The title says it all. Marketing attribution can be the bane of many a marketer, director and business owner wishing to fully understand from where exactly their customers are being acquired. If I were to define channel or marketing attribution in a statement, it would be as so:
Marketing attribution is the process of giving credit to channels that are successfully producing clients or sales.
We could get all pedantic at this point and say that channel attribution is far more complex than the above statement – which it is – but from a business owner or director’s point of view, asking which channels are producing revenue, paying for themselves and ensuring the long-term success of the business will be the most important question on their mind.
I can imagine a typical discussion in many businesses plays out like this:
Business owner: So, Bob. Which channel produced the most sales in the previous financial period?
Bob (marketing director): Erm, well it doesn’t really work like that. You see, people weave in and out of various marketing channels, and attributing sales to one single channel isn’t as straightforward as you may-
Business owner: Yes yes, but you see Bob, I’ve invested $150,000 into these so-called “paid channels” in order to acquire new sources of leads. How have these really fared? That’s money I could have used to pay for two Business Development Managers, you know.
Bob (marketing director): The PPC campaign has performed exceptionally well. As a matter of fact, we can attribute around 65% of our total leads last year from our paid search efforts. It’s a great source of acquiring new sales and, as my previous report shows, it’s a far cheaper route than the traditional outbound sales approach.
Business owner: So you keep saying. But whenever I log into that Google Analytics account you set up, if I set the date range to the previous year then all I see is a very low conversion rate of leads coming from the paid channels. That custom report you set up for me shows an estimated cost-per-lead of $130. Our average lead value is $250, and after taking into account additional costs, such as my sales team, post-sale support and your very significant wage, Bob, then I’m afraid it all just doesn’t stack up.
Bob (marketing director): Well, Google Analytics uses a different attribution model to Adwords and, as I say, it’s not as straightforward as-
Business owner: Heard you loud and clearly the first time, Bob. But I have a feeling it’s money I could save and invest elsewhere. This year, I want to trial us not running any of this Pay-For-Clickings, or whatever you call it. We’ll soon see just how influential all of this new-fangled advertising is. I’m betting my money that we’ll save a lot of advertising costs this year, and nobody will know any different.
Bob (marketing director): Sigh.
How Google Analytics Attribution Works
At its most basic level, the data that you see in Google Analytics shows what happened during a specific visit (or Session, as it is called in Google Analytics). A user and, by extension, a customer, may visit your website multiple times prior to them converting into a lead or sale.
As such, the data that is presented to you in Google Analytics is, by default, incredibly two-dimensional. We do not see what happened before the conversion took place, only what happened when the conversion occurred. Here’s a good example of this in action when examining the Channels report. If you look at the Revenue column, you’ll be forgiven for thinking that the revenue is displaying the amount of money produced by that channel. But this is not the case. Instead, it is simply showing you how much revenue was generated during visits in which the user completed the purchased after arriving on the website through that channel. In short, it was the last channel through which the user interacted when that sale was generated.
Using the above example, we can see that around $1.2million was generated by users when the last channel with which they interacted when purchasing from the website was from a Referral.
Incorrect assumptions are often made by business owners or top-level decision makers who examine channel performance in Google Analytics. Using the above example, one incorrect – and potentially dangerous – assumption is that Referral traffic produces the vast amount of sales, and therefore all other marketing spend – SEO, PPC, etc – should be reallocated to earn more Referral traffic. After all, just look at that enormous Referral conversion rate of 9.49%, and compare it to Organic Search, which produces a meagre conversion rate of just 0.62%.
This is where many (fatal) marketing mistakes are made in business, as top-level decision makers incorrectly assume that the data they are seeing shows which channels are producing new prospects, leads, sales, customers or readers. This is very rarely the case, particularly amongst e-commerce businesses (where users need to consider their product and research other competitors before purchasing), as well as lead generation websites that have a lead time longer than 24 hours.
So, in short: the majority of reports in Google Analytics are simply showing you what happens when the user visits the website through that channel. As such Google Analytics attributes sales or leads to the last channel through which the user interacted when that sale or lead was generated.
The Conversion Path
Let’s present this information more succinctly by displaying the top 10 paths users are taking in order for the conversion to take place. In the below screenshot, we can see the first channel with which the user interacted, plus ant subsequent channels in their conversion journey, ending with the final channel with which they interacted when the sale or lead occured. Notice that in none of the conversion journeys in the below example were sales initiated by the Referral channel, but that the Referral channel is always the closing channel. In fact, looking at the Top 10 conversion journeys, it’s clear that Organic traffic is initiating the most revenue:
We typically refer to the initiating channel as the First-Click channel, and the closing channel as the Last-Click channel.
This is why marketing attribution can be difficult. To which channel would you give the credit to a sale or lead, if the user is navigating through multiple channels?
One school of thought is that you should always credit a lead or sale to the First-Click channel, as this was the channel which initiated the sale in the first place. Whilst this is sound logic, it’s worth bearing in mind that the channels in-between and towards the end of the conversion path can often be the deciding factors in the user’s decision to complete their conversion. I once read online a great anecdote of this fact some years ago. It went along the following lines: in your wedding speech, you would never credit the first romantic love of your life as the sole reason for you marrying your now husband/wife. The initial interaction may of course have led you down the path to finding your soulmate whom you ultimately marry, but in no ways would you base your entire decision to marry your true love on you having met your first love. The decision to only factor in the First Click as the deciding factor of a sale or lead being produced is no different, although in my experience First Click is a far greater indicator of a marketing channel’s influence over new acquisitions.
The reality is that users will discover your brand or website through different channels, and each person’s path will be unique to them.
How Businesses Can Better Report on Channel Attribution
There are many ways to present attribution clearly to business owners and top-level decision makers, without bamboozling them with meaningless stats.
One option is to navigate to the Attribution Modelling report, which can be found under Conversions > Attribution > Model Comparison Tool. Using this report, you can compare First Click performance with Last Click performance, as per the below screenshot which shows that Paid Search is Initiating the most leads, and also Closing the most leads. Notice, however, that Direct initiates only 3.5% of leads, but closes around 15%. This means that Direct is getting more credit than it deserves:
The next report to consider investigating is the Assists report, which can be found in Conversions > Multi Channel Funnels > Assisted Conversions. This is a slightly different report, because it shows the performance of all leads or revenue where that channel was part of the conversion journey, but when that channel did not get any credit. I prefer to refer to the final column on the right, the Assisted/Last Click or Direct Conversions column. Figures that are higher than 1 show more of a tendency to Initiate or Assist a lead/sale, as opposed to Close a sale. In the below example, the Display channel has a high assist influence of 2.00 – perhaps it’s worth trialling some additional spend in the Display channel, to see if this produces more sales across the other channels:
Once the core channels that Initiate, Assist and Close sales or leads have been identified, I prefer to produce reports for my clients which show what happened when users interacted with the key channels. In the below example, a segment has been activated which shows all users who have Interacted with Paid Search advertising. Notice how only 52% of all revenue is actually generated during the Paid Search session, and nearly half of all revenue is instead attributed to other channels, including Organic, Referral and Direct. This gives a clearer breakdown of what else happened when users interacted with each key channel:
This is an easier method of showcasing the influence that specific channels have on overall revenue, and a much easier-to-digest form of reporting for top-level decision makers. Using the above example, we can say that Paid Search has helped to generate $33,463 in revenue in the selected date range, regardless of which channel closed the sale.
Once the idea of attribution can be instilled into the minds of all key business decision makers, more advanced reporting can be delivered to them, including:
- Top sales awareness / initiating / First Click channels.
- Top sales-bolstering / assisting channels.
- Top sales-closing / Last Click channels.
The top-level decision makers will then be presented with all the information they need to make long-term decisions, with bulleted recommendations from the marketing expert, of course.
I hope you have found this dummies guide to Google Analytics channel attribution useful. Whilst I have written it from the perspective of businesses trying to understand channel attribution, the same process applies for blogs, news websites, affiliate publishers and any other website that relies on Google Analytics to understand conversion data.
For further reading, I also recommend reading this related article on why optimising your website for bounce rate can be a dangerous tactic.