A Practical Guide To Multi-Touch Attribution

Posted by

The consumer journey includes multiple interactions in between the customer and the merchant or service provider.

We call each interaction in the customer journey a touch point.

According to Salesforce.com, it takes, usually, six to 8 touches to produce a lead in the B2B area.

The variety of touchpoints is even greater for a client purchase.

Multi-touch attribution is the mechanism to assess each touch point’s contribution towards conversion and offers the appropriate credits to every touch point associated with the consumer journey.

Performing a multi-touch attribution analysis can help online marketers understand the customer journey and determine opportunities to further enhance the conversion paths.

In this short article, you will discover the fundamentals of multi-touch attribution, and the actions of performing multi-touch attribution analysis with easily available tools.

What To Think About Before Performing Multi-Touch Attribution Analysis

Specify Business Goal

What do you wish to achieve from the multi-touch attribution analysis?

Do you wish to examine the return on investment (ROI) of a specific marketing channel, understand your customer’s journey, or determine important pages on your site for A/B testing?

Various business goals may require various attribution analysis methods.

Defining what you want to achieve from the beginning assists you get the outcomes much faster.

Specify Conversion

Conversion is the desired action you want your customers to take.

For ecommerce websites, it’s normally making a purchase, specified by the order conclusion event.

For other industries, it may be an account sign-up or a subscription.

Different kinds of conversion likely have various conversion paths.

If you want to perform multi-touch attribution on several preferred actions, I would suggest separating them into different analyses to prevent confusion.

Specify Touch Point

Touch point could be any interaction in between your brand name and your clients.

If this is your first time running a multi-touch attribution analysis, I would recommend defining it as a check out to your website from a specific marketing channel. Channel-based attribution is simple to conduct, and it could give you an overview of the consumer journey.

If you wish to understand how your customers engage with your website, I would suggest specifying touchpoints based upon pageviews on your website.

If you wish to include interactions beyond the site, such as mobile app setup, e-mail open, or social engagement, you can include those events in your touch point definition, as long as you have the information.

Regardless of your touch point meaning, the attribution system is the very same. The more granular the touch points are specified, the more detailed the attribution analysis is.

In this guide, we’ll concentrate on channel-based and pageview-based attribution.

You’ll find out about how to use Google Analytics and another open-source tool to conduct those attribution analyses.

An Intro To Multi-Touch Attribution Models

The methods of crediting touch points for their contributions to conversion are called attribution models.

The simplest attribution model is to give all the credit to either the very first touch point, for bringing in the client at first, or the last touch point, for driving the conversion.

These 2 designs are called the first-touch attribution model and the last-touch attribution model, respectively.

Clearly, neither the first-touch nor the last-touch attribution design is “reasonable” to the rest of the touch points.

Then, how about assigning credit equally throughout all touch points associated with transforming a customer? That sounds reasonable– and this is exactly how the linear attribution model works.

However, assigning credit uniformly across all touch points assumes the touch points are equally important, which doesn’t seem “fair”, either.

Some argue the touch points near the end of the conversion paths are more important, while others favor the opposite. As an outcome, we have the position-based attribution design that enables marketers to offer various weights to touchpoints based on their places in the conversion paths.

All the designs discussed above are under the category of heuristic, or rule-based, attribution models.

In addition to heuristic models, we have another design classification called data-driven attribution, which is now the default design utilized in Google Analytics.

What Is Data-Driven Attribution?

How is data-driven attribution different from the heuristic attribution designs?

Here are some highlights of the distinctions:

  • In a heuristic design, the guideline of attribution is predetermined. Regardless of first-touch, last-touch, linear, or position-based design, the attribution rules are embeded in advance and then used to the information. In a data-driven attribution design, the attribution rule is produced based on historic information, and for that reason, it is unique for each circumstance.
  • A heuristic model looks at just the courses that cause a conversion and overlooks the non-converting courses. A data-driven model utilizes data from both converting and non-converting courses.
  • A heuristic design associates conversions to a channel based upon how many touches a touch point has with regard to the attribution guidelines. In a data-driven model, the attribution is made based on the impact of the touches of each touch point.

How To Assess The Effect Of A Touch Point

A common algorithm used by data-driven attribution is called Markov Chain. At the heart of the Markov Chain algorithm is a concept called the Elimination Effect.

The Removal Effect, as the name recommends, is the impact on conversion rate when a touch point is eliminated from the pathing information.

This post will not go into the mathematical details of the Markov Chain algorithm.

Below is an example illustrating how the algorithm attributes conversion to each touch point.

The Removal Impact

Assuming we have a scenario where there are 100 conversions from 1,000 visitors pertaining to a website by means of 3 channels, Channel A, B, & C. In this case, the conversion rate is 10%.

Intuitively, if a certain channel is gotten rid of from the conversion courses, those paths involving that particular channel will be “cut off” and end with less conversions overall.

If the conversion rate is reduced to 5%, 2%, and 1% when Channels A, B, & C are removed from the data, respectively, we can compute the Elimination Effect as the percentage decline of the conversion rate when a specific channel is gotten rid of using the formula:

Image from author, November 2022 Then, the last step is attributing conversions to each channel based upon the share of the Removal Effect of each channel. Here is the attribution result: Channel Removal Impact Share of Elimination Result Associated Conversions

A 1–(5%/ 10% )=0.5 0.5/(0.5 +0.8+ 0.9 )=0.23 100 * 0.23 =23 B 1–(2%/ 10%
) = 0.8 0.8/ (0.5 + 0.8 + 0.9) = 0.36 100 * 0.36 = 36
C 1– (1%/ 10% )=0.9 0.9/(0.5 +0.8 + 0.9) = 0.41 100
* 0.41 = 41 In a nutshell, data-driven attribution does not rely on the number or

position of the touch points however on the impact of those touch points on conversion as the basis of attribution. Multi-Touch Attribution With Google Analytics Enough

of theories, let’s take a look at how we can utilize the ubiquitous Google Analytics to perform multi-touch attribution analysis. As Google will stop supporting Universal Analytics(UA)from July 2023,

this tutorial will be based on Google Analytics 4(GA4 )and we’ll utilize Google’s Product Store demonstration account as an example. In GA4, the attribution reports are under Marketing Snapshot as shown below on the left navigation menu. After landing on the Advertising Photo page, the first step is selecting a proper conversion occasion. GA4, by default, includes all conversion events for its attribution reports.

To prevent confusion, I extremely advise you pick just one conversion occasion(“purchase”in the

listed below example)for the analysis. Screenshot from GA4, November 2022 Understand The Conversion Courses In

GA4 Under the Attribution area on the left navigation bar, you can open the Conversion Paths report. Scroll down to the conversion course table, which reveals all the courses causing conversion. At the top of this table, you can discover the typical variety of days and number

of touch points that lead to conversions. Screenshot from GA4, November 2022 In this example, you can see that Google customers take, on average

, practically 9 days and 6 visits before making a purchase on its Merchandise Store. Find Each Channel’s Contribution In GA4 Next, click the All Channels report under the Efficiency area on the left navigation bar. In this report, you can discover the attributed conversions for each channel of your selected conversion event–“purchase”, in this case. Screenshot from GA4, November 2022 Now, you understand Organic Browse, together with Direct and Email, drove the majority of the purchases on Google’s Merchandise Shop. Examine Outcomes

From Various Attribution Designs In GA4 By default, GA4 uses the data-driven attribution design to identify how many credits each channel receives. However, you can take a look at how

different attribution models appoint credits for each channel. Click Model Contrast under the Attribution section on the left navigation bar. For instance, comparing the data-driven attribution model with the first touch attribution model (aka” first click design “in the below figure), you can see more conversions are attributed to Organic Search under the very first click design (735 )than the data-driven design (646.80). On the other hand, Email has actually more associated conversions under the data-driven attribution model(727.82 )than the very first click model (552 ).< img src="// www.w3.org/2000/svg%22%20viewBox=%220%200%201666%20676%22%3E%3C/svg%3E" alt="Attribution models for channel grouping GA4"width=" 1666"height ="676 "data-src ="https://cdn.searchenginejournal.com/wp-content/uploads/2022/11/attribution-model-comparison-6371b20148538-sej.png"/ > Screenshot from GA4, November 2022 The data informs us that Organic Browse plays an essential function in bringing prospective clients to the shop, but it requires help from other channels to convert visitors(i.e., for consumers to make real purchases). On the other

hand, Email, by nature, engages with visitors who have checked out the site previously and helps to convert returning visitors who initially pertained to the website from other channels. Which Attribution Model Is The Best? A common concern, when it concerns attribution model comparison, is which attribution design is the very best. I ‘d argue this is the wrong question for marketers to ask. The reality is that nobody model is absolutely much better than the others as each design highlights one aspect of the consumer journey. Online marketers need to welcome multiple models as they see fit. From Channel-Based To Pageview-Based Attribution Google Analytics is easy to use, however it works well for channel-based attribution. If you want to further comprehend how customers browse through your website prior to converting, and what pages affect their choices, you require to carry out attribution analysis on pageviews.

While Google Analytics does not support pageview-based

attribution, there are other tools you can utilize. We recently carried out such a pageview-based attribution analysis on AdRoll’s site and I ‘d more than happy to share with you the steps we went through and what we discovered. Collect Pageview Series Information The first and most tough step is collecting data

on the series of pageviews for each visitor on your site. Most web analytics systems record this information in some type

. If your analytics system doesn’t provide a method to extract the data from the user interface, you might require to pull the information from the system’s database.

Comparable to the actions we went through on GA4

, the primary step is defining the conversion. With pageview-based attribution analysis, you likewise require to determine the pages that are

part of the conversion process. As an example, for an ecommerce website with online purchase as the conversion occasion, the shopping cart page, the billing page, and the

order verification page become part of the conversion process, as every conversion goes through those pages. You must omit those pages from the pageview data because you do not need an attribution analysis to tell you those

pages are essential for converting your clients. The function of this analysis is to understand what pages your potential consumers visited prior to the conversion event and how they affected the consumers’choices. Prepare Your Data For Attribution Analysis Once the data is ready, the next step is to sum up and manipulate your data into the following four-column format. Here is an example.

Screenshot from author, November 2022 The Course column reveals all the pageview sequences. You can utilize any unique page identifier, but I ‘d advise using the url or page path due to the fact that it permits you to analyze the outcome by page types utilizing the url structure.”>”is a separator utilized in between pages. The Total_Conversions column reveals the total variety of conversions a specific pageview path resulted in. The Total_Conversion_Value column reveals the overall financial value of the conversions from a particular pageview path. This column is

optional and is mostly suitable to ecommerce websites. The Total_Null column reveals the total variety of times a particular pageview path stopped working to convert. Construct Your Page-Level Attribution Designs To construct the attribution models, we leverage the open-source library called

ChannelAttribution. While this library was originally created for use in R and Python programs languages, the authors

now provide a totally free Web app for it, so we can use this library without composing any code. Upon signing into the Web app, you can publish your information and begin constructing the designs. For first-time users, I

‘d suggest clicking the Load Demo Data button for a trial run. Make certain to examine the parameter setup with the demo information. Screenshot from author, November 2022 When you’re prepared, click the Run button to develop the models. As soon as the models are created, you’ll be directed to the Output tab , which shows the attribution arises from four different attribution models– first-touch, last-touch, direct, and data-drive(Markov Chain). Keep in mind to download the outcome data for further analysis. For your referral, while this tool is called ChannelAttribution, it’s not limited to channel-specific information. Since the attribution modeling mechanism is agnostic to the kind of data provided to it, it ‘d attribute conversions to channels if channel-specific data is supplied, and to websites if pageview information is offered. Analyze Your Attribution Data Arrange Pages Into Page Groups Depending on the variety of pages on your site, it might make more sense to first evaluate your attribution information by page groups instead of individual pages. A page group can include as couple of as just one page to as numerous pages as you desire, as long as it makes good sense to you. Taking AdRoll’s site as an example, we have a Homepage group that contains simply

the homepage and a Blog site group which contains all of our post. For

ecommerce sites, you may think about organizing your pages by product classifications too. Beginning with page groups rather of specific pages enables marketers to have an overview

of the attribution results throughout various parts of the website. You can constantly drill below the page group to individual pages when needed. Recognize The Entries And Exits Of The Conversion Courses After all the data preparation and model building, let’s get to the fun part– the analysis. I

‘d recommend very first identifying the pages that your possible consumers enter your site and the

pages that direct them to transform by examining the patterns of the first-touch and last-touch attribution designs. Pages with particularly high first-touch and last-touch attribution values are the starting points and endpoints, respectively, of the conversion paths.

These are what I call entrance pages. Ensure these pages are optimized for conversion. Keep in mind that this kind of gateway page might not have really high traffic volume.

For example, as a SaaS platform, AdRoll’s prices page does not have high traffic volume compared to some other pages on the site but it’s the page lots of visitors checked out prior to transforming. Discover Other Pages With Strong Influence On Clients’Choices After the gateway pages, the next action is to find out what other pages have a high impact on your consumers’ choices. For this analysis, we search for non-gateway pages with high attribution worth under the Markov Chain models.

Taking the group of product function pages on AdRoll.com as an example, the pattern

of their attribution worth across the four models(revealed listed below )shows they have the greatest attribution worth under the Markov Chain design, followed by the direct design. This is an indicator that they are

visited in the middle of the conversion courses and played an important function in influencing consumers’decisions. Image from author, November 2022

These kinds of pages are also prime prospects for conversion rate optimization (CRO). Making them simpler to be discovered by your site visitors and their material more convincing would help lift your conversion rate. To Summarize Multi-touch attribution enables a company to comprehend the contribution of different marketing channels and identify opportunities to further enhance the conversion paths. Start simply with Google Analytics for channel-based attribution. Then, dig deeper into a customer’s pathway to conversion with pageview-based attribution. Do not stress over selecting the best attribution model. Take advantage of numerous attribution designs, as each attribution model reveals different aspects of the consumer journey. More resources: Included Image: Black Salmon/Best SMM Panel