control’ baseline for results for your measurement, which can then inform more advanced marketing strategies. This process can help you assign real value to your channels and better understand optimal target CPAs by channel or campaign.

What does that look like in practice? A remarketing ad campaign might get 10 conversions on last-touch. But if a holdout group that didn’t get remarketing ads still had nine conversions, then the incrementality would be one conversion and not 10, making the CAC 10x higher than what it would be assumed to be had you not done incrementality testing.

Verifying results through media mix modeling

The key to incrementality testing is to start small and begin with single-channel tests. As your marketing programs become more sophisticated, move to holistic incrementality testing across your entire channel mix. This allows you to look at discrete points along a continuum of spend to understand how much each channel is impacting value. The ultimate goal is to understand marginal CPAs in order to optimize your spend. Media mix modeling is also worth considering as a method to verify the findings of other models.

Marketing Attribution Technology, Tools, and Tactics

There are countless forms of tech and tools out there that can help take the manual legwork out of attribution. We already mentioned Google Analytics as a good starting point, but there are more sophisticated tools and even third-party vendors that can help you get the customer data you need. Regardless of what tools you use, you should never rely on one ad platform’s data unless you’re only running ads on that one platform.

It’s particularly important to utilize the right tools to track your offline channels. Because it’s not click-based and you can’t get data on an individual customer level, offline measurement is much more difficult. You’ll most likely need to use a partner, but even then, there’s no perfect solution for measurement, so you should continually validate their data using HDYHAU surveys, vanity URLs and promo codes, brand awareness studies, and incrementality testing. And don’t forget that while surveys and studies are useful, you shouldn’t trust them completely—customers often forget touchpoints or don’t know why exactly they were driven to make a purchase.

Attribution challenges post-iOS 14.5

In addition to offline channels, another instance where it’s smart to use specialized tools is mobile apps that are dealing with the privacy changes in the post-iOS 14.5 world. The iOS update released in April 2021 made it more difficult for brands to track users individually, and therefore, gives us a murkier picture of the true attribution path for your converting customers. This shift toward increased user privacy is likely to continue and marketers will need to be creative in how they gather attribution data in the face of these challenges. We’re likely to see companies shift more efforts into media mix modeling and running incrementality tests.

For companies struggling with attribution due to privacy policies, working with a mobile measurement partner (MMP), like AppsFlyer, can provide valuable anonymized data in a landscape that’s otherwise void of customer information. It’s also a good idea to implement server-side tracking whenever possible. And don’t forget, just like offline attribution, you should always analyze your data against alternative sources (like surveys and media mix models) to verify its accuracy.

Avoid These Common Marketing Attribution Mistakes

As we’ve covered, attribution isn’t always straightforward and it’s only going to continue getting more opaque as privacy policies become more strict in the future. Because of this, there are some common traps we see growth marketers fall into when using attribution models to inform business decisions:

  • Relying on platform data, which almost always overestimates that platform’s impact
  • Assuming that a customer’s first or last touch is what led to conversion
  • Overinvesting in new tech or tools when you don’t have the infrastructure or capacity to implement them
  • Expecting channels to continue performing at proportional efficiency with higher spends
  • Misattributing to bottom funnel tactics instead of top of funnel tactics that actually created the value
There are several common marketing attribution mistakes that can skew your data.

A quick note on ad platform data: This information is helpful for understanding the difference between campaigns within the platform and gaining insights into creative engagement. Basically, if you’re trying to make a decision about how to optimize and scale campaigns within that platform, it’s generally okay to trust platform data and use it directionally. However, keep in mind that platforms tend to take as much credit as possible, especially when you use their default settings. Make sure to hold all tactics to the same standard of attribution. And avoid double counting your data, for example by factoring in that Meta conversions might also be counted as Google or affiliate conversions.

Making Your Attribution Strategy Successful

As with all facets of marketing, your ideal attribution strategy depends on your unique company, growth stage, and business goals. By understanding the different types of attribution models and the importance of incrementality testing, you can determine which mix is most useful to you for understanding what makes customers convert. And although there are some common pitfalls and increasing privacy challenges, there’s a growing ecosystem of attribution tools and partners that can help ensure you’re getting the clearest, most accurate data to inform your marketing efforts.

Want to talk shop about your attribution strategy? We’d love to chat. Drop us a line at growth@rightsideup.co to get the conversation started.

Chelsea started her career as a marketing assistant and fell in love with Google Analytics before focusing on analytics full-time. She's worked across ad agencies, early/late stage startups and tech companies analyzing offline, digital, and social channels. Currently full-time at Credit Karma, she spends her days solving attribution problems and helping marketers better understand the user journey.  When she's not crunching numbers, you can find her walking her dog around Brooklyn or trying to find the best caesar salad in NYC.

Rian began his career in analytics and has gone on to marketing leadership roles at then early stage businesses such as FanDuel, Wag! and GoodRx. Presently he is working with Right Side Up clients as a fractional marketing leader. Rian brings a data-driven approach to his marketing teams based on his analytical foundation and is experienced in navigating attribution in a wide variety of business models and business stages across a range of industries. He is also adept at shaping attribution to varying media mixes, including digital, broadcast and offline media.

Ryan leads the project strategy team at Right Side Up, helping clients develop and resource best-in-class growth marketing strategies. With over 10 years of experience building customer acquisition and engagement programs at companies like Expedia, Prosper Marketplace, and Trip Advisor, he's a big marketing nerd and always eager to geek out over things like incrementality and Excel index match functions.

Bridget is a growth marketer with deep performance marketing and analytics experience. Prior to joining Right Side Up, Bridget ran the international marketing team at Zulily, where she managed user acquisition and retention across paid social, paid search, email, and onsite offers in 70+ countries. Bridget also led incrementality testing and measurement that led to significant improvements in the company’s attribution model.

Let's talk growth

Get in touch

Let's talk growth

Get in touch

Let's talk growth

Get in touch