Customer journeys have undergone a vast change over the past 10 years. Go backwards in time to the early-to-mid 2000’s and the customer journey was a fairly simple and linear one. The internet was significantly less developed than it is today and there was little to no capability to check reviews, price compare or delve deeper for information.
Fast forward to today and we’re surrounded by an abundance of data at our fingertips. Suddenly the purchase of something small still requires a quick bit of research here and a price compare over there. Add in the further complication of additional devices and all of a sudden we’re looking at a much more complicated customer journey, let alone the marketing that goes with it (See figure 1).
In the same way that user behaviour has undergone change, so must the methods of reporting and analysis. That’s why giving full consideration towards attribution has become invaluable.
What is marketing attribution?
In the earlier ages of the internet, it made sense to simply view the website interactions that directly led to a user completing a desired action as being the valuable ones.
In the traditional UK household, buying something online required you to sit down in front of the family computer, find a desired item, and then reach for the credit card. A single interaction with a single transaction.
With advancements in technology, along with the evolution of the ‘art’ of digital marketing, the thought of this type of purchase journey happening these days is almost absurd.
It’s common for customers to conduct their initial research on their mobile devices, only to complete a purchase or sign-up in the evenings when they get home, sat down at their computers (See figure 2).
You’ll also find that there are users that add an item to their basket and purposefully leave the website with the hope that an automated email containing a discount code will be soon to follow (See figure 3). There are also those that will stare longingly at an item for days at a time, counting down the hours until their next paycheque, all the while being “followed” around the web by ad banners displaying that very same product (See figure 4).
There are hundreds more examples, but as marketers the question remains the same: out of all of the marketing channels I utilise, just how can I prove which are the most effective at generating revenue?
It’s easy to let this question remain rhetorical. The mindset of viewing the last interaction before a purchase is made as the ‘driver’ or ‘winner’ has been long-established, treated as being simple and perfectly adequate.
Most marketing platforms sustain this mindset by using this as their default method of reporting conversions gained, along with revenue generated. Historically, reporting anything above and beyond was deemed to be reserved for only the biggest enterprises, multi-nationals and umbrella companies.
And yet, even the smallest online business will experience the effects of modern browsing behaviour; sales made as the result of multiple website interactions, devices and channels. This is exactly why the old way of thinking no longer applies. Modern marketing success calls for an evolution in the way we consider the value of our marketing efforts.
That’s where marketing attribution comes in. Econsultancy defines marketing attribution as:
The practice of determining the role that any given channel plays in informing and influencing the customer journey.
This perfectly encapsulates the concept, although the key benefits require some expansion:
Shining a light on customer behaviour, namely their interactions with different marketing channels at the various stages in the purchasing journey.
Crediting channels with their value more accurately, so that business decisions can be based on data that better represents the ‘whole story.’ Without an understanding of marketing attribution, supporting channels such as paid display advertising can be quickly cast aside because their direct volume of sales commonly appears lower than other paid-for channels.
However, by reducing, limiting or completely removing this channel, overall sales are likely to drop. This is down to how effective it is at driving the journey to purchase forward, typically during the mid-stages. Being able to properly represent this value and the channel’s contribution via n attribution can stop ill-informed decisions being made that will be detrimental to bottom-line figures.
Providing visibility to different parts of the business, particularly those that are responsible for different types of marketing activity, allows for not only a better understanding of marketing effectiveness but also how each department can speak to a customer at every point in their journey to purchase.
As a hypothetical example, imagine that without marketing attribution it is observed that your company’s paid search activity seems to be generating a lot more sales and revenue than your unpaid (or organic) social media activity (See figure 5).
This may cause a reaction that calls for the company’s social media team to increase their efforts, or perhaps even worse, have their operating budget reduced in-line with the perceived value that their work is contributing to the marketing mix (See figure 6).
However, when proper marketing attribution is applied to the data, it is revealed that the majority of customers that make a purchase are often found to have been driven to the website via posts on social media early-on, perhaps in their consideration stage, or even in cases where the customer has not been previously aware of the brand.
Later in the journey, perhaps once they have solidified their intention to buy, they are then revealed to be searching for the name of exact product they desire through a search engine, and interact with a PPC ad. The business has now learned three very valuable insights:
- Unpaid social media is likely attracting customers, putting many on the journey to purchase. Therefore, it is more valuable to the business than first realised.
- The team responsible for running the social media accounts now know that they are speaking to users that are typically early on in the journey. They can now tailor their content to more effectively address this.
- The team responsible for handing paid search also know that the messaging a user has been exposed to early-on is likely that found on the company’s social media accounts. It can react to this by ‘carrying-on’ the conversation through paid advertising, more safely assuming that the user has a base-level knowledge of the product and therefore can dedicate more ad space to offers, urgency, and specifics to better drive sales quickly and more effectively.
Now the business has a more accurate understanding of how each team contributes value, and also how the two teams can work together more effectively for better overall improvement.
There’s also a key insight that these three benefits provide: being able to take the ‘mystery’ out of maximising the return on investment (ROI) or return on ad spend (ROAS) from the entire marketing mix. Where ‘credit’ can be spread across the marketing channels involved in a customer’s journey to purchase, so can the revenue generated from that purchase.
In turn, this can be calculated against the total cost for the business to run each channel (ROI) or simply the media spend of each (ROAS) (See figure 7).
This provides a picture of the efficiency of each marketing channel, which can be used as a basis for measuring improvement alongside total volume of sales and revenue generated.
Limitations of attribution
Before looking to incorporate attribution to your marketing activity, there are some limitations that need to be considered first.
The accuracy of the data you are collecting Can your data be trusted? For attribution to be accurate, you need to be able to depend on the data you collect.
Offline activity Do you have a strategy in place to tie any offline marketing activity to online? This can be difficult to achieve, and without any digital measure of how customers have interacted with any offline advertising or physical store locations, the ‘full’ customer journey will be impossible to plot in many cases.
Immeasurable factors Related marketing aspects such as brand awareness, trust and reputation will influence a user journey and their likelihood of becoming a customer. These factors will not be able to be plotted directly when reporting on attribution.
Even with these limitations in mind, this should not dissuade you from moving towards marketing attribution. There is still a great deal of value in moving away from old-fashioned measurement techniques, just so long as there’s an awareness that a ‘perfect’ level of clarity will be extremely difficult achieve.
Before embarking on incorporating attribution into your reporting, it is best to start by gaining a foundation understanding of what is meant by conversions and assisted conversions. Think of this as a primitive type of marketing attribution, and may well prove useful when beginning to educate other staff members and stakeholders in your organisation.
When a conversion (be it a purchase, lead, sign up or other desired action) takes place, the channel that brought them to the website is referred to as the ‘last interaction.’
The default reporting method on various popular platforms (such as Google Analytics) credits the last interaction entirely for driving that conversion (See figure 8).
Notice that the final touchpoint of the journey is solely credited with the conversion, but the previous interactions receive no value assignment whatsoever.
In order to shine a light on the role the previous steps played in helping the customer move down the journey to converting, assisted conversions can be awarded to the corresponding marketing channels so that when overall totals are viewed, their contributions are visible (See figure 9).
This journey would appear as such (See figure 10) when looking at overall totals by channel.
It is important to note that assisted conversions should not be added together with your conversion count, as this would artificially inflate the amount of sales (or leads, sign-ups etc.) that have actually taken place.
The reason for adding and reporting assisted conversions is simply to give visibility to the channels that are helping drive the customer journey, particularly those that do not typically appear as the last interaction. As mentioned previously in the article, without these channels being present, the customer may not have reached the conversion point, and therefore they are likely essential to achieving bottom-line success.
To see how this looks when applied to your own data, you will be able to view this quickly and easily within Google Analytics so long as you have one of the following features set up:
- Goals or Ecommerce
- Enhanced ecommerce tracking
Simply navigate to the Conversions > Multi-Channel Funnels > Assisted Conversions report in the left-side toolbar (See figure 11).
From here, you will be able to see all of the traffic sources that have brought customers to your website, with the total conversions and assisted conversions sorted neatly in table form (See figure 12). You will typically find that the marketing channels that play a more ‘supportive’ role in the customer journey will have a higher ratio of assisted conversions to full conversions.
Moving to attribution modelling
The next step along from measuring assisted conversions is to start using an attribution model. Like the assisted conversion report, doing so will provide visibility of the value that the supporting channels bring.
The difference is that it disregards the assisted conversion count, and instead assigns conversion credit to the channel that makes the most logical sense with regards to your marketing objectives.
In the case of a multi-touch attribution model, a whole conversion can even be ‘split’ into parts, distributing these parts to the channels that feature in the journey (See figure 13).
How the splits are performed and allocated is determined by the attribution model you use, of which there are many, each with different purposes and benefits.
Now you’re familiar with the base concepts of marketing attribution, you can start looking at choosing which digital attribution model will be the right fit for your business or service.
The second article in this series – ‘Which attribution model is right for my business?’ is a great place to start.
Photo by Stephen Phillips - Hostreviews.co.uk