Attribution can be described as assigning credit for sales/conversions to touchpoints in a user’s conversion path
(Google)
Understanding what you’re measuring and reporting on is essential to turning data into insights that can be used for informed decision making. Like most marketers in digital advertising I have been very used to reporting on the default setting of ‘last interaction’ attribution, for the simple reason that it is the attribution rule which is easiest to comprehend and the easiest to work out! But there are many more interesting ones to consider... are you familiar with them?
The simple rules - first or last?
The last interaction rule is just that, giving credit to the last touchpoint source which delivered the user to your website from which they then convert.
When using a web analytics platform, it is imperative that all sources are tracked via the same system in order to ensure that the true last action credit is awarded. Proprietary tracking systems like Google Ads, Facebook Pixel, Twitter Ads and LinkedIn Insight Tags are only interested in their own performance, so using these in isolation will never give you a last action view and their stats should never be added together.
In Google Analytics it is worth noting though that the default attribution rule for all Acquisition and Behaviour reports is actually last non-direct click.
Why is this? Google acknowledges that it’s hard for any person to arrive at a site directly without prior knowledge of its existence, if any other source has been involved in the journey, then Google Analytics automatically helps you work towards a truer picture of the most applicable last action touchpoint. In this rule, all direct traffic is ignored, and 100% of the credit for the sale goes to the last channel that the customer clicked through from before converting.
You can see how this has impacted your results by using the Model Comparison Tool in Google Analytics and comparing last interaction with last non-direct click. You’ll notice % change increases in most channels (if not all) except direct, and a large % decrease for direct.
Some paid advertisers opt for a First Interaction rule, as it can help determine whether the channel is influencing conversions which later convert via cheap brand PPC, organic or direct visits at a later date. In the same vein as it’s easy to track all last interactions, in this rule you’re simply turning the rule on its head and crediting all first touchpoints.
When you’re spending thousands on search terms or display advertising which is delivering awareness for the brand and new visitors to the website, it can often be short sighted to report solely on a last action basis. You can also run the risk of over-crediting the bottom of the funnel which makes funding (important) influential upstream channels even harder to justify.
Feeding your Google Ads or programmatic campaigns with first interaction data can help the system find more new visitors and work more effectively (as it has access to more information). Be aware that you may need to adjust your cost per acquisition (CPA) targets if you discover that the follow on channels also cost money to acquire the traffic - a blended cost per acquisition across digital marketing budgets may be required when the user journeys are more complicated.
Getting more realistic - weighting your data
The second most useful attribution rule in my opinion (after last non-direct click) is Position Based. This offers 80% credit to the first and last channel (40% each) in the user’s journey but also acknowledges all the other channels in the middle, which receive an equal share of the remaining 20%.
Using this rule allows you to acknowledge the effort and the importance of the channel introducer and the closer, and takes into account longer path lengths to conversion.
In Google Analytics, if you don’t have Analytics 360, this is the rule which shows you’re serious about attribution, and want to award credit all round, but acknowledge that depending on the position in the journey this credit needs to be weighted.
A Linear attribution rule gives equal credit to each touchpoint in the journey - this is probably one of the weaker rules because it’s unlikely that they each play an equal share. As we know it can often be harder to reach someone for the first time, and also be the touchpoint which does the hard work at closing out the conversion, so why should either or both of these events share the credit equally with other touchpoints?
Time Decay is a version which places greater emphasis on weighting according to how close to the conversion the channel touchpoint was. This is perhaps a step up from linear as some weighting and logic has been applied, however it is flawed in that it neglects the significant role the first channel may play which could need to happen days or weeks in advance of making the decision.
The attribution rule nirvana is to get to a Data-Driven model - this is when your conversion data is used to calculate the “actual” contribution of each touchpoint across the conversion path. Systems like Google Ads, Google Analytics 360 and other paid tracking providers will try and identify patterns within converting and non-converting user journeys in order to fairly apportion conversion credit.
By discovering touchpoint patterns which lead to higher converting probabilities, these behaviours are assigned weighted scores to attribute a single conversion to multiple touchpoints.
Like all other attribution rules (except first and last interactions), the number of conversions typically come out with decimals rather than nice round numbers!
I use the term ‘actual’ very loosely here, because it is also worth noting that each data driven model is specific to an advertiser. And like all attribution rules whilst they can provide more insight than a simple first or last click, none of them will paint an exact picture of what happened. Don’t get lost in trying to find the perfect attribution rule (I’ll let you into a secret, it doesn’t exist!) just be content with finding one that helps you spend wisely and effectively and allows you to scale your business in an affordable way.
How far are you willing to look back?
There is one other aspect of attribution that platforms like Facebook can offer marketers. The ability to look at post-impression (‘seeing’ an ad but not clicking) and post-click attribution in 1 day, 7 day and 28 day look-back windows. Google analytics offers look-back windows up to 90 days before a converting session whilst platforms like Ruler Analytics are indefinite.
Determining what look-back window (how far you’re willing to look back in time to involve a touchpoint in the user journey) and what type of look-back (impression or click) you accept can have serious consequences on your data, cost per conversion values and decision making.
My advice, think logically about how these interaction types can over or under credit the role your activity plays, consider how long it should take people to make a decision to frame what is an acceptable window of deliberation and wherever possible frame the results against your website tracking like Google Analytics and final conversion count. At the end of the day, you can’t claim to have more conversions than you actually do!
The last thing you want to do is overclaim value and waste money, but if you can’t achieve some of your sales without this influencing activity then you’ll be stifling your success.
Test, test and test some more
Whilst you find your feet have a play around with different attribution rules. In Google Analytics you can use the Model Comparison Tool, but you can also visualise the paths to conversion in the Multi-Channel Funnels report.
Find a rule that fits your business and your customers’ typical behaviour, and share this across all your departments so that everyone is on-board with your decision as it’ll affect their stats too!
