What is multi-channel attribution and why should B2Bs care?
The B2B marketer today operates across a wide selection of channels. Paid ads in the form of Google ads, Capterra or Linkedin, organic, email, referrers, social, you name it.
Yet, 48% of marketers struggle to understand the value these channels bring and the role they play in the customer journey. And even when they do, they struggle to compare their efforts in each channel like-for-like. That is, some channels start journeys, others finish them, others have no impact at all. Let’s dig into multi-channel attribution models.
As we’ll cover in the post below, this is due to the inherent limitations of the platforms they’re working with in their tech stack. These are more often than not siloed and ill-equipped for multi-touch and account-based attribution. Presenting customer journeys that are disconnected, when they are really dots on the same line.
Because of these limitations, marketers (and their bosses) are unable to gain insight into how their marketing channels are helping generate pipeline and revenue, and delivering ROI.
A B2B revenue attribution tool offers these insights, by connecting all the dots on the customer journey. And in this post, we’re going to show you how.
More on revenue attribution models here.
To do this, we’ll be answering the following questions:
What is multi-channel attribution?
Why is channel attribution important?
Why do B2B marketers struggle to link channel performance to revenue?
How does multi-channel attribution overcome these challenges?
What does multi-channel attribution actually look like?
Already sold on multi-channel attribution, 👆 jump straight here
What is multi-channel attribution?
Let’s start with a couple of basic definitions to make sure we’re all on the same page.
What is a (digital) marketing channel?
To borrow Wikipedia’s definition, “Digital Marketing Channels are systems based on the Internet that can create, accelerate, and transmit product value from producer to a consumer terminal, through digital networks.”
In layman’s terms, channels are platforms or methods you use to reach your target audience. In digital marketing we’re talking: paid ads, emails, referrals, social media, search (organic), etc.
What is multi-channel attribution?
Channel attribution, or multi-channel attribution, is the process of attributing credit to these channels based on the customer’s touches, and where they happen, in the buying cycle.
Why is channel attribution important?
Most readers will already be guessing at the obvious answer to this question, but it’s worth laying it out here all the same.
In fact, the reason most readers will be (painfully) aware of the answer is that you likely fall into the category of: “63% of marketers under extreme pressure to deliver revenue growth” (CMO Council Report 2021)
This highlights the fact that management and the company leadership are growing less interested in the number of MQL conversions you’ve generated. Instead, they’re interested in pipeline and revenue generated. After all, these better inform ROI on each channel, marketing performance generally, and by extension, budget allocation.
In reality, these metrics are not only about appeasing leadership and keeping one’s skin in the game. Even where these pressures are not being exerted, accurate revenue performance of channels can help the multi-channel marketer optimise and scale their efforts.
In this way we can say that multi-channel attribution models are important for:
Understanding channel acquisition performance by revenue and pipeline generated
Running cross channel attribution - to see channel impact across the funnel
Comparing channel cost and performance against revenue
How multi-channel attribution models can help B2B marketers who struggle to link channel performance to revenue
So, given that it’s so important, why aren’t all marketers attributing revenue and pipeline generated to their channels?
Well, it turns out that accurately bridging marketing efforts to revenue is not the easiest of endeavours, especially for B2Bs - with our long and complex customer journeys.
As mentioned in the intro, the reason for this is that the tools sitting on the B2B tech stack are siloed and simply ill-equipped for the multi-touch and account-based attribution necessary for B2B multi-channel attribution.
Specifically, B2B marketers encounter three problems:
Not connecting and storing first-party data from across the ecosystem
Not viewing the customer journeys as account journeys
Running a stopgap single-touch attribution solution that doesn’t overcome these first two points👆
Let’s unpack these a little further:
1. Connecting and storing first-party data from across the ecosystem
In order to bring all the channels under one roof and measure them against revenue and pipeline generation, it’s necessary to link data from across the revenue ecosystem.
Problem is, almost without exception, these tools are siloed. Meaning that the customer journey data is disjointed. So any attempt to bring it under one roof is pretty much impossible.
B2B marketers lack the on-site tracking, integrations of ad platforms and CRM, and data processing (joining, cleaning, storing) to allow them to create a unified customer journey - to which they can run attribution.
2. B2B customer journeys as account-based journeys
Even where there is tracking of (parts of) the journey, the tools focus on the individual user and not all the stakeholders - account - that form part of the buyer journey.
In the typical B2B buying scenario, the user to first interact with your brand is typically not the one signing off on the deal at the end of it.
For this reason, not working with accounts (by storing, identifying and linking and users into accounts - i.e. connecting all the data) it is impossible to gain all the benefits of B2B multi-channel attribution.
3. Stopgap single-touch attribution solutions don’t solve these 👆
“Wait a minute. I can attribute with my CRM, the Original Source field tells me where (the channel) my leads are coming from.”
Yes, that’s right. And yes, this does mean that in these cases there is a link between data sitting in different tools.
However, relying on a CRM’s “original source” (whether HubSpot, Salesforce or other) field to describe the initial stage of the already complex B2B buyer journey, can only ever deliver but a fraction of what’s actually taking place.
That’s because using a single-touch model, only attributes credit to one single player. But this methodology means not capturing the entirety of the customer (account) journey, and so missing out crucial touches which might happen further down the funnel - think retargeting ad.
Similarly, it means attributing incomplete data - a big no, no. The issue here becomes that you’re not only missing out crucial touchpoints, but that in doing so you’re giving greater credit to a touch/channel than it is actually owed. Something that becomes particularly problematic when scaling, as you might end up throwing tons of cash on a channel that might not have been driving the lead down the funnel.
You can read more about the Original Source here.
“What about Google Analytics’ ‘multi-channel funnels’? That’s multi-touch, isnt’ it?”
Yes. But, while it offers a multi-channel view of conversions (not necessarily revenue), it suffers from three important limitations:
Google Analytics tracks individual devices, not accounts
Google Analytics’s “look-back window” is only 90 days long
Only Google data is accurately tracked
You can catch more details on the limits of Google Analytics in this post, and of Google’s Click ID here.
Look at how these approaches compare for a single customer before and after switching to multi-channel attribution. 👇
How does multi-channel attribution overcome these challenges?
Multi-touch attribution, ideally through a B2B attribution tool like Dreamdata, overcomes these challenges by creating a single unified customer journey where all touches across all channels are stored and processed in a single platform.
A single, unified customer journey
The central element of multi-touch attribution is taking into account every one of the touches that happen across the customer journey. This allows marketers to see the impact of all the events that took place on their channels - and when!
How does it do this?
Introduces a tracking script on your website to track all anonymous_id/user_id and url (more on Dreamdata’s tracking script here).
Makes sure auto-tagging and tagging on your channel platforms, such as Google Ads, is on.
Fetches reports from your channels, such as Click Performance Report from Google Adwords API to link the GCLID with the relevant campaign, ad group and ad
Connects this data with data from across your ecosystem through integrations, such as Hubspot CRM.
Stores all this lead and customer data in a data warehouse.
Merges and cleans the data to make sure there’s no duplication, empty values, etc.
Runs multi-touch attribution modelling on the data, including running reports by pipeline stage as well as by revenue generated.
Displays analytics and reports on dashboards. Or alternatively, invites customers to pull their clean, modelled data, to their own preferred BI tools and dashboards.
Multi-channel attribution modelling
Let’s dig a little deeper with the most iconic part of the process: the attribution models.
Important caveat: the attribution modelling will only be as good as the data that is fed into it. Which means that you need to nail the data collection process set out above before you can get the most out of your attribution!
And even still, it’s impossible to get 100% accuracy - we can’t read human minds!
Or as George Box puts it,
“All models are approximations. Essentially, all models are wrong, but some are useful”.
Models:
First touch attribution model - 100% of the credit is given to the first step of the journey.
Last touch attribution model - The same as first touch but to the final step of the journey.
Linear attribution model - Distributes credit evenly to every single touchpoint in the customer journey before converting.
Time-decay attribution model - Similar to Linear attribution, but gives more credit to the touchpoints closest to the conversion.
U-shaped attribution model - Tracks every single touchpoint, but gives 40% credit to the anonymous first touch and 40% credit to the lead conversion touch. The remaining touchpoints get 20% split between them.
W-shaped attribution model - Same as the U-shaped model, but the W-shaped model also emphasises the opportunity creation touchpoint. These three key touchpoints receive 30% credit each, and the last 10% is split evenly among the remaining touchpoints.
Custom attribution models - Where a data scientist helps build an attribution model that best suits your specific customer journey.
Ideally, you want to have multiple models in place to allow you to compare how your channels are performing at different stages - you can find an example of this below.
Get the full lowdown on all things multi-touch attribution models here.
What does a multi-channel attribution model actually look?
Now that we’ve got a clearer picture of what happens under the hood, we can take a sneak peek at the actual outputs of multi-channel attribution.
Broadly speaking, there are two main analyses that can be done when running multi-channel attribution:
Revenue (and pipeline) attribution: Where you get a snapshot of how your channels are performing in terms of revenue and pipeline generated. And, from which you can deep dive into campaign level analysis, for optimisation, scaling, etc.
Attribution model comparison (cross-channel attribution): Where you gain a better understanding of when and how each of your channels and sources contributes to your company’s buyers journey. This not only gives you better insight into your funnel but also the chance to optimise content and ads to where in the funnel they best perform.
Let’s look at some examples.
*N.B in these examples, we’re looking at the Dreamdata B2B revenue attribution platform.*
In this report, we find the snapshot of channel and source performance by deals and revenue generated. That is, the product of achieving the Holy Grail of connecting channel activity to revenue and pipeline generated.
It also includes the more traditional metrics of the number of visitors, contacts (individual identified users in the pipeline), and companies.
From this report we can learn that direct traffic, paid ads (on Google), have been the main driving force in generating deals and traffic.
It also tells us how valuable (in terms of revenue) each channel is by traffic generated. Which in turn can influence planning and investment.
For instance, here we see that Organic traffic generated 1,571 visitors which produced 19 deals. Doubling organic traffic, say through a determined SEO campaign, could yield a better deal-to-visitor ratio than Paid.
The same data can then be reported in time-limited ways, or to include specific campaigns, and measure against deals and revenue independently, as shown in the charts below. This can help better understand what within channels and when has worked.
Cross-channel attribution
What this reporting doesn’t tell us, however, is when in the pipeline these channels are helping move the customer forward. For this, we need to compare attribution models.
Here we can see more clearly that Direct peaks under the last touch model. This means, that it’s likely the touch that takes place when Sales send buyers to read documentation or make the purchase.
The example also looks to have a very well working email flow that both contributes to initiating customer journeys and to a lesser degree closing them.
A final word on multi-channel attribution
The verdict is in. Multi-channel attribution has become as relevant to the B2B marketer’s toolkit as to B2B board rooms.
The need for multi-channel attribution arises from having to:
Understand channel acquisition performance by revenue and pipeline generated
Run cross-channel attribution - to see channel impact across the funnel
Compare channel cost and performance against revenue
However, the typical marketing tech stack, with all its siloed data and misleading single-touch logic, prevents the B2B marketer from doing this. That is unless they set up a B2B revenue attribution solution.
An attribution tool like Dreamdata helps overcome these challenges by creating a single unified customer journey where all touches across all channels are stored and processed in a single platform. From which multi-channel attribution modelling can be run.
