Daasity’s Complete Guide to Multi-Touch Attribution
The question keeping eCommerce marketers everywhere up at night: are we killing it or are we just burning money? Are we spending the right amount in the right channels to maximize profitability?
It’s a tough one. And the more complex your business gets, the harder it is to answer.
What’s that you say? You look at Facebook reporting? That’s very trusting of you. Remember, there’s an inherent bias to vendor-reported numbers so you keep spending with them. If you want to get a realistic view of how your marketing is performing, it will require a broader perspective and more data points.
Without data from multiple sources and a multi-touch attribution model that helps you compare the effectiveness of different marketing channels along the customer journey, you won’t know what an order really costs you and if your marketing is keeping you in the black.
Why multi-touch attribution matters to growing your business
Getting serious about multi-touch attribution is even more critical now, in the aftermath of Apple’s iOS 14 update, which requires users to opt-in to in-app tracking. This is particularly detrimental to merchants who rely on Facebook advertising. Only about 20% of iOS users have opted in, making the Facebook algorithm less effective and therefore less valuable (and more expensive) for advertisers.
Understanding marketing performance is essential for making smart decisions about your marketing budget. If you want to optimize your marketing spending and ensure long-term profitability, attribution data and performance metrics should be the basis of all the conversations you have about performance and budget.
Who is this post for? We find that businesses start having the serious talk about multi-touch attribution at the point where their marketing is becoming more complex; for example, they’ve added channels beyond Google and Facebook and are selling in several channels, not just Shopify or Amazon.
Maybe they’ve been using Google Analytics to track attribution to date, but they want to get more nuanced to try and figure out how effective their awareness- and brand-building advertising is performing.
How can they? What are the options?
There are certainly many options. That’s what makes the conversation about multi-touch attribution so interesting. Attribution models will be different for every business. It depends on multiple factors, including what stage your business is in and what model is going to provide the information you need to grow a healthy business.
This post will help you understand your options, as well as best practices on what you should consider when deciding which multi-touch attribution model is right for your business.
What is multi-touch attribution?
Multi-touch attribution is about determining which marketing channels along the customer journey get credit for a sale. It helps you answer the question, how did this customer get to my website to make a purchase?
Using additional metrics, this information then helps you understand what it cost you to get that order and if that marketing spend was ultimately profitable. Then you can determine which marketing channels were most effective and where to continue or stop investing.
While this may sound simple in theory in the real world, the customer journey can be complex and nonlinear. There is very rarely clear action and result, such as “customer clicks on ad and immediately purchases.”
More likely, the customer has interactions with your brand along several different touchpoints before they purchase. For example, they could see a Google ad and move on, click through to your site from a Facebook ad but not purchase, read reviews on Amazon, and then come back to your site to purchase at a later point by directly typing your website URL into a browser.
Each of these channels get credit for the purchase, but how do you decide how much?
Here's another example of a customer journey:
In the above example, a customer has 8 interactions with a brand before they finally convert. Although they used a podcast promo code, their last click channel was Paid Social.
Challenges of multi-touch attribution
Multi-touch attribution is hard, no doubt about it. There are several factors that make it particularly challenging.
Factor #1: There’s no right answer
When discussing your approach to marketing attribution, it helps to have the right mindset. That is, an open mindset. That’s because there’s no one “right” way to do multi-touch attribution. Every company will set it up differently.
The key to doing it effectively is two-fold:
- Understand how the model you choose works
- Ensure the information you can get from the model supports the sales cycle, your business goals, and your marketing strategies
While you could argue that multi-touch attribution is sort of a crap-shoot—an “imperfect science,” let’s say—it’s still important to use attribution models to give you a base to work from. Otherwise, without any or one-sided data (e.g., vendor-reported only), you’re determining your marketing allocation in the dark.
Also, our experience shows that attribution models don’t have to be super complex, requiring a team of developers to build your own algorithm. You can get a lot of useful information from a relatively simple model if you understand how the different models work, your customer journey and your business goals.
Factor #2: Not every customer interaction is digital
As your marketing mix becomes more complex, you will likely have a combination of online and offline touchpoints. There are strategies and best practices that can help you track offline marketing and analytics platforms that can help you bring it all together.
It’s good to have this knowledge on your radar, but offline attribution is a whole other beast that we’re not going to get into in this post.
Factor #3: Not all interactions have the same worth
Is first touch more important than last touch? What about all that stuff in the middle? Assigning weight or credit to the different touches a customer has in their journey can be confusing.
The key is to think about your business goals and what you want to find out about your marketing efforts. Understanding the different models and what they can help you measure is a great place to start.
Understand your options: Types of multi-touch attribution models
There are a handful of attribution models that you should know. We’ll review first-touch, last-touch, linear, time-decay, U-shaped, and W-shaped. Each one gives more or less credit to different touchpoints in the customer journey.
First-touch attribution model
What it is: A first-touch attribution model assigns 100% of the credit for a sale to the first channel a customer who purchased interacted with.
What it’s good for: Understanding the effectiveness of awareness-building or brand-building campaigns. That first-touch is the most important metric because it helps track reach.
Primary Use Case: Your young business is in full-on acquisition mode to grow your customer base and you want a simple way to track which channels are reaching customers who ultimately purchase.
Last-touch attribution model
What it is: A last-touch attribution model assigns 100% of the credit for a sale to the last channel the customer interacted with before they purchased.
What it’s good for: Understanding which channels are most effective to deliver the messages, offers, and content that convert a customer to purchase.
Primary Use Case: You’ve built awareness about your brand and built up a customer base. Now, you want to figure out which channels and which CTAs on these channels will help you increase sales.
A straightforward way to start analyzing multi-touch attribution that will give you a lot of useful information is to compare first-touch and last-touch attribution. Neither would be completely right or wrong, but give you access to which channels convert more effectively at different stages of your funnel.
Linear attribution model
What it is: Linear attribution considers the multiple touchpoints in the customer journey and breaks down the credit into equal parts.
What it’s good for: Understanding the (not too complex) customer journey and for getting a high-level indication if your marketing strategy is working.
Primary Use Case: You want to add additional channels to support and nurture the customer journey, and you need to track and test which combination of channels drive more customers towards a purchase.
Time-decay attribution model
What it is: Time-decay attribution is a more nuanced version of linear attribution, giving more credit to the touchpoints closer in time to the sale.
What it’s good for: Understanding the most effective “conversion path,” i.e., which channels are most effective at ultimately converting to a purchase.
Primary Use Case: You want to fine-tune your marketing strategy and need to figure out where to shift or invest more budget to increase conversions. This model can help you understand the customer journey at the bottom of the funnel as the customer becomes closer to buying.
U-shaped attribution model
What it is: U-shaped attribution, also called position-based attribution or bathtub attribution, gives more credit to the first and last touches, with less credit to touchpoints in the middle of the customer journey. Generally, 40% of credit is assigned to the first and last channel, and the rest of the credit is divided among the remaining channels.
What it’s good for: Understanding which top-of-funnel and bottom-of-funnel touchpoints are most effective in the context of a longer customer journey with multiple interactions.
Primary Use Case: U-shaped attribution overweights the first and last touches. When you pick a U- shaped model, it's typically because you place more importance on top of funnel and bottom of funnel channels.
W-shaped attribution model
What it is: W-shaped attribution is another version of position-based attribution that gives more weight to first- and last-touch and also credits a mid-journey touchpoint in the funnel.
What it’s good for: Understanding more complex campaigns and longer customer journeys with many different touchpoints. It helps you test what mid-journey touchpoints may be most effective in nurturing a lead to drive customers to purchase.
Primary Use Case: You have a longer sales cycle that requires multiple touchpoints before customers decide to purchase. You want to figure out which mid-funnel touchpoints are most effective in moving customers faster through the funnel.
Setting up the Attribution Models
You can set up all these attribution models (besides the W-Shaped Attribution Model, which requires a custom Google Analytics implementation or attribution tool) on Google Analytics out of the box. Shown in the image below is the basic set up.
- On the left sidebar, click Conversions
- Next, click Multi-Channel Funnels
- Then, Click Model Comparison Tool
- Name your Attribution Model
- Choose a model: you can choose from u-shaped, first-touch, last touch, Time Decay, and others.
- (If applicable: adjust the amount of conversion credit)
- Set custom rules, such as lookback window (e.g., a period of 30 days) and behavior-based credit
How to pick a multi-touch attribution model for your business
The attribution model you decide on for your business depends on… your business. It should make sense for the length of your business’ sales cycle, how you run your business, your marketing strategy, and—let’s be honest—what your finance team will let you get away with.
Marketing attribution is an important decision. Where you decide where credit for a sale goes should be the foundation for all performance and budget conversations.
Some considerations to help you determine which attribution model to use include the following:
- Consider your business goals and marketing strategy: Is your business young, more established, or mature? Are you focused on acquisition or retention? Are you expanding to new markets? All these considerations should influence which attribution model you choose.
- Know your cash flow: Often, it’s budget that influences your marketing decisions. Do you need cash now? Or can you invest in longer-term awareness and brand-building? It’s important to understand how patient your team is from a financial perspective as you develop your marketing plan–that patience will influence which attribution model you choose.
- Understand your sales cycle: To develop an effective marketing strategy, you need to know how long it takes to convince customers to buy your product(s). Do your customers need to do research before they buy? Are they making an expensive or long-term investment? Or are your products relatively inexpensive or impulse buys? The length and complexity of your sales cycle influences your marketing as well as how you will measure its performance.
Best practices for how to build a multi-touch attribution model
- Start simple: If your business is young and small, it’s smart to not make attribution too complicated. That said, you should at least look at first-touch and last-touch–and yes, you can also look at vendor-reported numbers. Taking all these models into consideration will help give you a clearer picture–likely the “real” results for attribution and CPA are somewhere in the middle.
- Broaden your view beyond Google Analytics: If your business is more mature, with a more complex marketing strategy, use an eCommerce analytics platform like Daasity in addition to the attribution model you set up in Google Analytics to get a combined view of first-, last- and vendor-reported metrics and also track specific promo code usage, post-checkout survey results, or map orders to marketing channels.
- Tap your analytics team: If you are a larger company that has at least one dedicated analyst, work with them to develop an attribution model that is truly tailored to your business and can better give you the data you need.
- Agree on a model: Because there is no one “right” answer on attribution, chances are everyone will have a different opinion. The key is to agree on one and have everyone on the team use the same model and numbers.
- Check in using the MER: While you should monitor and test your different marketing channels to try and optimize them on a per-channel basis, a great, additional data point to add is an old-school metric: the “marketing efficiency ratio” or “media efficiency ratio” (MER). This will help you keep tabs on your overall cross-channel profitability. To calculate MER, you divide total revenue by related marketing spend. When you test specific channels, e.g., increase Facebook budget, if it’s effective, your MER will remain the same because overall revenue will increase.
Be strategic and data-driven
It’s time to give credit where credit is due. That probably means less credit to your social media platforms. That might hurt your team’s pride in the short-term, but it’s better to know what your marketing is really costing you so that you can optimize to make your business more profitable and healthier in the long term. Not sure where to start? We can help. Daasity makes it easier to track your data and understand your true marketing costs. Give us a shout or request a demo.