A Marketing Attribution Model that Works for Marketing Automation
The past few months we have been working with the team at PerfectoMobile to help take advantage of everything Marketo has to offer. One key project goal was to conduct multi-touch attribution without Marketo’s Revenue Cycle Explorer. The team at Perfecto wanted to know why leads responded to an offer and where they came in from to be able to measure marketing offer and channel effectiveness. A few of the key questions posed included:
- Did that newsletter sponsorship generate better webinar attendance than our benchmark?
- How much did that cost?
- Did a particular tactic lead to revenue and did that get associated to the content?
- Is content marketing achieving an ROI?
That’s a big challenge and we gave it a lot of thought on white boards real and virtual.
As I worked on this, I thought about a conversation I had with Jonathan Beaton about Lead Sources. He pointed out that most firms have a list similar to these:
But a lead does not appear from a Webinar or White Paper. The content is why they are there, not how. So the standard list of Lead Sources confuses content with channel. As Jonathan noted,
“Why is ‘Web’ not a lead source? Web is not a lead source because visitors to your website may be there for a variety of reasons.”
Similarly, a whitepaper is not a “source” of a lead. As I pondered this question for a great solution, I went into my friend Jeff Coveney’s presentation on Lead Sources at the 2014 Marketo Summit. I found his idea that you could track the Source and the Asset detail one part of the puzzle, especially for naming Sources and for First Touch, Last Touch tracking. I wanted to develop this system further, so I kept working at it.
One way to look at this situation is to ask, “What is the origin of the Lead and where were they going?” I like to think of this as the Origin and the End Point.
If we take a look at this image, we can see that a Channel Origin can go to more than one End Point. I might promote my event on a PPC platform while also using my in-house list and an outbound call campaign. Longer term and large marketing programs might use five or ten promotion platforms to get the word out.
And it is a good idea to remember that your promotion channels are promotion platforms, just as if you were standing up in the town square and had a loudspeaker. There are several platforms in that square and each one has a different audience. Thus, Google AdWords has one, very large group, while your trade association has a smaller, focused platform.
From a systems perspective, your Channels and database might look like this. As you can see, promotion platforms drive traffic to a website (or landing page) and then into Marketo.
With this understanding, I began working out the best methods to understand the data we could collect that would explain how well marketing was doing. When we came to PerfectoMobile, we needed to look at how to separate Lead Source data from Content request data in Marketo and Salesforce. Thus, we created a new attribution concept: Offer and Channel.
The Offer and Channel Attribution Method
The Offer is the content–the reason the lead entered your database. The Channel is how your lead got to your site in the first place.
- Offer Examples: webinar, roadshow, whitepaper, blog post, free trial, or demo.
- Channel Examples: Adwords, Banner Ad, Paid Email Blast, Content Syndication, LinkedIn
One way to think about it is that Offers are content, while Channels are equivalent to Lead Sources or “platforms” used to promote the Offer. You always promote an Offer on a Channel, not the other way around.
With that in mind, we developed several iterations of the actual system to collect data. As we met to finalize the details, Linda Duchin, Demand Generation Director, said,
“When designing tracking models, I try to picture myself as the lead recipient and supply them with the most valuable information for follow up: Why did the lead respond, and how did they find out about us.”
Immediately I saw the simplicity she crystallized for all marketers who need closed loop attribution. This goes right back to elementary school essay lessons: the 5Ws and the 1H: Who, What, When, Where, Why, and How.
Let’s take a closer look at how this concept works. Most marketing automation platforms (MAPs) take care of the Who and When because those are baseline features. Tracking things like the What, Why, Where, and How are a bit trickier. Usually those items have to be tracked and setup by the user (meaning you) with Forms, hidden fields, and URL parameters. A system like Marketo will let you track the Offer and the Channel via the Program concept. Yet, in Marketo you can only choose Channel or Offer with the Program Channel Type. For Perfecto, we chose to track Offer at the Program level.
In your MAP, the flow could look like this, with URL parameters driving data collection in your MAP. We will talk about how that is done in more detail next time.
Once you understand these concepts, the hard part is putting together a system that lets you track each of these touches and report on them. Marketo already handles the Who and the When. The individual touch data will be used to answer the questions Perfecto posed at the outset:
|Question||How We Answer It|
|Did that newsletter sponsorship generate better webinar attendance than our benchmark?||
|How much did that cost?||Program Cost DataSFDC Cost Data|
|Did a particular tactic lead to revenue and did that get associated to the content?||Use a campaign or RCE cross tab to show Offer X used Channel Y to obtain 58 leads and 4 opportunities with $120,000 in pipeline|
The next step is to develop the system for First, Last, and Multi-touch attribution to record the entire history of interactions and then be able to report on that history in a way that meaningful decisions can be made.
In upcoming posts, we will discuss just how to do this using Marketo and SFDC with, and without, Revenue Cycle Explorer.