understand the contribution of each of your Marketing efforts – WAU
Multichannel Attribution is a strategy based on the use of different channels, direct or indirect, with the aim of diversifying points of contact with the target audience.
If you have already managed a Marketing team or strategy, you have certainly had to answer at least one of the following questions:
- Which channel generates the most results?
- In Marketing, what type of strategy brings more customers?
- Where do most of our customers come from?
- How to allocate budget for the different Marketing strategies?
This type of questioning comes from all sides, from Marketing professionals, co-workers, the financial department to the CEO of a company. And they are capable of driving any Marketing manager crazy. After all, how to determine the right answer?
In other words, if the Marketing strategy was a football team, which player would take credit for the goal?
If you’ve ever had that kind of doubt, you’re not alone! In this article, I will explain to you how the channel assignment problem arises and what your options are for solving it.
When you finish reading, I hope you have a clearer view of your strategy, and that you can start making more accurate decisions!
When does attribution become a problem in your life?
To define the importance of studying attribution, we need to go back to one of the most basic theories of Marketing: the Marketing Mix. She says that success in acquiring customers does not come from one place. On the contrary, it is necessary to balance different elements and reach potential customers in different ways so that they come to buy from your company.
Likewise, a football team made up only of center forward would not be so successful, right?
Therefore, a good Marketing strategy invests in several points of contact. And the more robust your strategy gets, the more you invest in different channels. It is even more difficult to determine what contributed to a potential customer becoming a customer!
When you need to set budgets, allocate resources and make strategic hires for a Marketing team, you need to understand which channels bring the best results. And having an attribution model will help you make the best decisions in the medium and long term.
So far you have already understood that the attribution of Marketing is the process that helps to determine which tactic takes credit when bringing customers, and understanding this is important for making strategic decisions.
But why, exactly, is it difficult to know who was responsible for a purchase? Shouldn’t that be a simple answer?
You’ve also caused headaches for professional colleagues
Imagine the following situation:
You will start training for a marathon and need to buy running shoes. One day, you go to Google and your search for “running shoes” begins to see what’s out there on the market. You even go to the website of a tennis brand that a friend indicated and even sign up for your email to receive news, but in the middle of your search, a call interrupts and you decide to leave your purchase for later.
A week later, you go to Facebook and see in a advertisement that the sneakers you had liked is 10% off at the store. You click on the ad, but you are still in doubt and decide to wait to receive your salary to buy the sneakers.
You wait another week for your salary to drop, and when that happens you promptly returns to the running shoe website but see that the 10% discount no longer exists. As it is an expensive item, you decide to look for other cheaper options and go back to Google.
While you’re looking for other options on Google, a message on Whatsapp distracts you. After reading the message, do you decide to check your inbox before going back to Google, and what are you faced with? a email offering 25% discount on the tennis you wanted so much! You click on the email, access the brand’s website and buy the sneakers!
Now let’s imagine that you are the Marketing Manager for the running shoe brand.
Your company is cutting marketing costs and you need to determine which channel brings more customers to be able to prioritize resource allocation.
In the situation described above, who was responsible for taking you to the purchase?
- The SEO strategy that placed the company’s website in the first results for “running shoes”?
- The Facebook ad that increased your interest in the product with a discount coupon?
- Or the email that sent you the second discount coupon that actually led to your purchase?
Surely you have also been through this type of questioning (otherwise you would not be reading this article). It is important to know that this happens in all companies, and that is why some attribution models that help to determine, in each company, which channel takes credit.
The 3 universal attribution challenges
When trying to determine which part of Marketing contributes most to the acquisition of customers, we come across several fronts. You might want to know, for example:
- What was the impact of an event on an online purchase? Or
- How many people accessed content in an app and then made a purchase? Or
- How do top funnel content impact customer conversion?
Among thousands of other questions that arise daily.
Although everyone wants to give credit to a certain thing, each of the points above represents a different challenge, and needs to be solved differently.
Therefore, we can divide the attribution challenges into 3 categories:
Every marketing operation has at least one of them! Do you identify where your difficulties fit?
1. Impact of Online on Offline (and vice versa)
If your company has physical stores, you will easily identify with this challenge.
Can you tell how your online strategies impact store results? What is their role in the customer’s life until he enters the store? Who is more important: a Facebook post for the company or a promotional ad in the store window?
The challenge of determining the impact of online on offline, and vice versa, happens even in companies that do not have physical stores. Sponsoring an in-person event, for example, adds one more variable to the customer acquisition equation.
2. Impact through different devices
We have known for a long time the importance of a strategy that reaches the potential customer on different devices.
Being present at various times is essential to maintain the relevance of your brand. But cell phones and tablets brought yet another challenge in the lives of marketers: how to follow a customer’s journey through different devices?
How do I get a person’s activity history as they change from device to device?
3. Impact of multiple channels on a conversion (Multichannel Attribution)
This is the challenge we presented with the example at the beginning of the article, where a person wanted to buy running shoes and had contact with several channels before making the purchase.
This is where we’ll find templates to help allocate resources and people to each channel in your Marketing strategy.
Multichannel attribution is the focus of this article for being one of the great challenges for marketers. What can we use to solve it?
Multichannel Attribution Models: how to choose the most important player
When it comes to multichannel attribution, the final decision of which channel contributed most to a key conversion will always be arbitrary. That’s because there is no right answer.
The right answer to attribution questions will depend on your strategy, your company and the decision makers.
There are some models that we can use to help determine how we are going to see multichannel attribution. Therefore, whenever an attribution challenge arises, it is necessary to make a strategic choice as to which attribution model will be used.
Read each one carefully before deciding which one makes the most sense for your company’s reality!
To understand each of the multichannel attribution models, we will use as an example the journey we described at the beginning of the article. That is, from a person who needed to buy running shoes and had contact with the company in the following order:
- Organic traffic
- Facebook ad
- Direct traffic
- E-mail marketing
- Conversion (purchase)
Last click / last interaction attribution
This is the model that gives 100% credit for the last contact channel.
In our example, email marketing would take all the credit for the purchase of running shoes, while the other channels would not take credit.
If the company had to choose a channel to reduce costs, it would most likely not be Email Marketing.
No attribution model is perfect (in fact, they are far from it). To make the right choice, it is important to understand the pros and cons of each.
Let’s go to the positives and negatives of the last click / last interaction attribution!
In addition to being simple to understand, the attribution of last click / last interaction is one of the easiest to implement.
Tools like Google Analytics, for example, allow you to see the last channel visited by the user before a conversion.
In addition, it’s easy to calculate a channel’s share in a conversion, as a single channel will always earn full credit.
This type of attribution completely ignores the participation of other channels in the conversion.
The downside of this is that each channel has a different role in the marketing strategy, and consequently some act more at the time of conversion than others. For example, an AdWords ad is more likely to generate more running shoe sales than an Instagram post, since the ad is more targeted.
Thus, the attribution of last click / last interaction favors strategies that are directed towards conversion and disfavors channels that focus more on relationship and branding.
[EBOOK GRATUITO]: Marketing on Instagram
[EBOOK GRATUITO]: Google Adwords
First click / first interaction attribution
Understanding this attribution model will not be difficult, as it is the exact opposite of first click / first interaction attribution!
That is, this model gives 100% of credits for the channel where the first interaction took place.
Following our example, the organic traffic would take credit for the purchase of running shoes.
If the company had to answer the question “How many customers do we generate with our SEO strategy?”, With the attribution of first click / first interaction the answer would be only the customers whose first interaction with the company was through organic traffic.
As with the attribution of last click / last interaction, this type of model is easy to understand and easy to implement because it assigns all the credit to a single channel.
Most Marketing automation tools use this model as the main framework. You can see in these tools that they establish the original source of the contact as the first channel where the contact interacted with the company.
Therefore, first click / first interaction attribution is used in several tools and this makes data analysis very easy, since there’s no need to create new mechanisms to track contact interactions.
This attribution model has the same negative point as the last click / last interaction attribution.
In other words, the fact that these models ignore other channels that contributed to the conversion can generate a skewed view that some channels are more important than others, when in fact they can be channels that naturally contribute more at the time of conversion.
Remember when you were at school, and in some competitions did all students win participation medals?
This is the linear model.
It assumes that all channels contribute the same amount to lead to conversion, and therefore should receive exactly the same amount of credit.
In our example of running shoes, credit would be distributed equally among the 3 channels: organic traffic would earn 25% of the credit, the Facebook ad would earn 25% of the credit, direct traffic would earn another 25% and the Email Marketing ditto.
While assignments last interaction it’s from first interaction ignore channels that were not responsible for the first or last interaction, linear attribution gives due credit to all channels that participated in the conversion, and that can be a positive point.
This model eliminates the possibility of not giving due credit to any channel.
Assigning credit to all channels can eliminate some recency bias, but on the other hand, with the linear model it is not possible to determine which channel generates the most value.
If we had the challenge of choosing a channel to prioritize and a channel to reduce costs, and we used the linear attribution model, the decision would be more difficult. A slightly more complex historical analysis would be necessary to determine which channel received the most credit accumulated over a period of time.
Allocation of time devaluation
This attribution model prioritizes the recency. That is, he assumes that all channels have an important role in the final conversion, but that this importance is relative to how long before the conversion the contact on the channel took place.
Simply put, an interaction that takes place a week before conversion is less relevant than an interaction that occurs 1 day before conversion.
So what would happen in the case of our example of buying running shoes?
Here is the timeline of our purchase:
Day 1: Organic traffic
Day 7: Facebook ad
Day 14: Direct traffic
Day 14: Email Marketing
Day 14: Purchase
Therefore, the Facebook announcement happened 7 days before the conversion, and the organic search visit happened 14 days before the conversion;
To analyze this model, it is always necessary to define a half life, that is, a period of time that will serve as a comparison to make the analysis. In this case, our half-life will be 7 days.
This means that the Facebook ad, where you interacted 7 days before the purchase, will receive half the credit received by Email Marketing for direct traffic, while organic traffic will have ¼ the value of the channels that participated on the same day of the conversion (because it impacted 14 days before conversion).
Among the options presented so far, it is one of the most realistic attribution models. In addition to giving credit to all channels, taking into account their participation in the conversion, it prioritizes the most recent interactions. Which makes sense!
The downside of attributing time devaluation is that it can be difficult to analyze, depending on the tools you use to track your key conversions.
In Google Analytics it is possible to determine a time allocation model to analyze the goals defined in Analytics. If you’re tracking conversions using Google Analytics, great!
If you use another tool, such as marketing automation software, that does not serve the same purpose, attributing time devaluation may not be the best option for you.
Now that we have gone through attribution models that take into account the order of interactions, the contribution of channel channel and the influence of time, we come to the last attribution model whose biggest factor is the channel order until you get to the conversion.
If we had to materialize the attribution models, the placement attribution would be a sandwich.
As with linear attribution and time devaluation, its premise is that all channels contribute to the conversion. However, placement attribution assumes that the first and the last channel are the most important in the purchase journey.
The logic behind this model is that the first channel is very important because it contributed to attracting the customer to the company, while the last channel of interaction contributed to the key conversion. And the interactions that took place between them also deserve merit because they were part of the client’s journey, but they don’t deserve so much credit.
Like the attribution of time devaluation, the attribution of positioning is one of the models that makes the most logical sense.
In addition to considering the importance of all channels in the conversion, he is able to highlight the channels that had the most impact on the journey (as opposed to linear attribution, which gives the same credit to everyone). In this way, it is possible to determine which channels have the greatest impact on the customer journey.
Since nothing in life comes easy, placement attribution is one of the most difficult to implement outside of Google Analytics.
If you use Google Analytics to track conversions, you can consider using this model as it will not be difficult to implement. Otherwise, you will need to study other ways to analyze your channel’s performance using this model.
[EBOOK GRATUITO]: How to optimize your conversions with Facebook Ads
[EBOOK GRATUITO]: Social Media Marketing Kit
[EBOOK GRATUITO]: Websites Are Us’s Ultimate SEO Guide
[KIT GRATUITO]: Master Google
Now it’s up to you!
If you’ve had time to answer questions like “Which channel do more customers come from?” or “Which of our channels generates the most value?”, may be the right time to determine a multichannel attribution model for your marketing strategy.
I will repeat:
Determine ONE multichannel attribution model for your marketing strategy.
That’s it. The most important point here is to choose only one model that will guide the analysis of all your Marketing strategies. After all, you are doing this to standardize the analysis of results (and this will be an impossible task if your team uses several models at the same time).
No attribution model is perfect. It is essential to remember this at all times. Everyone will have negative points, and it will be necessary to live with that.
So think hard before determining your attribution model and talk to other people who make Marketing decisions so that together you can weigh the pros and cons of each model.
I hope that after reading this you will be able to make more informed decisions and be able to understand more about how your Marketing strategies impact sales results!
Do you want to continue maturing your Digital Marketing strategy? Check out our free Ebook with Leads reactivation campaign templetes!