What is Marketing Attribution?
The Short Definition: Put simply, marketing attribution is the analytical science of determining which marketing tactics are contributing to sales or conversions.
The Long Definition: Marketing attribution is the practice of evaluating the marketing touchpoints a consumer encounters on their path to purchase. The goal of attribution is to determine which channels and messages had the greatest impact on the decision to convert, or take the desired next step. There are several popular attribution models used by marketers today, such as multi-touch attribution, lift studies, time decay, and more. The insights provided by these models into how, where, and when a consumer interacts with brand messages allows marketing teams to alter and customize campaigns to meet the specific desires of individual consumers, thus improving marketing ROI.
Advanced marketing attribution programs require marketing teams to aggregate and normalize consumer data from across channels to ensure each interaction is properly weighted. For example, if a consumer is exposed to a display ad and an email campaign, but only converts after seeing a special promotion in the email, marketers can note that this piece of collateral played a bigger role in driving the sale than the display ad. They can then devote more resources to creating targeted email campaigns.
To achieve the level of data granularity required for effective attribution, marketing teams need advanced analytics platforms that can accurately and efficiently distill big data into person-level insights that can be used for in-campaign optimizations.
What is a Marketing Attribution Model?
Marketing attribution models assign value to marketing campaigns through statistical analysis at the user-level. This is in contrast to models such as marketing mix modeling that use aggregate data. This person-centric approach is why attribution models are more typically applied to digital campaigns than those conducted offline, such as print advertising. Each attribution model relies on different analytical techniques, which will be explored further later on.
Attribution models are typically categorized as single-touch or multi-touch.
Single-Touch Attribution Models
Single-touch models are less widely used today, as they fail to provide a nuanced look at the customer journey. These models attribute a conversion to a single touchpoint, often the first or last one engaged with by the consumer. An example of single-touch attribution is the last click model, which attributes a conversion to the last piece of marketing material a consumer clicked on before converting. However, this neglects to look at the wider customer journey and touchpoints engaged with.
Multi-touch attribution models look at all of the touchpoints engaged with by the consumer leading up to a purchase. As a result, these are considered more accurate models. Depending on which multi-touch model you use, they might assign value to channels differently. For example, some assign value based on when a consumer interacted with a touchpoint relative to the conversion, while others weigh all touchpoints equally.
The most effective attribution models will provide insight into:
- Which messages a consumer was exposed to and on what channel
- Which touchpoint had the greatest impact on their decision to purchase
- The role brand perception played in the decision to convert
- The role of message sequencing
- Which messaging gets the best results from each consumer
- The impact of external factors (e.g. how gas prices affect car sales)
Benefits of Marketing Attribution
Advanced attribution models can be time and resource intensive to get right, especially complex models that evaluate a variety of datasets for online and offline campaigns. However, when done effectively, attribution brings a myriad of benefits including:
Optimized Marketing Spend: Attribution models give marketers insights into how marketing dollars are best spent by showing touchpoints that earn the most engagements. This allows marketing teams to adjust budget accordingly.
Increased ROI: Effective attribution enables marketers to reach the right consumer, at the right time, with the right message – leading to increased conversions and higher marketing ROI.
Improved Personalization: Marketers can use attribution data to understand the messaging and channels preferred by individual customers for more effective targeting throughout the customer journey.
Improve Product Development: Person-level attribution allows marketers to better understand the needs of their consumers. These insights can then be referenced when making updates to the product to target the functionality consumers want.
Optimized Creative: Attribution models that can evaluate the creative elements of a campaign allow marketers to hone messaging and visual elements in addition to better understanding how and when to communicate with users.
Common Marketing Attribution Mistakes
While marketing attribution can offer many benefits, there are a host of common mistakes that can result in misattribution, obscuring the success of campaigns for marketers.
To ensure they are getting the most accurate data that reflects their users’ customer journey, marketers should avoid:
Attribution models can be subject to correlation-based biases when analyzing the customer journey, causing it to look like one event cause another, when it may not have.
In-Market Bias: This refers to consumers who may have been in the market to buy the product and would have purchased it whether they had seen the ad or not. However, the ad gets the attribution for converting this user.
Cheap Inventory Bias: This gives an inaccurate view of how media is performing, making lower cost media appear to perform better due to the natural conversion rate for the targeted consumers, when the ads may not have played a role.
Each of the biases threatens to have marketers make optimizations in favor the less effective messaging, causing immense damage to ROI.
Digital Signal Bias
This occurs when attribution models do not factor in the relationship of online activity and offline sales. For marketers who make sales both online and offline, they must make optimization decisions based on both online and offline data, not only what they can trace digitally.
Brand & Behavior
Attribution models can often overlook the relationship between brand perception and consumer behavior, or will only look at them at a trend regression level.
Marketers must ensure their attribution models are able to detect relationships between brand building initiatives and conversions. Not understanding how their attribution model measures branding impact is a common and detrimental mistake, leading marketers to make decisions based on incomplete recommendations that devalue brand building.
Missing Message Signal
Creative and messaging are just as important to consumers as the medium on which they see your ad. One common attribution mistake is evaluating creative in aggregate and determining that one message is ineffective, when in reality it would be effective for a smaller, more targeted audience. This emphasizes the importance of person-level analytics.
Different Types of Marketing Attribution Models
As noted earlier, there are two main categories of attribution: single touch, and multi touch. Within these categories there are several core models which each provide different insights:
First-Touch Attribution: First-touch attribution assumes that the consumer chose to convert after the first advertisement they encountered. Therefore, it gives full attribution to this first touchpoint, regardless of additional messaging seen subsequently.
Last-Touch Attribution: Conversely, last-touch attribution gives full attribution credit to the last touchpoint the consumer interacted with before making the purchase, without accounting for prior engagements.
Each of these methods fails to factor in the broader customer journey, as such marketers should avoid relying solely on these methods.
These models are largely differentiated by how they divide credit between touchpoints on the path to purchase.
Linear: Linear attribution records each touchpoint engaged with by the consumer leading to purchase. It weighs each of these interactions equally, giving each message the same amount of credit toward driving the conversion.
U-Shaped: Unlike linear attribution, the U-Shaped attribution model scores engagements separately, noting that some are more impactful than others on the path to purchase. Specifically, both the first touch and lead conversion touch are each credited with 40 percent of responsibility for the lead. The other 20 percent is divided amongst the touchpoints engaged with between the first and lead conversion touch.
Time Decay: The time decay model also weighs each touchpoint differently on the path to purchase. This model gives the touchpoints engaged with closer to the conversion more weight than those engaged with early on, assuming those had a greater impact on the sale.
W-Shaped: This model uses the same idea as the U-Shaped model, however it includes one more core touchpoint – the opportunity stage. Thus, for the W-Shaped model the touchpoints credited with first touch, lead conversion, and opportunity creation each receive 30 percent of the credit. The remaining 10 percent is divided amongst the additional engagements.
Choosing the Right Attribution Model for Your Organization
Marketers need to make several considerations when selecting which attribution model to rely on at their organization.
First, think about the type of sales cycle you use, and how long it typically runs, and how much of it is done online or offline. Ecommerce sites may not have to factor in offline conversions, but most major retailers will
It is also important to consider how much of your marketing efforts are focused on offline methods such as print, broadcast, and television. Organizations that place a lot of value in these mediums will need an attribution model and platform that is able to correlate and normalize online and offline efforts together for the most accurate insights. For example, multi-touch attribution is often credited with working better for digital mediums, while marketing mix modeling provides stronger insights into offline campaigns. Unifying both of these measurements enhances overall visibility.
Ultimately, your organization will likely have to utilize several attribution models in tandem for the most complete understanding of the impact of your efforts.
Marketing Attribution Software and Tools
To get the most reliable insights, marketers will need to use a combination of models and correlate the data from each to determine the correct optimizations to make for online and offline campaigns.
Doing this will require a powerful analytics platform, though many marketers have been disillusioned with these platforms before. The marketers that find the right platform that can provide in-campaign insights into online and offline marketing optimizations will be at a distinct advantage.
There are several categories that marketers should evaluate when selecting a marketing attribution tool or software.
- Connection of branding and performance
- Cross-channel insights
Here are a few questions to ask when selecting an attribution model:
- Can you get visibility into branding impact?
- Do you get visibility into the impact of creative during the consumer journey?
- Can you get person-level insights for non-digital, offline efforts?
- Are you only measuring lift, and not inevitable events?
- Do you use experimental design to avoid correlation bias?
- Can you get insights to optimize during the campaign, or only at the end?
- Do you get insight into external factors that impact campaigns?
- Does the solution provide quality analysis in addition to accurate data?
Additional Resources for CMOs and Marketing Professionals
- Marketing Attribution Models: How Did We Get Here? A History of Measurement
- Inaccurate Attribution is Bad Attribution
- Attribution Buyers Guide: The Light at the End of the Tunnel
- Attribution Buyers Guide: Measuring at the Person Level – Part One
- Attribution Buyers Guide: Measuring the Power of Your Brand