Content

Back to Basics: Attribution Modelling for Content

Historically, it has been difficult to measure the true ROI of digital content. As part of my role, I'm responsible for the creation of digital content that aligns and supports SEO performance goals for clients. However, with every discipline, there are challenges which the dedicated teams at iProspect are looking to overcome through data, creativity and innovative opportunities. 

On average the modern global consumer owns 3.64 devices, according to research from Global Web Index, making touchpoint opportunities exponentially wider. Mobile is fast becoming the dominant device for consumers when shopping online: Adobe’s Christmas 2017 report shows that smartphones had 46% of traffic over the Christmas period.

With this in mind, the challenge for brands of capturing these opportunities with content across devices, and being able to measure it accurately, is something that is still a hot topic for content marketers globally.

This post will focus on one method that is helping to address this: attribution modelling. In the simplest terms, attribution is the process of assigning credit to a marketing channel for each completed sale or lead generated.

Attribution modelling enables brands to track the performance of their onsite content by measuring both the journey that users take to find it and ultimately, its ROI.

Hannah Killick, a Data Intelligence Manager at iProspect, understands this challenge. “It can be daunting as there are many model variations and platforms available to choose from, such as GA360, Abakus, Adobe and VisualIQ. It’s also important to prove that your touchpoints are reaching people, so you can optimise better and evolve your content.”

It's clear to see that there isn’t a one size fits all solution, due to unique business factors such as sales cycle length and number of digital channels a brand has. This means that a degree of customisation is needed, but there are a number of widely used and accepted models out there.

The two main categories of attribution models are Single Touch and Multi Touch. Single Touch are models that assign full credit to one interaction, whereas Multi Touch rewards more than one interaction. There are variants within each, which are explained in a little more detail below.

Single Touch: First Click & Last Click

First Click

This model provides full credit to the first touchpoint that brought the user to the asset; this can work for channels where there is a very short sales cycle. An example would be if a lead is made from filling out a form to download an e-book.

Last Click
This model provides full credit to the last interaction a user has before a conversion is made. An example of this would be an organic search which then resulted in a purchase on the site.

Multi Touch: Linear, Time Decay, Weighted and Position Based

Linear

This model assumes that each stage of the consumer journey is equally responsible for the conversion. In other words, if there are four interactions before a purchase, 25% is given to each interaction. This could be an organic search, a click on a display advert, watching a YouTube video and signing up for an email newsletter.

Time Decay

In this model, more credit is assigned to touchpoints that were closer in terms of time to the decision to purchase. An example of this could be watching a ‘how-to’ video five minutes before the conversion, as opposed to an organic search completed 24 hours ago.

Weighted

This model is more bespoke. Different touchpoints can be given a custom weighting depending on which interaction. For example, reading the email newsletter can be weighted more prominently than an organic search.

Positional

The principles of this are a little more sophisticated. This model states that the first and last touch points are worth a defined amount of credit, and all other touchpoints between this are equally rewarded for their contribution with the remaining percentage.

Algorithmic

This is the most recent and sophisticated model, as it uses machine learning to create a bespoke model, based on all the touchpoints. It creates custom weighting for each stage to represent how customers interact with your channels.

Key Considerations for Content

Here are our top tips to consider with attribution for content marketing:

  • Consider how long your sales cycle is. This will determine what content you need and whether single or multi touch will work for you to track it.

  • Question the purpose you want content to serve in the consumer journey. This should help with when you define weighting.

  • Define where you want people to find your content. You can optimise channels to ensure important content pieces are seen on multiple channels and increase likelihood of conversion.

  • Ensure your content is optimised for devices. This will increase the likelihood of your content being seen and engaged with as a positive experience.

Measuring the success of content is still something the industry is trying to figure out due to the complexities. This won’t be a quick solution. However, having these considerations should give you a steer on what to keep in mind when measuring the success of your content. Think about the outcome and ensure the right tracking is in place.

The encouraging thing to note is that technology is becoming more sophisticated. Machine learning will play a key part in progressing attribution, which will enable more accurate and agile changes in consumer behaviour on your site. This at least provides an indicative measurement of success. 

Please get in touch with our teams to find out how your business can better measure your content. 

comments powered by Disqus