Introduction
Attribution modelling is one of my favorite topics while dealing with Google analytics. It’s complicated, challenging and not anyone can fake their knowledge when it comes to attribution modelling. It is something which requires a deep understanding of a business model and how different marketing channels work together to create a customer journey.
So before you start understanding what is attribution modelling, please make sure that you understand your business model, your industry and target market.
Because if you don’t, you would end up making applying/creating some wrong attribution models and lose a lot of money. Attribution modelling goes beyond Google analytics and we can categorize it as a completely separate topic and not just a subtopic of google analytics
But this blog talks about Attribution modelling with respect to google analytics since the majority of businesses today use Google Analytics as the primary analytical tool.
What is Attribution Modelling?
Attribution modelling is the process of understanding the buying behavior of your customers and to determine the most effective marketing channels for investment.
Why Attribution Modelling is Important?
As I said earlier, attribution modelling is a complex subject and the reason is that it deals with topics and analysis which cannot be carried out by an inexperienced professional. It deals with topics which seem intangible for a person unaware of how their business model work.
You should use attribution modelling to understand the buying behavior of your website visitors:
- Why do people buy from your website?
- What content do they consume before they buy?
- What happens before they buy from your website?
- What prompted them to complete the predefined goal?
The biggest insight that you can get from the attribution model is that you can determine the most effective marketing channels for investment.
Some marketers still think that tracking conversion is the only way to understand the effectiveness of a marketing channel. But it does not give you the complete picture.
Some marketing channels may not be directly linked with conversions but may have started the conversation or assisted in the decision-making process. So before you discard any marketing channel as ineffective, try to gauge these two questions:
- The number of conversions initiated by the marketing channel
- The number of conversions assisted by the marketing channel
For example – A person may have heard about your offer via an email marketing campaign, later they might have visited your blog to read about the product features and a few days later visits the website directly to make a purchase.
Here your email marketing campaign initiated the conversation, blogging channel assisted it and the customer ended up making a purchase by visiting your website directly.
In the case of multi-channel marketing, one channel is not solely responsible for conversions.
Before we deep dive into Google Analytics, you need to understand some terminologies, which will help you better understand this blog post.
Acquisition Channels
Also known as ‘marketing channels’, ‘digital channels’ or ‘channels’ are the sources of traffic coming to your website. For example – ‘paid search’, ‘social media, ‘email’, ‘referral’ etc. In multi-channel marketing, acquisition channels are just called ‘marketing channels’ or ‘channels’.
Google analytics provide your reports about how your website is acquiring the visitors.
Hence to understand your acquisition report, you need to first understand what are channels and how they are defined in google analytics.
In order to understand channels, you first need to understand: traffic sources, medium and campaigns.
What are Traffic Sources in Google Analytics?
Traffic source is the source of your website’s traffic. It also includes the traffic tracked from ‘utm_parameter’
In the case of google/cpc – Google is the source of traffic
In the case of yahoo/organic – Yahoo is the traffic source
In the case of MakeMyTrip/referral – makemytrip is the source of traffic
What is Medium in Google Analytics?
Medium is the category of the traffic source as defined by Google. It also includes traffic medium tracked via utm_medium parameter.
In the case of google/cpc – cpc is the medium
In the case of Yahoo/organic – organic is the medium
In the case of MakeMyTrip/referral – referral is the medium
What Is Campaign In Google Analytics?
The campaign is the name of your Google AdWords campaign or any other custom campaign.
A custom campaign is that marketing campaign which has been tagged with campaign tracking parameter ‘utm_campaign’.
What Is Channel In Google Analytics?
In google analytics, ‘channel’ refers to the group of traffic sources with the same medium.
For example – ‘organic marketing’ is a marketing channel. It can be made up of following traffic sources.
- Google (as in google/organic)
- Yahoo (as in yahoo/organic)
- Bing (as in bing/organic)
Acquisition > Overview report
What is Conversion in Multi-channel Funnel Report?
The definition of conversion in multi-channel reports is slightly different. It encompasses both the goal conversion and an e-commerce transaction. The total conversions in multi-channel funnel reports are the sum of total number of Goal conversions and the total number of e-commerce transactions.
Introduction to the Google Analytics Attribution Models
An attribution model is a set of rules that determine how value for sales and conversion path should be attributed to various touch points of a user’s conversion path.
Attribution models are broadly classified into two categories:
- Baseline Attribution model
- Custom attribution model
Baseline Attribution Model
There are basically 8 types of baseline attribution models:
Last interaction attribution model: This is a default setting in the google analytics multi-channel report. This attribution model applies 100% credit to the last interaction that took place before the conversion.
First interaction attribution model: As the name indicates, this model applies the credit to the first interaction in a conversion path.
Linear attribution model: This model equally divides the credit to each interaction in the conversion path.
Time Decay attribution model: This model assigns more credit to the interactions which are closest in time to the conversion.
Position based attribution model: By default, this attribution model applies 40% credit to the first interaction, 20% credit to middle interaction, and 40% credit to last interaction.
Last non-direct click model: This model assigns all the credit for conversions to the last non-direct click in a conversion path. Google Analytics uses this model by default for non-Multi channel funnel reports.
Last Adwords Click: This model assigns 100% credit to last adword click in the customer’s conversion path.
Data-driven attribution model: This model uses an algorithm to credit value to the interaction that takes place in the various phases of a conversion path.
There is no one size fits all approach. You need to have a deep understanding of your business model to correctly implement these attribution models. A clear framework of advertising objectives must be laid down before deciding which attribution models to follow.
Custom Attribution Models
As the name indicates, these models are designed by individuals to meet their specific requirements.
When you build your own custom attribution model, you assign your own rules about how to assign a value for interaction along each customer path, This is termed under ‘custom credit rules’ in Google Analytics.
Attribution Reports in Google Analytics
Google analytics provide the following attribution report:
- Model Comparison Tool
- ROI analysis (Available only in Google 360)
- Data-driven analysis (Available only in Google 360)
You can access these reports by navigating to: Conversions > Attribution in your GA view:
Model Comparison Tool
The model comparison tool, as the name suggests is used to compare different attribution models in google analytics.
Through model comparison tool you could compare different baseline attribution models and custom built models to each other. This is used to gain the understanding the perspectives of different marketing channels.
You can also use this tool to create custom attribution models in Google Analytics.
ROI Analysis in Google Analytics
You can do ROI analysis in Google analytics by analyzing the ‘ROI analysis’ and ‘Cost Analysis’ reports.
Through this report, you can calculate ROAS (Return on Advertising Spend) of various marketing campaigns in different attribution models.
Don’t let the name misled you, ROI analysis in Google analytics is done by analyzing ROAS. Even though its name says ‘ROI’ analysis.
Through the cost analysis report, you can determine ‘Cost per click’ and ‘ROAS’ of all marketing channels for which you have downloaded cost expenditure reports.
In ROI analysis, you can determine ‘cost per acquisition’ and ‘Return on Advertising spend’ for each marketing channel under different attribution models.
Understanding User Journey
User does not follow your desired path: There is no guarantee that a person clicking on your Adword link will navigate through different pages of your website before making a purchase.
Users switch between marketing channels: A user may visit your website through social media but can make a purchase through Google organic search or search engine ads. You may conclude that your social media is now working but in reality, it is.
Users may not always use the same device to access your website: The whole multichannel marketing term arises from the fact that consumers today are accessing websites and content from different devices. The conversation may start on a desktop but later move to mobile and consumer may end up getting converted over a tablet.
Google analytics treats these interactions as a different user as cookie information is not shared across devices.
This is how Google Analytics will record these interactions:
User 1 accessed your website through google ad but didn’t make a purchase, so google ad will not get the credit of conversion.
User 2 browsed your product pages but didn’t make a purchase.
User 3 directly visited your website and via a mobile device and made a purchase.
How Different Marketing Channels Affect Each Other?
Because of multi-channel marketing and multi-device attribution, any improvement or decline of one particular channel may affect the performance of the other. So if you suddenly switch all your PPC campaigns, you may witness a decline in direct traffic.
Organic traffic can affect the CTR of PPC ads, as people who see both your organic listing and paid ad in the search engine, they are more likely to click on the paid ad. As your organic ranking decline, it may adversely affect your CTR of PPC campaigns. Similarly, people who are exposed to a brand for the first time via social media may return to the website via branded organic search or directly. So any decline in social media efforts may negatively impact branded organic search traffic or direct traffic.
In the multi-channel world, we always optimize our channels to create a seamless user experience.
5 Keys To Success In Attribution Modelling
Understand Your Audience
Understanding your customer purchase journey is the key to fixing the attribution problem.
No any tool or marketing trick will help you if you fail to understand your customers. Not just the attribution battle, but any battle in the business and marketing world.
Understand the Concept of Missing Touchpoints
Google analytics attribution modelling is based on only known touch points. That is it accounts for only those touch points which can be tracked and recorded.
Thus there is no guarantee that your attribution modelling knows all the touch points involved in a customer journey.
Fixing Data Integration Issues
You should always aim to minimize the missing touch points in your conversion path by integrating as much data as possible from different data sources. With so many tools at disposal gathering data has become easy, almost every business is doing it in a small or large scale proportion. The success lies in extracting insights from the data available, marketers should use data integration to drive marketing strategies and the first step is to centralize all your available data.
Attribution is Beyond Analytics
Attribution is much more than google analytics. Don’t limit your capabilities by using just google analytics attribution models.
There are a lot of attribution tools available in the market. After all, its the custom made attribution model that will actually make a difference, Especially for large organizations.
Understand Not All Touchpoints Are Equally Valuable
It is important to know that not all touch points are equally valuable in a conversion path.
The marketing channels which assist the most should get the maximum credit for conversion and maximum resources should be allocated to it regardless of whether it’s first middle or last interaction. All the other touches get the credit in proportion to the contribution it makes in the conversion path.
