A Refresher on Marketing Attribution
Table of Contents:
Introduction
If you spend time and money on Digital Marketing, it is likely that you want to measure the results of your efforts. The way you measure your results will define how you optimize your campaigns and that will be absolutely key to the success of your business. However, modern marketers face one major challenge regarding how they are measuring user engagement and ROAS. It is the way they look at their data. We tend to measure results by traffic source, programmatic, search or social, rather than having an omni channel approach and a general understanding of what we produce overall across all digital media.
The solution is to build a truly unified Attribution Model. Understanding how the various traffic sources influence each other within your Marketing mix is one of the holy quests of every marketer who is trying to accurately evaluate the return on their Marketing Investments.
What is a Marketing Attribution Model?
A unified Marketing Attribution Model is the best way to ensure that credit is given where it is due. It is the best way to make sure you are not going to miss out on the channel, tactic or campaign that initiated a journey for a converting user. If you have that visibility, you will take more informed optimization decisions using accurate data, which should lead to better overall results.
Over the last decade, the world has become more connected than ever, the mobile phone trend has reached 5 billion plus unique users globally (GSMA Intelligence 2020) and the revenue of online businesses keeps increasing at an astonishing rate every year. This new, and more complicated take on a user’s journey online, has created new challenges for Direct Response and Performance marketers everywhere. It has rendered some classic measurement and attribution tools obsolete and inaccurate. Having a unified Marketing Attribution Model across all the online channels you are buying traffic from is the most effective way to respond to this multi-device, user behavioral shift.
Why create a Marketing Attribution Model?
We all want to get the best out of our Marketing investments, but our perception of what constitutes a good return on investment can be biased by data making you feel like you are doing as well as you possibly could. I recently came across a marketing professional that told me that his CRM (salesforce) was the best attribution tool available on the market today. I shared my surprise and disagreement with him, citing the fact the majority of his employer’s traffic took place on mobile, which is known for cookie/IDFA dropping. I also mentioned the fact that Salesforce doesn’t do view through conversions, or that it is only using ‘last click’ to attribute conversions at the campaign ID level. He seemed confused. Not everybody is aware of how poorly they may be compiling their conversion data or how inaccurately they are looking at it.
So whether or not you can identify the areas where you could improve, why should you create a unified Marketing Attribution Model for your organization?
First, to avoid the aforementioned issues created by relying on your CRM for attribution, which can be summarized in an overall lack of visibility that will lead you down a path of bad decision making. And second, it will help you understand who your customer is, what is the online journey of a converting prospect and how to leverage it to find more like minded customers. It will also show you what conditions have driven purchase intent for your product or your services and how you can recreate those conditions.
The Marketing Attribution Model will help you dedicate time and money to the touch points that really generate revenue for your business, wherever those touch points might be.
Which Marketing Attribution Model?
There are two main components to achieving a truly unified Marketing Attribution Model for the traffic you buy online, across all your channels:
- Using the same set of pixels across all channels, for impressions and clicks.
- Using the actual type of Marketing Attribution Model you think represents best the average purchase intent for your product or service. Details below.
The Marketing Attribution Model type you pick will define how you give credit where it is supposedly due. The default attribution model you will find on most platforms is ‘last click’. Obviously ‘last click’ is not ideal if you aim to gain complete visibility and If you want to attribute credit where it is really due. That being said, it can be useful in some cases. For example, I personally like to use ‘last click’ in the early days of a new campaign or project, especially with Search Engine Marketing, just to get an idea of the type of conversion volume I can expect, and more importantly to help me reach the necessary conversion volumes required to get started on automated conversion optimization. Once you are ready to move on from ‘last click’, you should rely on the following predefined models:
Predefined models:
Linear:
Each touch on the conversion path gets the same amount of credit.
Time Decay:
The touch closer to the conversion gets the most credit, the rest of the credit is distributed decrescendo back to the first touch of the conversion path.
Reverse Time Decay:
The touch starting the conversion path gets the most credit and the touches after that get the rest of the credit decrescendo all the way to the last touch.
More models exist in the predefined category but I feel like these three are the best to get started with.
The data driven and custom models below require more time and more data to reach their maximum potential, if the Marketing Attribution Modeling is still an early concept for you I would suggest starting with predefined models.
Data Driven Model:
Data Driven:
This model uses your data to determine which touch points have the most impact on your bottom line, and it builds the according attribution model for you.
Custom Models:
This one allows you to build your own attribution model as you see fit. This requires a personal deep dive into your existing data.
Setting Up your Marketing Attribution Model:
Now that you have picked a model all you have to do is pick the set of pixels you will be using to track the traffic you are buying.
Your choice in this regard may be influenced by the sources of traffic you are focusing on.
Search Engine Marketing (SEM):
Everybody should be buying a healthy amount of SEM traffic, it is how you close the loop, when the customer intent is at its highest. Keep in mind that Google generates 85% to 90% of the world’s SEM traffic (outside mainland China). That is the reason I would most definitely recommend a google product if your mix is including significant SEM traffic. The Search layer of the Google Marketing Platform is called SA360 and is probably the best Search automation tool out there in 2021.
Display and Programmatic:
I personally believe that the best Programmatic platform out there today is DV360, another Google product, formerly DoubleClick Bid Manager(DBM), its algorithm is without a doubt the most successful, its targeting and reporting capabilities are unparalleled and it has preferential access to YouTube inventory.
For direct buys, CM360 (also part of the Google Marketing Platform) provides a really cost effective ad server for campaigns with publishers that will not host actual ads but only redirect links
It is also important to mention that the Google Marketing Suite makes it easy to append your tracking scripts onto click URLs of third-party Programmatic platforms, like Adroll or Amazon. So you can buy from many different publishers and still have a truly unified attribution model. Not all publishers give you that option.
Social Media:
Social media is usually a significant part of an advertiser’s overall online strategy. For B2C products or services FB is an absolute must mainly due to scale. For B2B, LinkedIn is the platform that will offer you the most granular and sophisticated way to reach the best fitting audience for your business. The likes of Reddit and Twitter are useful for Branding but usually not the best converters. The Google Marketing Platform allow you to append tracking scripts to the aforementioned publishers URLs (and thus complete your unified attribution model throughout the entire funnel), with some restrictions however:
FB will only give you conversion data (view throughs and click throughs) for reach campaigns. For every other type of campaign, you will only get impression data, but without visibility on actions that are assisting conversions. There are ways to circumvent this issue, however none are producing results as accurate as the ones you would see across the rest of the funnel.
Twitter is not giving advertisers a lot of flexibility when it comes to creating custom trackable events, it is apparently being worked on. Just be aware of that limitation if you are already buying Twitter traffic.
Some other tech providers propose Marketing Attribution Modeling solutions, but if the bulk of the traffic you are going to buy is Google labelled, why would you not also trust them to help build the Marketing Attribution Model for your business?
The Google Marketing Platform is the best option on the market today to help you build a successful customized Marketing Attribution Model for your business. And we have already helped many great organizations build these customized solutions. It often comes with revolutionizing the way they are looking at Marketing metrics and measurement tools, it takes time and effort but in the end, this new approach never fails to help them increase their conversion volume and decrease their cost per conversion.
General Observations
Even if you do everything by the book, a unified Marketing Attribution Model across your entire digital spectrum can take some time to show the expected success. The first stage of that evolution is that it will redistribute the credit you allocate to various channels, publishers or campaigns. Consequently it might seem like it is having a negative impact on your former conversion champions. I was raised on SEM tactics before real attribution modeling was even a thing, SEM traffic is expensive and is your best converter, I understand that nobody wants to see its performance decrease. However your SEM campaigns will start looking a bit more expensive with a new Marketing Attribution Model, you will also start seeing real traction on upper funnel tactics, and the traffic that used to show limited results will start getting credit for conversions. It is about having a mid to long term vision of your goals.
Remember that the whole point of having an unified Marketing Attribution Model across the whole online spectrum is giving credit where it is due.
While your bottom funnel tactics will look a bit more expensive, the upper funnel ones should make up for that, leaving your overall cost per conversions in the same range as before.
It is only after a few weeks or months, depending on spend and market conditions, that you will start seeing the real benefits of a unified Attribution Model.
With better data, and a more accurate overall vision of what your Marketing investments produce you will simply take better optimization decisions, and with that will come drastically better results down the line.
After spending more on upper funnel tactics, the ones you never knew were producing or guiding your users on a converting path, the virtuous cycle you initiated will impact your entire Marketing spectrum and bottom funnel tactics will benefit from that change as well. It will result in an increase in conversion volume, a decrease in cost per conversion or maybe even both.
- Redistribution of conversion credit, lower funnel looks more expensive, upper funnel starts producing.
- With better visibility, comes better decision making, and better budget allocation. Spend more on conversion assisters and initiators.
- Get more overall conversions and/or decrease your cost per conversion.
Given enough effort and time this conclusion is inevitable, regardless of what industry or geolocation you are in.
I hope you enjoyed reading this article as much as I enjoyed writing it.
Depending on the device you are currently using or the publication you found this on, you should be able to send me comments, questions or all sorts of requests about the above material. I would be more than happy to discuss with you how we could help your business create the perfect customized Marketing Attribution Model and get a lot more out of your Marketing investments.
Franck Lacroix for Huddlefifty.
Franck@huddlefifty.com
Contact us to setup a free consultation on your Marketing needs.
Privacy Policy | Terms of Use | Cookie Declaration |
Huddle Fifty
99 Wall Street #425, New York, NY 10005