Based in Ottawa, canada, Tuesday Standard is a blog about consumer insights, digital measurement, online engagement and marketing.

What is Attribution and how to Deal with the Analytics behind it?

What is Attribution and how to Deal with the Analytics behind it?

Last week we explored the topic of analytics and measurement at different levels, depending on the purpose and audience of the insights. Next, I thought it was important to dig deeper into what makes up a detailed view of analytics – one where the nuances of visitors and how they behave and what got them to take the final action you wanted them to take (i.e. online purchase, sign-up for social action, create account, etc.).

Attribution is a marketing term that focuses on the customer journey (used liberally – could be citizen, etc.) and is the concept of giving credit to the various touchpoints or campaigns or channels that a customer interacts with on its way to a decision. Effective analytics programs provide the necessary data to the organization to understand what touchpoints attribute to a sale or conversion that is important.

It seems simple enough when you think about the final transaction. Your analytics will tell you what happened in the buying or converting visit – what keyword was used to get the visitor to your site – what email link did they click on – what social media post did they respond to. But how often does a customer ever go straight to a website and make a purchase? Rarely. Multiple channels and messages were often responsible for the final purchase or conversion decision. In an ideal world, you would be able to track the entire customer journey from start to finish with the use of old fashioned personal anecdotes from each customer about why they made the decisions they did along the way. But that’s not realistic, or scalable.


To determine what was involved in the journey, and how the customer was influenced to make their decision, there are various attribution models that organizations can go by. These models could be divided into 3 categories (this from various sources but has an interesting take on it: ):

1.  Single source attribution – where credit is given to the first or last touch with the customer before the purchase or conversion happens. The first touch is the channel that the customer first engaged with – the very first thing they clicked or saw to get them to your site and then eventually purchase. It focuses on the beginning of the journey which can happen over time. Last touch focuses on what happened just before the purchase happened and ignores the full customer journey over time. What was the last influencing step before the button.

2.  Multi-Source Attribution – where credit is given to each channel along the journey that contributes to the final conversion or sale. While each channel receives credit, there is little knowledge of which one played the greatest part in the customer’s final decision. What was the tipping point when the journey includes a search engine click, then a newsletter sign-up, then a webinar, then a demo, etc.? Multi-source attribution models include: (i) linear, (ii) time decay, (iii) u-shaped, (iv) w-shaped, (v) full path, and (vi) custom.

3. Weighted Multi-Source Attribution – where the touchpoints along the journey are weighted depending on the part you feel has done the heaviest lifting in influencing the customer to make a purchase or conversion. This is a complex category and one that can be challenging to get completely accurate.

The attribution model you choose is important for organizations as they look at trying to decide which of the various channels they are using give the most credit for sales, and ultimately which one to fund or allocate the greatest amount of time to from a marketing perspective. It also helps decision where marketing ends and sales begins – which has always been an often contentious conversation when the sales process is longer and more complex with high value purchases.

From an analytics or measurement perspective, here is an example of what the journey could look like and the challenge with measuring influence along the way.

XYZ Company is looking for data on which of their marketing and communication efforts has the most impact in getting a customer to purchase one of their offerings. A user starts with a Google search of a generic industry term, sees their ad on LinkEDIN for a product specific webinar and signs up, later they receive a follow-up email to a white-paper, then a direct mail postcard highlighting customer stories, then visits your website where they end up converting.

Attribution tells you the actual contribution of each of these touchpoints:

  • Google search 45% responsible for the conversion
  • LinkEDIN 15% responsible for the conversion
  • Webinar 15% responsible for the conversion
  • Email 10% responsible for the conversion
  • Direct mail 10% responsible for the conversion
  • Final direct web visit was only 5% responsible for the conversion

Being able to provide a complete picture of this data to the marketing, communications, and/or sales teams is an important aspect of the program. Many marketers only look at the analytics data, but this may not be enough to tell the proper story. Attribution needs to be included in the analysis.

Google has a great analogy here: think of attribution as the peanut butter to analytics' jelly. Yes, each is great on its own, but for many, they're even better together. With both analytics and attribution in measurement, you can establish a complete measurement plan with key performance indicators (KPIs) that take all touchpoints into account. With this information, your organization is better able to manage marketing spend to best reach, influence, motivate, and drive your audience to those places where they can engage and convert.

As you look at your attribution needs, there are a few questions you can ask yourself to determine if you are in fact not telling the complete story with your analytics:

  1. Do I see all touchpoints when looking at my marketing and sales funnel? If your analytics tells the full picture, this may be all you need. If not, layering in attribution features may be required.
  2. What added insights will attribution provide? Will the added complexity add value?
  3. Does Marketing know what to do with this added data and does the analytics team understand how to put the measurement in place to capture the attribution data?
  4. Does the cost and effort to collect this new data provide positive ROI in your situation to make it worthwhile, or is it just noise that complicates what should be a fairly simple process for decision making? 

Working together, the various stakeholders involved in arriving at an attribution model that works for your organization, and implementing the appropriate measurement and tracking mechanisms (analytics, CRM, Marketing Automation, etc), you can build the appropriate attribution framework that meets your needs. The proof will be in more accurate story telling about why your customers do what they do – and better allocation of resources to capture more visitors and turn visitors into customers.

The bottom line is that attribution for your organization’s marketing efforts may be a necessary element of your measurement strategy. As analytics professionals, we should proactively work with marketing or communications to make sure everyone is on the same page with how your measurement is configured and what is being reported so there is no confusion.

I would be interested in hearing what your organization is doing about attribution and whether this is a group strategy or handled solely by the marketing or communication department with tools such as and Marketing Cloud or Pardot, Eloqua, Marketo, etc. In my organization, we handle the analytics around marketing campaigns combining the data from our analytics tool with the CRM and marketing automation software in place.

Thank you. Please let me know if you have any questions or comments.

Conversion Metrics: Not just for E-commerce

Conversion Metrics: Not just for E-commerce

What is your story's perspective? Telling stories with analytics at different levels...

What is your story's perspective? Telling stories with analytics at different levels...