Profile picture for user Andrew Nguyen

In this post we review multi-channel attribution, reviewing the attribution capabilities and methods available through different technologies. We’ll also discuss how to evaluate which multi-channel attribution model is best for your organization. 

What is Multi-Channel Attribution?

Multi-channel attribution is a set of rules that assigns credit for sales and revenue to touchpoints across the customer journey.

What’s a touchpoint? It’s simply information on a point of engagement that describes who, what, when and where. It’s a lot like point-of-sales information, which describes the who, what, when and where of a purchase. For B2B marketers, touchpoints are basically a data table with each row being a touchpoint and each column having information describing that touchpoint, like the person, opportunity, content, web page, form, location, time, campaign and revenue value, associated with that touchpoint. Multi-channel attribution has to do with that last part: revenue.

Attribution refers to the rules and methods used to assign value across multiple channels and touchpoints. Revenue amount is typically the value that is assigned to touchpoints.

You can assign conversion credit to touchpoints too. Conversion means any actions that are deemed valuable. You can also assign creation credit, telling you the number of leads or opps that were created from a group of touchpoints.

Why Care?

The purpose of attribution is to help marketers understand what marketing actions and touchpoints influence revenue and conversions. Without attribution, we’re left with massive data tables that don’t provide insights like how much value we generated from marketing efforts.

Once marketers know what’s valuable and what’s not, it opens up a whole new set of actions to pursue. Actions around optimization, budgeting, and strategic planning are all based on understanding value.

So we know that marketing attribution can be important to your business and your role as a decision maker, let’s now discuss the different technologies and implications for how you’ll go about assigning value to your marketing actions, campaigns, and investments. 

Reviewing What’s Available To You Now

Adopting multi-channel attribution doesn’t happen overnight. It’s a technology and measurement method that affects multiple teams, requiring coordination and change management.

If you already have Google Analytics, Salesforce and marketing automation like Marketo, you already have options for implementing marketing attribution. You may quickly find that you have needs beyond these existing technologies. So let’s evaluate what you get with existing technologies.

Multi-Channel Attribution With Google Analytics

Google Analytics provides a variety of multi-channel models, from time decay to custom. But the weakness here is that reporting is all in Google Analytics and revenue and opportunity data is housed in the CRM, which tracks information on upsells, opportunities in an account, account-based marketing campaigns and offline marketing.

In addition, while digital marketing likely gets a healthy level of investment, there are other channels that customers engage with along the buyer journey, including sales activity with BDRs, attending events/conferences, and offline ABM (mailers and field marketing).

Google Analytics is a great start if your buying cycle happens exclusively on your website, like ecommerce. If you have a complex sales cycle spanning multiple channels, online and offline, and have a sales team, then a more complete attribution model is needed.

Salesforce Marketing Attribution

With Salesforce you are highly limited in multi-channel attribution reporting. In Salesforce, marketing attribution is based on campaign membership and is highly simplified. You can do single-touch attribution and very rudimentary multi-channel attribution.

You can attribute revenue to campaign membership in binary fashion. If a prospect is part of a campaign and an opportunity is created, that campaign gets credit for the entire opportunity. Even if they are part of multiple campaigns, revenue gets credited to the last campaign associated with the contact. This is essentially last click attribution.

You can attempt to do multi-channel attribution with Salesforce with campaign influence reporting. But the major downside to this is that it credits the entire opportunity amount to each associated campaign, leading you to over-report pipeline and revenue.

Using Salesforce alone is not a viable option. It’s covered in greater detail in this article, Marketing Attribution, How It’s Done In Salesforce

Attribution With Marketing Automation

Lastly, let’s look at marketing automation like Marketo RCA. For attribution within marketing automation there are a few rules to keep in mind.

Marketo takes the entire opportunity amount and splits it evenly, no matter which model is being used (first touch or multi-channel)

Credit cannot be given after an opportunity closes. For example, if a customer engages with a campaign after an opportunity closes, that touchpoint won’t get any credit. This can lead to under reporting of customer marketing and upsells. Solutions like Bizible assign revenue credit to post opportunity creation and customer marketing for more complete reporting.

Lastly, rules in marketing automation like Marketo don’t allow for time-based weighting, meaning if a deal takes a year to close, the touchpoints all get equal credit. With more advanced models, more weight is placed on the touchpoint right before opportunity creation.

There are several more limitations. This comprehensive review of capabilities and limitations for marketing attribution using Marketo, by LeadMD, is a great read. 

We’ve reviewed using existing tools for revenue attribution and now it’s time to review the variety of multi-channel models and whether they are a good fit for your go-to-market strategy.

List of Multi-Channel Attribution Models

In this section we evaluate the different multi-channel attribution models and the scenarios where they work the best This is a great section if you’re wondering which attribution model to use.

See a more a complete list of marketing attribution models, including single touch models.

Linear Attribution Model  


The Linear Attribution Model has strengths and weaknesses. Its strength is that it is unbiased in how it distributes revenue credit. It simply credits each touchpoint with conversions or revenue equally across the customer journey.

This may be helpful if you’re interested in understanding the customer journey and want to look at touchpoints in aggregate. For example, if you invested in a PR campaign and want to know if a referral touchpoint is present in the customer journey, you could use a linear attribution model to surface touchpoint information, counting how many referral touchpoints there are. But, if you’re concerned with revenue then a linear model may not be the best.

A linear attribution model doesn’t overtly show which touchpoints are influential, and this can lead to data overload when you have thousands of touchpoints that all get credited with the same amount of revenue. Your data starts to get very noisy.

While you can study engagement, i.e. the amount of channel touches or digital campaign touches with a linear model, there are better multi-channel models that can help you define more clearly which touchpoints are important in terms of revenue.

Use a linear model If you want to understand the customer journey, study engagement, and have a shorter sales cycle. Keep in mind, studying the many different funnel paths is a difficult endeavor.  

Time Decay Multi-Channel Model


A time decay multi-channel model gives more revenue or conversion credit to touchpoints that happen closer to a closed deal. This can be helpful for B2B organizations with longer sales cycles and who are investing in sales enablement.

If you are doing sales enablement activities and account-based marketing, a time decay model can help you understand what marketing campaigns are influencing deals and revenue by virtue of the touchpoints happening closest to the time of winning a deal. You’re giving credit to your bottom-of-funnel activities.

With a long sales cycle, it doesn’t always make sense to give too much credit to touchpoints that are years old. If you’re a marketer that spends a lot of effort in bottom-of-funnel tactics then a time decay model can help you make sense of the effectiveness of those efforts.   

Position-Based Attribution Models


Position-based attribution models give more revenue credit to certain positions. By position, we mean stages of the buying journey. This model is the preferred model for B2B marketers who have adopted a clear framework for understanding their buyer journey, and who separate their broad marketing actions based on specific goals such as lead generation, opportunity generation and net new customers.

A position based model gives more credit to stages like First-Touch, Lead Creation Touchpoint, and the Opportunity Creation Touchpoint. These touchpoint positions represent an important conversion or stage: brand discovery, lead conversion, and opportunity conversion, respectively.

Use a position based model if you have teams and campaigns dedicated to improving conversions or performance at each stage of journey. A position based model like W-Shape, U-Shape, and Z-Shape are all appropriate here.


Custom Attribution Models

Custom models can also be thought of as position based and are a feature for marketers who want greater customization in how they credit revenue to the different positions.

If you have a more granular sales cycle than the one described above, then a custom model is right for you. For example, you may have more stages in your buyer journey than is prescribed in existing position based models like W-Shape

Another use case for a custom model is if a larger company has multiple campaigns, and these campaigns have different buyer journeys and stages. Because these campaigns have different GTM strategies they each need their own custom model.

Adopting multiple custom models make a lot of sense for large marketing organizations with different divisions, different product lines, different target geographic regions, and different GTM strategies.  

Algorithmic Attribution Models

Algorithmic Attribution Models are the most advanced models to consider. So far we’ve covered models with revenue distribution rules that are based on time, position, and simple division (linear). These rules are determined by marketing leaders in a somewhat arbitrary fashion based on what they think are important groups of touchpoints.

An algorithmic attribution model creates rules for distributing revenue credit based on an analysis of all your past touchpoints data and deciding which touchpoints are predictive of revenue.

The difference here is that the rules are created by data scientists and algorithms predict the successful outcome of a deal or the amount of revenue. In predicting these outcomes or amounts, touchpoints data is used and touchpoints are given weights based on how predictive they are of a deal closing (or revenue).

There are a variety of scenarios where an algorithmic model makes sense. If small differences in weights matter to you then it makes sense to get help creating a customized model for your business.

A good analogy here is a nutrition plan. At a broad level, a simple nutrition plan works for most people who want to make minor improvements in their diet and health. But for an athlete who has to perform at peak conditions, what’s required is a personalized nutrition and diet plan based off of the athlete’s specific body type and performance requirements. For athletes, a small change can yield relatively high gains, or many small changes result in even larger gains. 

For complex businesses with many channels, campaigns and touchpoints, a custom algorithmic model is a good solution. These models make a lot of sense for marketers in mature markets where improving efficiency is important but also difficult. An advanced multi-channel model can improve reporting and help marketers shift their spending based on more specific weighting.

If an algorithmic model shows that certain touchpoints or channels should receive 35 percent of revenue credit, compared to 33 percent under a W-Shaped model, this might not be a big deal for smaller companies. The result is a minor difference in recommended spending changes.

But a 1-2 percent spending change in a channel or demand generation campaign can amount to a lot of money if you’re talking about multi-million dollar campaign budgets. So you’d want to be right in how you configure the rules in your attribution model.

Small percentages of large dollar amounts are still large dollar amounts. So if you have large campaign budgets and need highly detailed attribution rules to help with planning and optimization, then algorithmic attribution is for you.


We hope this article has helped you better evaluate your multi-channel marketing attribution needs. To better help in your evaluation process, here are additional resources:

We’ve covered in detail the why of attribution, what you get with widely adopted marketing technologies, and how to evaluate the need for more advanced models.

For a complete ebook on multi-channel attribution, download the guide below.