This week I completed the Attribution course on CXL.com by Russell McAthy, the CEO and co-founder of Ringside Data. I also managed to score a 90% on the exam on my first attempt and get certified. IF you’ve played around with the Facebook attribution tool (prior to the iOS 14 update of course), then you’re probably familiar with attribution models. However, the course does go a little deeper on what attribution is and is not, where it can be considered and helpful as well as how to strategically use it.
Unfortunately, come August 2021, Facebook will be retiring their current attribution tool partially due to their inability to collect data because of the iOs 14 update. But just because Facebook takes it down, doesn’t mean attribution is not important. Before we jump into the types of attribution models, we first need to understand why attribution is important.
Look at your phone, and take note of the brand/ model. Now think back to the first ad or the first time you heard about the brand or the model you have? Can you think about the number of times you engaged with the brand until you decided to buy your phone? The blogs, the reviews, the youtube unboxing videos, the brand’s website, announcements, maybe there were offline events, newspaper ads, billboard ads, tweets by the CEO, etc that nudged you towards investing in the phone you currently have.
These days, marketers have more channels through which customers engage with their brand than ever before. And while we may map out a linear buyer’s journey, in reality, it’s not that simple. Consumers interact with a brand on multiple touchpoints before they make a purchase, and most businesses fail to credit these multiple touchpoints for a conversion. Why? Because it’s a big challenge.
As a marketer, we need to be experts in understanding which channels are converting customers most successfully and why. This information ideally should help us properly allocate our limited resources into these channels and accurately predict how it would impact the business.
When it comes to digital marketing channels such as SEO, PPC, Affiliate campaigns, email marketing, etc each channel does provide you with reports of their own, however, unless you know how these touchpoints work together to push the consumer further down the conversion path, these reports are just vanity metrics.
A lot of businesses, spend hours crunching numbers from various reports into spreadsheets, in order to get some visibility out of all this information. However, before you sit down to create this elaborate tracking system to make the most out of your data, you should first be aware of the various attribution models available.
Different attribution models distribute different amounts of credit for conversions across all your channels, and this has to align with the type of business you’re in and what makes sense to you.
There are 6 types of general attribution models:
First Touch Attribution:
Here complete credit is given to the first time the user landed on your website. It’s to be considered if your conversion success is defined within a longer attribution window, or if your product/service requires a longer consideration period. This method does not consider the middle or last touchpoints, which might have helped “push” the user further down the buyer’s journey.
Last Touch Attribution:
This model gives 100% credit to only the last click/visit that happened in the conversion path. It can be considered for businesses with a short attribution window or when there are low-consideration conversions involved. This method does not take into account the first and middle touchpoints that made your customer aware of your brand/product/service. It also ignores those that helped push them through the “consideration phase” of the buyer’s journey.
Even Credit/ Linear Attribution:
As the name suggests this model gives an equal percentage of credit to each touchpoint in the conversion path. This model can help you understand how to value the first touchpoint that introduced the product, the middle touchpoints that build consideration, and the last touchpoint that helped people get to the point of conversion. However, even credit models are more illustrative than actionable, as it's unlikely that all touchpoints have the same effectiveness, thereby, not giving you information on how to redistribute your resources better.
Position Based/ Positional Attribution:
Over here, a specific percentage of the credit for a conversion is given to the first and last touchpoints in a conversion path, with the remaining credit distributed evenly across all other touchpoints. Here weighted importance is given to the first and last touchpoint – how the consumer found you and what finally pushed them to convert. This is done, without overlooking or ignoring the middle touchpoints.
Time Decay Attribution:
This model gives a decreasing percentage of the credit for a conversion to touchpoints as they get further away in time to the conversion. It can help you understand all touchpoints in the conversion path while also giving more importance to the most recent touchpoints.
The last one is specifically for those who are familiar with the Facebook Attribution tool - Data-driven attribution model.
However, on April 1, 2021, Facebook discontinued this model. But it’s still important to know, should you consider implementing a custom attribution model.
The data-driven attribution model was made to assign fractional credit for a conversion to Facebook touchpoints based on their estimated incremental effect. This was a statistical model developed b Facebook and periodically updated. Unfortunately, it was limited to only campaign data related to Facebook, Instagram, and the Audience Network.
The only constant is change and policies such as GDRP and market conditions such as the iOs 14 update, which has made complete attribution quite the task. As more attention is being drawn to user privacy, the digital marketing industry will be forced to change tides by changing the way they manage and track campaign performance, and I’m excited to see what’s next. Russell McAthy mentions in the course that, soon most if not all businesses will depend on machine learning to understand and improve their touchpoint attribution, and I can’t wait!
About Russell McAthy
From what I could get from his website, he's a big fan of dinosaurs, cooking, and the NFL - National Football League ( The American Football). As mentioned before, he's also the CEO and co-founder of www.ringside.ai
His course was insightful and gave me a lot to think about in terms of how data can be stitched together and made more useful than it once was. The course, however, doesn't teach you how to set up these attribution tracking channels. But overall, if you're signing up to CXL.com, consider watching his course on attribution.
P.S I was only joking when I said you’d have to manually track everything on a spreadsheet. If you’re looking for a few attribution tools, this link should help: https://segment.com/blog/choosing-the-right-attribution-tool/