It’s not what you know, but who you know. That quote is supposed to be about networking. But it’s even more important to know who’s converting at your ecommerce shop.
Enter multi-channel attribution. Multi-channel attribution refers to the data and analytics you use to measure your marketing success. More specifically, it addresses who your customers are and where they came from using metrics that are relevant to your programs.
Did they buy your product because they saw it on Facebook ads? What if they dropped out of a cart, received an abandoned cart email, and then finally made the purchase? To what do you attribute that purchase?
Answering these questions is complicated. But it’s also rewarding. With multi-channel attribution strategies in place, you’ll know where your future ad dollars need to go to generate the highest possible ROI for your shop.
The advantages of using multi-channel attribution
Understanding where your conversions are coming from is critical for evaluating the success of your marketing efforts. Here are some of the benefits of using multi-channel attribution in your analytics.
Track top-performing digital channels
Multi-attribution isn’t just about identifying the different channels where customers come from. It’s also about evaluating where a business should spend its marketing budget. Usually, a buyer will interact with several touchpoints before converting.
Without proper attribution, marketers are essentially blind regarding where buyers come from.
It’s critical to have a strategy to credit these different digital channels properly. It’s clear only multi-channel attribution can accurately track which channels fuel engagement with so many touchpoints.
Understand the customer journey
You can better understand the customer journey with access to accurate attribution data. Using a multi-channel attribution model, you can answer one of the most important questions––which channel did the most work to convert a visitor into a customer?
Let’s say you’re an online shoe shop. Using a multi-channel model, you may find that 70 percent of conversions come from your mobile site. Most shoppers see an Instagram ad and click through to your site.
With this knowledge, your team can optimize your Instagram page and mobile site, potentially allocating more budget to these channels.
You can also identify which channels aren’t driving conversions. Then either double down on these channels or focus on other high-performing channels.
The challenges of using multi-channel attribution
While it’s known for its accuracy, multi-channel attribution isn’t always straightforward. Some of the most common challenges associated with attributing conversions across multiple channels are data regulations and misattributed conversions.
Data regulations
With the death of third-party cookies on the horizon and increasing global data-privacy restrictions, it’s much harder to see what customers get up to online before converting. Businesses usually need first-party data to determine what customers are doing before converting.
Companies need a lot of data from all their marketing campaigns to attribute sales to specific channels.
Misattributing conversions
Setting up multi-channel attribution effectively is a complex process. It’s easy to misattribute conversions, leading you to believe one channel was more effective in fueling conversions than the others.
Let’s say you allocate more budget to paid Instagram ads. You then head into Google Analytics and see that revenue has increased, attributing to organic search.
You may then assume that your organic search campaigns are much more effective for driving conversions than more costly paid Instagram ads.
But later on, you dig deeper and find that the growth in organic search was from an increase in search for your brand name.
What appeared to be huge growth from search engine optimization (SEO) was actually misattributed revenue from searches increasing following Instagram ad campaigns.
Weighing up offline attributions
Not all attributions happen online. What happens if a customer’s colleague recommends your product to them? You could use customer questionnaires to fill in the gaps.
Following a purchase or sign-up, you can use the survey to target those channels you haven’t been able to reach with multi-channel attribution––particularly those offline conversions that happen at first interaction.
It may take some experimenting with multi-channel attribution models to get an accurate picture of where your conversions come from.
Types of multi-channel attribution models
Single-touch attribution models
This is the simplest form of all marketing attribution models because it credits one channel for each purchase. But you can still break it up into different strategies depending on how you want to weigh your attribution.
First-click attribution. This is the simplest form of attribution. Whatever generated the customer’s first click gets credited with the conversion. This is true no matter how many channels the customer visits between the first click and their order.
Last-click attribution. The same as above, but reversed. What was the final channel that pushed the customer to the order page?
Last non-direct click attribution. Let’s say a customer clicked around your website a little bit before making a purchase. What was the last touchpoint before their purchase? Under this last touch attribution model, that’s the click that will get the credit.
Multi-touch attribution models
You might have guessed by now that multi-touch attribution models tend to be more accurate. After all, if a customer finds you on Facebook, reads your blog, and decides they love your company, shouldn’t both the blog and the Facebook ad get credit?
Here are some of the multi-touch attribution models that you might incorporate:
Linear attribution. Someone clicks on an Instagram ad, which takes them to a newsletter signup page, which sends them an email, which results in a purchase.
What channel made the sale? Under linear attribution, you would divide the conversion across all channels involved equally, or 33% of a conversion for each.
Last-channel attribution. Similar to last-click attribution, this attribution model disproportionately awards credit to the last channel a customer visited before a purchase. The difference? You would still share attribution with the initial reasons a customer clicked.
Time-delay or time-decay attribution. We’ve previously written about delayed attribution during seasonal promotional events.
The time decay model of attribution can alter how much a channel is credited for a conversion depending on its time distance from when the sale goes through.
Position-based attribution model
Finally, position-based attribution models are a hybrid of single-touch and multi-touch styles. Also called first-and-last channel attribution, this model puts added weight to the customer’s first and last clicks.
But rather than giving these clicks the complete weighting of the conversion—splitting them 50-50 with no attribution to any channels in between—there’s still some left for any other channels that user may have visited.
If a customer visited four channels, this attribution model might award 40% each to the first and last channels that brought in the conversion. You’d then distribute something like 20% to the channels in the middle.
Getting started with multi-channel attribution
Setting up your custom attribution model typically involves three key steps: choosing your attribution style, configuring your analytics, and testing and analyzing your data.
1. Choose your attribution style
The first step sounds easier than it is. You’ll choose the style of attribution that most accurately paints a picture of where your conversions are coming from.
For most, a multi-channel attribution strategy is a slam dunk. But which multi-channel strategy is right for you? Should you emphasize last clicks? First clicks? Split attribution between both?
There are a few ways you can diagnose which will work for you:
Look back at existing analytics. Collect the data you already have and analyze it. What is the picture your analytics are painting? Try running different calculations between first and last-click attributions to see which set of data better resembles your experience.
Compare multiple data sources. If customers are telling you they converted because of an abandoned cart email, but you’re using first-click attribution, you need to switch strategies to resolve the discrepancy.
If you aren’t sure what to do yet, expand the types of data you collect to get a clearer picture. Then revisit the data in 90 days to see if there are insights that can help you choose your attribution strategy.
2. Configure your analytics
“Google’s shift to GA4 from UA will pose another challenge for multi-channel attribution,” said Joe Karasin, Head of Growth at CircleIt Inc. “Most marketers will need to get a firm understanding of how to use GA4 in conjunction with Tag Manager, Meta, and Google Ads to keep attribution models relevant.”
Google Analytics 4, for example, might be set to single-touch attribution by default. And before you choose the attribution model you want to use in Google Analytics, you’ll want to be clear about how it attributes each conversion.
Cross-channel first click. A single-touch attribution channel that gives 100% of conversion credit to the first channel your customer clicked.
Cross-channel linear. Multi-touch attribution that divides the attribution equally along all channels the customer engaged with before converting.
Cross-channel position-based. 40% to the first interactions and last interactions, and then 20% divided among “middle” interactions during the customer journey.
Cross-channel time decay. Prioritizes touch points that happened closer to the actual conversion. “A click 8 days before a conversion gets half as much credit as a click 1 day before a conversion,” writes Google Analytics.
3. Test your data and analyze
Once you’ve chosen an attribution model and configured it in your analytics, your next step is to wait.
Why wait? You need to gather enough of a sample size to determine whether your multi-channel attribution is accurately reflecting the conversion data.
“Test,” says Nate Nead of SEO.co. “Test, test, test. … Nothing speaks better than your own data. Your own page will likely convert very differently than a competitor’s—even if someone landed on both pages from the same search.”
In other words, there is no universal answer here, no rule of thumb that says you should always attribute 33.178392% of your conversions to Facebook ads on a Wednesday. However, there are a few steps you can take during the tweaking phase:
Roll out post-purchase surveys. How do you know what to test in the first place? If you roll out post-purchase surveys, you’ll have data you can compare to your multi-channel analytics.
Here’s where you can find the gaps, such as if customers say your landing page made the sale, but you’re awarding 50% of that conversion to paid social media.
Watch the timing. Even if you’re not sure if you’ve chosen the best attribution model yet, you can get an idea of how accurate it is by watching the timing of your conversions.
For example, if conversions spike after you launch a Twitter ads campaign but every other aspect of your funnel is the same, you may think about weighting conversions towards the first click.
Evaluating the success of a multi-channel marketing model
Customers expect to interact with your brand across multiple channels. Plus, when surveyed in 2022, 75 percent of marketers said they use a multi-channel approach to campaigns.
But with so many sources of campaign and customer data, it can be challenging for marketers to evaluate the success of multi-channel marketing efforts.
As you add new marketing channels to your efforts, you’ll need to evaluate each campaign’s success—which depend on your campaigns’ goals. To get started, here are four questions to ask while evaluating the overall success of your multi-channel efforts.
1. Is short and long-term revenue increasing due to your campaign?
The clearest indicator that your campaign is successful is an increase in revenue for the product or service you’re selling. When you have a solid multi-channel attribution model, you can determine which marketing channel or campaign this revenue is coming from.
2. Is the cost of customer acquisition going down?
As your campaign progresses, acquiring each new customer should cost your business less. If your customer acquisition costs go up, you may need to rethink how you’re allocating your campaign budget.
3. Are you reaching conversion targets?
Conversions will look different depending on your business model and campaign goals. For instance, conversions could be signing up for a free trial, scheduling a demo, or purchasing a product.
Before you launch any marketing campaign, clearly define your conversion targets. Then evaluate whether you’re reaching them on a monthly and quarterly basis.
4. Do you know when and where customers convert?
Marketing campaigns can’t be successful if you can’t attribute their success to anything tangible. If you can identify which campaign elements prompted customers to convert, you have a solid attribution model in place.
