Attribution modeling is incredibly important to the operations and strategy of your ecommerce or multichannel business - even though you might not think of it in those terms, you’re doing it to some degree. Attribution modeling is the process by which you assign credit for purchases to different touchpoints in the customer journey. In a multichannel world, it helps answer the question - which channel is ultimately responsible for driving a sale? There are several different methods of multichannel attribution - in fact, businesses will often use more than one to analyze different things, depending on their strategy. Here, we’ll go over what multichannel attribution is, why it’s important, what options you have, and how Glew solves attribution for merchants.
What is multichannel attribution?
Multichannel attribution is the set of rules that determines which of your marketing channels receive credit for the sales and revenue you bring in. In a multichannel world - where customers may interact with you on multiple platforms and see multiple campaigns before making a purchase - knowing what ultimately drove them to make a sale is key. Here’s a real-world example: Say you have a customer who visits your website through a Google Ads campaign, but then leaves and sees an ad on Facebook before coming back and making a purchase. How do you attribute that sale - to Google or Facebook? And what about when that customer comes back to make another purchase after receiving an email campaign? Where do you credit that - and future - purchases? Multichannel attribution helps you navigate that complicated web of information - for every customer and purchase - so you can better understand and evaluate your marketing channels. When it comes to multichannel attribution, you have a few options. We’ll go over the main ones below - including the ones Glew uses.
Attribution models available in Glew:
First-order attribution - the default attribution model in Glew - assigns credit for a sale, and all future sales, to the channel that drove the customer’s first purchase. (This applies to web orders only for most ecommerce and analytics platforms. However, for Shopify users, Glew is able to default to Shopify’s order source for non-web orders like marketplaces, retail point of sale, ReCharge or draft orders.) EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook before making a purchase. Under first-order attribution, Facebook would get credit for that sale. If they come back a month later and make another purchase after clicking on an email campaign, that second purchase would still be attributed to Facebook, since Facebook drove their first purchase.
Last-click attribution - also available in Glew - assigns credit for a sale to the channel of the customer’s last click prior to purchasing. (This applies to web orders only for most ecommerce and analytics platforms. However, for Shopify users, Glew is able to default to Shopify’s order source for non-web orders like marketplaces, retail point of sale, ReCharge or draft orders.) EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook before making a purchase. Under last-click attribution, Facebook would get credit for that sale. If they come back a month later and make another purchase after clicking on an email campaign, that second purchase would be attributed to email, since that was their last click prior to purchasing.
Glew allows you to toggle between first-order and last-click attribution on certain tables throughout your order and channel reporting. That means you can choose the attribution model that works best for your business - just keep in mind that you’ll see performance metrics for your channels change when you switch attribution models. Here's where you can change your attribution model in Glew:
- Performance > Overview: Revenue by Channel, Net Profit by Channel and Order Source by Channel tables (note: Order Source by Channel availability may vary depending on your ecommerce platform)
- Customers > Lifetime Value: Lifetime Profitability by Channel table
- Channel Details: All metric groups
- Orders List: All metrics
Why does Glew use these two attribution models? We think they do the best job at helping most merchants understand their channel performance in both the long-term and the short-term. We’ll get into why (and when) you may want to use each model in just a bit.
Other attribution models:
First-click attribution assigns credit for a sale to the channel that receives the customer’s first click in their journey to purchasing. EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook and interacts with an organic search result before making a purchase. Under first-click attribution, Google Ads would get credit for that sale, since it was the channel of their first click.
Linear attribution assigns credit for a sale equally to each channel on a customer’s journey to purchasing. EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook and interacts with an organic search result before making a purchase. Under linear attribution, Google Ads, Facebook, and organic search would all get equal credit for that sale, since the customer interacted with each one before purchasing.
Position-based attribution, similar to linear attribution, assigns credit for a sale to each channel on a customer’s journey to purchasing - however, it assigns credit proportionally depending on where those channels fall in a customer’s journey. Typically, position-based attribution assigns 40% of the credit to the first and last touchpoints in a customer’s journey before purchasing, with the remaining 20% assigned evenly to the middle touchpoints. (However, you can adjust the weighting as it makes sense for your business.) EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook and interacts with an organic search result before making a purchase. Under position-based attribution, Google Ads and organic search would each receive 40% of the credit for the sale, while Facebook would get 20%.
Similar to linear and position-based attribution, time-decay attribution assigns credit for a sale to each channel on a customer’s journey to purchasing - however, it assigns credit proportionally, giving more weight to channels closer to purchase. EXAMPLE: A customer visits a website via Google Ads, then sees an ad on Facebook and interacts with an organic search result before making a purchase. Under time-decay attribution, organic search would get most of the credit, while Facebook and then Google Ads would get proportionally less, depending on how you decide to allocate it.
You can choose from all of these attribution models in Google Analytics, depending on your business needs and strategy (see more on that below). However, keep in mind that if you’re analyzing data in Glew, you’ll want to stick to first-order or last-click attribution.
Why multi-channel attribution is important
These days, the customer journey is anything but linear. Every sale is different - you’re promoting your products and interacting with your customers through multiple channels, and processing orders in various places (from marketplaces to subscriptions to ecommerce to point of sale). A customer might have multiple points of contact with your channels and campaigns before ultimately making a purchase. Attribution modeling helps you determine which channel gets credit (and how much) for each purchase - which in turn helps you better understand your channel performance and make more informed decisions going forward.
Gain insight into the customer journey
Understanding the path that your customers take on the way to purchasing can tell you a lot about your customers and their buying behavior. Do you tend to have customers who interact with multiple channels and come back to your website several times before purchasing? That might indicate that they’re doing research and evaluating other products before purchasing - some educational content could be helpful. Do you have a lot of success converting customers quickly from paid ads? It may be worth increasing your budget.
Understand your channel performance
Landing on a successful attribution model can help you get to the holy grail of marketing reporting - which channels are actually helping you drive the most sales? Not to mention which ones add the most revenue, have the highest margin, help you acquire the most new customers and have the highest LTV - all metrics available by channel in Glew. There’s nothing more important when it comes to evaluating your marketing activities and allocating resources and advertising dollars wisely.
Develop smart strategies for future marketing efforts
At the end of the day, gaining insight into your customer journey and evaluating your channel performance helps you with one major thing: developing smart, data-driven marketing strategies going forward. Attribution modeling helps you figure out what’s working and what isn’t, where you should invest more and where you should cut back. It helps you ensure your marketing strategy remains profitable and you put your dollars - and resources - behind strategies that drive results.
How Glew solves multi-channel attribution
In order for orders to be matched to the correct channels, your ecommerce or analytics platform needs to be able to identify where a customer came from before making a purchase. This typically happens through a process called channel mapping, used by Glew (and other platforms, like Google Analytics) to identify the correct source, medium and campaign for purchases in order to attribute them to the correct channel. Channel mapping works through UTM tagging, using the unique identifiers added to a URL to indicate the referring source, medium and campaign for purchases made through that URL. For example, a URL you use for a Facebook ad campaign might look like this: https://yoursite.com/utm_source=facebook&utm_medium=paidsocial&utm_campaign_springpromotion Most channel mapping models work by simple exact matching based on medium - for example, if the medium contains exactly “facebook,” “google,” or “email,” it will match it to the correct channel. Google’s channel mapping model specifically works this way - if the medium exactly matches the specific rules they have defined for their channels, it will map correctly. If not, it will go into an “unknown” bucket.
However, as you might guess, this method has some significant drawbacks:
- Exact matching fails to capture close-match terms (like “Facebook” and “facebook” or “video” and “videos”)
- Basing channel mapping on medium alone leads to a large number of orders falling into “undefined” or “unknown” - if the medium can’t be matched to a very specific set of rules, the channel is recorded as unknown
- Since it’s based on UTM identifiers in URLs, it’s limited to only orders that take place through your ecommerce store - it won’t help attribute orders from marketplaces like Amazon, Facebook and Google, orders placed through a subscription platform like ReCharge, or retail orders
Glew’s channel mapping model
Glew uses a similar channel mapping model to the standard one used by Google Analytics - however, we enhanced our model in a few key ways to help you better attribute orders with the correct channel. We also map purchases to channels based on the unique UTM parameters available in the URL, starting with the medium - but we take a few extra steps to ensure we’re matching channels accurately and capturing and attributing all the purchases we possibly can. Here’s how Glew’s model is different:
- Glew uses a waterfall model to identify and attribute all possible orders. First, we first try to exactly match medium or source to our predefined channels. If that’s not possible, we try to find the best match to our predefined channels - like “affiliates” and “affiliate.” Finally, if that’s not possible, we’ll take your medium or source and add it as a channel in Glew, even if it’s not in our predefined list.
- Glew’s model can account for non-web orders (including marketplaces like Amazon, Facebook and Google, subscriptions from ReCharge, retail point of sale and draft orders) for Shopify by defaulting to the order source from Shopify
- Glew’s channel mapping doesn’t rely on exact match - we capture close match and common variations (like “video” and “videos”) - and isn’t case sensitive, so we’ll understand that “Facebook” and “facebook” are the same thing Glew breaks out paid advertising into separately rather than bucketing them into one channel, including Facebook, Google, Instagram and AdRoll (Google buckets all paid advertising into one channel)
Because of these added steps, if you use both Glew and Google Analytics, you’ll notice some differences in attribution on a channel level, since Glew is able to attribute purchases with more precision and match more channels. Click here to see Glew’s full channel mapping model, including what’s included in each channel.
How Glew compares to attribution in other platforms
Since there are so many different ways of doing attribution, you’ll see different models - or slightly different takes on similar models - across analytics and ecommerce platforms. Here’s how Glew’s attribution models compare to other models you may encounter:
Google’s attribution models
Google Analytics allows you to choose between several attribution models, including Last Interaction, Last Non-Direct Click, Last Google Ads Click, First Interaction, Linear, Time Decay and Position-Based. Last Interaction is closest to Glew’s last-click model - Glew's model starts with Google's rules, with some additional capabilities and matching added on.
Shopify’s attribution models
Shopify offers two attribution models: first interaction and last interaction.
- First interaction: Assigns credit for a sale to the channel by which customers first discovered your store. This is essentially first-click attribution, in which the first click that brings a visitor to a website is given credit for a sale that takes place.
- Last interaction: Assigns credit for a sale to the last marketing activities that customers engaged with before buying. This is most closely related to last-click attribution in Glew and last interaction attribution in Google.
Understanding which attribution model to use
The attribution models we’ve outlined here all assign sales credit to channels in different ways, and all help you measure different things, depending on the goals of your business. Which one makes sense for you? That depends on your marketing strategies and what you’re trying to understand through your reporting. In Glew, we provide the option for you to toggle between first-order (credit for all sales to the channel that drove the first sale) and last-click attribution (credit for a sale to the channel of the customer’s last click prior to purchasing) because we believe these two models do a great job of giving you a holistic view of your channel performance from two different perspectives: long-term and lifetime value-focused (first-order) and immediate-term and conversion-focused (last-click). Here are a few pointers on which attribution model to use, depending on what you’re trying to understand: You want to tell which channels are driving the most lead generation: First-click You want to tell which channels are driving the most conversion: Last-click You want to understand which channels are driving the most value over time: First-order You have a longer buying cycle and want to look at all touchpoints: Linear, position-based or time-decay
Wrapping up attribution modeling and channel mapping
No matter which way you go about it, attribution modeling is something that’s important for every business to master. It helps you understand your buyer journey, gain insight into which of your marketing channels are bringing in the most new customers and revenue, drive smart decision-making around where to invest future resources and more. As we wrap up, remember these key takeaways:
- Glew allows users toggle between first-order and last-click attribution, providing a holistic view of channel performance that can be either lifetime value- or conversion-focused
- Other common attribution models include first-click, linear, position-based, time-decay and more - you can switch between these and other attribution models in Google Analytics
- Multichannel attribution in analytics platforms works through a method called channel mapping, where platforms match online orders to the correct channels using unique UTM identifiers
- Many channel mapping models, including Google Analytics’, works by mapping the UTM medium exactly to their specific set of predefined channels and rules, leading a lot of orders to be bucketed into an unknown/unattributed channel
- Glew’s channel mapping model takes several steps to account for this, including not relying on exact match and creating channels for any unknown mediums
And don’t forget why attribution is so important for businesses:
- It helps you gain insight into your customer journey, telling you a lot about the people who buy from you and the path(s) they take on the way to purchase
- It helps you understand your channel performance, including which channels are actually helping you acquire the most customers and drive the most sales
- It helps you develop smart marketing strategies, including dedicating resources to the most effective, profitable channels and figuring out where you should cut back
Want to see how Glew solves channel mapping for multichannel merchants for yourself? Start a free trial today to get insight into your attribution.
Get started with Glew Glew helps you understand attribution from all your channels - from Facebook to Google Ads to email, affiliates and beyond - so you can make smarter marketing decisions. Start a free trial to check it out: