The world of online business has witnessed significant transformations in recent years, particularly in how we understand customer behavior and how customers interact with online businesses.
In the past, attributing conversions to specific sources or touchpoints was relatively straightforward. All it took was using a simple marketing attribution model.
But in 2023 things are more complex…
Why?
Because an average customer interacts with an online business using 2-3 devices and the customer journey is now made of dozens of touchpoints.
As a result, the traditional concept of attribution doesn’t work as well as it did in the past.
In this article, I talk about how shifting from attribution models to contribution models in marketing & sales, can help a business to gain valuable insights into the factors that truly influence conversion.
What is attribution?
Attribution refers to the process of assigning credit or recognition to a particular source or touchpoint for a specific action, such as a sale in the context of online businesses.
Through different marketing attribution strategies, it identifies and acknowledges the factors that played a significant role in driving a successful transaction.
In simpler terms, it’s like saying: “This specific thing or person deserves recognition for their contribution to making this sale happen!“
In the past, people mainly used laptops or desktop computers.
This made gathering attribution data and determining the deserving source or touchpoint relatively straightforward.
There are several marketing attribution models such as linear attribution, time decay attribution, first-touch attribution, last-touch attribution, and many multi-touch attribution models.
However, with the proliferation of devices like smartphones, the task has become more complex and challenging.
The modern customer journey
Today, customer journeys are complex.
For most websites, the customer journey begins on mobile, and it may continue on a tablet or desktop as the user becomes more and more interested in the brand.
To solve the problem of tracking the same user using different devices, some platforms use user IDs.
For example, if you have an eCommerce platform and users create accounts, they can be identified with a unique ID.
If this ID is shared with GA, it allows you to connect all the different journeys the user took on various devices.
This way, all the devices are linked to a single account.
However, the problem with traditional marketing attribution is that it tries to assign the sole reason for conversion to one specific source or touchpoint.
Let’s call it “contribution”
Considering the complexity of this journey, it doesn’t quite make sense to attribute the conversion to a single source, like we used to do.
It’s the entire experience, including multiple touchpoints, that plays a crucial role in driving the conversion.
This is why we emphasize the concept of “contribution to conversion” rather than attribution.
Imagine a journey involving four different devices and around 50 different touchpoints.
In such a scenario, how can we say that the last interaction alone is the sole factor responsible for the conversion?
It would be incorrect to attribute all the revenue generated from that transaction to just one specific source.
Instead, we need to examine how all those 50 touchpoints collectively contributed to the conversion.
At the same time, contribution helps us understand which touchpoints attract more customers.
By analyzing a group of people who made a purchase, let’s say we consider 100 transactions, we can determine how many of them interacted with a particular source throughout their 50 or 60 touchpoints.
This facilitates our marketing measurements and makes it easier to redirect our marketing activities to the real sources of revenue.
With this method, we can explore how many of them came from a specific source and landed on a particular page.
This way, we can gain insights into the contribution of each touchpoint and source toward the ultimate conversion.
Since it makes more sense to talk about contribution rather than attribution, is it time to discard attributions altogether?
Is attribution really dead?
Attribution can actually still be useful for businesses.
It can be used to identify the sources of traffic that can be uniquely identified.
In this case, it’s correct to talk about attribution and it may make sense to apply an attribution model.
This is the case of the “first click”.
The first click is the interaction that brought the user to the site.
For this action, we can unequivocally identify 2 important properties:
- the “source” of the first click (the referral site where the user found the link, or the source type in case of search, direct, email, …);
- the page on which the user landed on his first visit.
This allows us to attribute the acquisition of the user to a specific source and page.
So, attribution still works when we talk about “acquisition” because it can be determined without any doubt.
But attribution has lost its meaning when we talk about conversion because there are simply too many interactions and their contribution to the conversion is diluted.
Companies want to determine conversion with absolute certainty and end up trusting attribution too much.
That’s why, while still using attribution to identify sources and pages responsible for acquisition, businesses should prefer a contribution-focused approach when talking about conversion.
Google handles contribution with a data-driven model, where the importance of each touchpoint is weighted depending on engagement factors.
That model is complex and it may be way more than what’s strictly necessary for a business to understand the contribution of each channel to the conversions.
A much simpler approach to contribution can be found in Funnelytics, where conversions can be linked with sources to explore how many of the people who converted interacted with that particular source at some point in their journey.
In my opinion, we are still in the infancy of contribution models and much can be done to improve them.
In any case, working with a simple contribution model (like the one Funnelytics offers) can give better results than working with an antiquated attribution model.
Conclusion
Customer journeys have become increasingly complex due to users interacting with websites through multiple devices.
These interactions are more extensive and occur over a longer period, resulting in a greater number of touchpoints, and these touchpoints and the related visit often require a double analysis.
As a result, the traditional approach of using a simple attribution model and attributing a conversion to a specific source has become ineffective.
Instead, it is more valuable to focus on the contribution of each source and landing page to the overall user experience that ultimately leads to a conversion, and relegate the use of attribution to when we want to analyze what brought the user to discover the brand.
