Throughout Squared Online, our students can learn from, and interact with, expert speakers who are thought leaders in their chosen fields. With attribution being a key part of digital marketing, we asked Gary Nissim, Managing Director of indago digital and an expert speaker on the course, to share his thoughts.
We hope you enjoy reading!
My name’s Gary Nissim, Managing Director of indago digital - a digital marketing agency that specialises in running acquisition campaigns focusing on attribution and search marketing. Today, I’d like to talk a bit about attribution.
Let me set the scene and explain why attribution is so important. Your marketing channels work together like a team in any sport, but currently you’re measuring their performance as individual players.
In soccer, how can a striker score without his defence obtaining possession from the other team? In rugby, how can the winger get the ball over for a try if the scrum half doesn’t get the ball out of the maul? It’s this collaboration and team interaction that attribution helps us understand.
Unfortunately, attribution is often discussed like it’s a private club for enterprise clients and those who can afford to employ specialist and expensive software. However, there is a great free tool that can provide clients who aren’t in that exclusive club enough insight to make decisions based on an attributed view – Google Analytics.
Is it perfect – no! Will it require man hours – yes! Can you make it work – hell yes!
What I’m going to do is show you how to create a few easy reports that will:
- Prove you need to implement attribution
- Provide guidance on the models that suit your needs
- Help start
- your attribution journey
But who am I and what do I know? I’m an acquisition marketeer with over 20 years’ experience and am one of the acquisition trainers for Squared Online. I’m not saying that I’m a leader in the field, but I know enough that you should stop skim reading and pay attention!
Prior to starting your journey, you need to understand that standard reports within Google Analytics work on a ‘Last Click / Interaction Wins’ model. That means that 100% of the value of any goal, event or eCommerce transaction is awarded to the last channel the customer interacted with prior toconverting.
So where do these golden reports sit? All the reports I’ll discuss can be found within the two areas shown on the right.
Multi-Channel Funnels – Assisted Conversion Report
The assisted conversion report perfectly explains why you need to employ attribution.
Focus on the far-right hand column, which is basically a ratio. It demonstrates a channel’s (or medium’s, as seen in GA) ability to drive a conversion as well as assist one. If the number in that column is 1, then it’s assisting the same amount of conversions that it’s directly driving. More than 1, and it’s assisting more than it’s directly driving.
In this example display is the best channel to prove why ‘Last Click / Interaction Wins’ is not a sensible way to measure your marketing activity.
What we can see here is that for every ‘Last Click / Interaction’ conversion display drives, it assists 11.25 conversions. Without this knowledge you’d be underestimating and under investing in this channel. In fact, every channel is assisting more than it’s converting but the ones like display that have the highest number need to be given more kudos and more budget.
Multi-Channel Funnels – Top Conversion Path
Here we are looking at the top 10 conversion paths a customer takes prior to purchase that include paid search.
Of those top 10 paths that equate to 228 conversions, only 24 would be associated to paid search in a normal Google Analytics report. (Remember: last click / interaction wins). What about those other sales? Did paid search not play a role - especially when it was the first channel a customer interacted with?
Short changing paid search will simply drive less transactions for your business.
So now we know that we need attribution, how do we start to build a suitable model for our unique requirements?
Multi-Channel Funnels – Time Lag
This report tells us how long (in days) a customer takes from first interaction with our campaigns until making a purchase.
41.96% of conversions happen on that same day. So, we know that remarketing over a prolonged period makes sense but hitting customers hard on the same day is maybe more important, which is great.
From an attribution point of view, we also know that the first day is extremely important and we can place more value to the channels that a customer initially interacts with as they are driving nearly 50% of all conversions.
Multi-Channel Funnels – Path Length
This report tells us how many times a customer interacts with our digital marketing prior to making a purchase.
A whopping 77.33% interact more than once. If that isn’t a signal that attribution is key, I don’t know what is! Overlay that with the Top Conversion Path report and we can easily start to see what channels are working most efficiently together.
Attribution – Model Comparison Tool
Here we can choose to compare models with a series of standard models that Google Analytics provides.
I suggest that measuring performance and associating budget on any model other than last interaction or last AdWords click will provide you with better results. (Last AdWords click basically favours AdWords).
In the next screenshot I have compared two models:
- Last Click or Interaction
- Linear (Where equal value is provided to all channels)
If you review the column on the far-right, you can see negative and positive numbers. Negative numbers mean that that channel is more likely to appear as a last click / interaction than assist a transaction. Positive means they are more likely to initiate or aid a transaction rather than convert it at the end.
Some would argue that these last click / interaction channels are stealing the glory from the channels that are actually generating the sale, and I would say that there’s a lot of truth to that.
Last click interaction is better than no tracking but it’s not a sound way of measuring your activity. Remember that by looking at these reports all you’re doing is looking at data – you’re not changing anything in Analytics or your marketing plan - so why not play away and produce hypotheses?
To test them, you will need to invest differently. Be sensible and start with small changes, make sure to measure results, become bolder and become an attribution marketeer.