Marketing attribution has been the buzzword of the digital marketing world in recent years, especially with the advent of various marketing attribution tools recently.
While those tools can certainly help your marketing activities, it won’t be much use if we don’t understand the concept of marketing attribution itself, and how to correctly implement it.
In this guide, we will discuss all the ins and outs of marketing attribution, the various available marketing attribution models, and some key things you should understand before implementing an attribution model.
Let us begin by discussing the definition of marketing attribution itself.
What Actually Is Marketing Attribution
Marketing attribution, in essence, is the process of defining a set of different user actions. These actions are often called “touchpoints” or “events”, and they contribute to a certain degree in producing a desired outcome. By assigning values to each of these actions, we can measure their “attributes”, hence the name.
Sounds confusing? Well, let’s try a more practical approach. The end goal or desired outcome of any marketing tactics and initiatives is conversion, where the website visitor turns prospect turns lead finally makes a purchase. So, marketing attribution is the science to determine which sets of actions (touch points) the customer interacted with before finally making that purchase. These actions will involve all interactions with your brand, from engaging your promotional email to clicking a banner advertising.
In practice, we put the theory of marketing attribution into use by creating a model, which we deem the “marketing attribution model”. With this model, we can determine how to assign campaign influence.
The Key Benefits Of Marketing Attribution
Based on the definition discussed above, the main purpose of the marketing attribution is to uncover everything about your marketing efforts, and use this clearer picture to develop an even better marketing strategy.
According to a research report released by iProspect (PDF), implementing a marketing attribution model can generally result in an increase of 15 to 35% in media efficiency and overall ROI. Surely that’s a significant number. Yet, the benefit doesn’t stay there. Here are some other benefits of implementing marketing attribution:
- More Accurate Revenue Decisions
Marketing attribution enables a clearer view o the buyer’s journey, allowing us to make more accurate decisions about each channel.
- Better View on The Sales Process
Attribution gives us a better look inside the marketing campaign to determine the ROI of each channel.
- Find and Solve Problems
With a proper marketing attribution model, marketers can have more insight to solve problems, both for the company and the customers.
- Better Understanding of The Audience
Marketing attribution reveals more data on actions/touch points with the highest engagement. As a result, we can have a clearer understanding of the buyer persona and can implement better targeting.
- Understand Customer’s Lifetime Value (LTV)
Attribution is especially useful in showing the influence of each action/touchpoint to the customer’s lifetime value. With this understanding, we can plan better marketing spending whether we should acquire a new customer or increase customer retention.
- Develop Better Marketing Strategy
We have briefly discussed this above, but the biggest advantage of implementing attribution is how you can get a detailed look on each individual aspect of your marketing strategy. As a result, you can develop a better marketing strategy that performs on every level.
Different Marketing Attribution Models
There can be many different variations of marketing attribution models to implement. Yet, we can generally divide them into seven different types. Here, we will discuss those types and the advantages of each type.
1. Last-Touch Attribution Model
This attribution model focuses on the end of the buyer’s journey, especially the actions/touch points that drove the customers in making a purchase.
Generally, this model is better-suited for marketers who are focused mostly on driving conversions and don’t really value non-converting factors.
Pros and Cons of Last-Touch Model
Pros: Very easy to set up, as most marketing automation and CRM platforms offer the feature to set up this model.
Cons: There are various factors that can be really important scattered throughout your marketing efforts. This model only focuses on the last interactions leading to the conversion, and might ignore important influences in the process.
2. First-Touch Attribution Model
The opposite of the last-touch attribution model, the first-touch model focuses on user actions that drove a visitor to the website the very first time, regardless of whether that visitor will convert to a lead and/or customer.
So, this model focuses on the beginning of the marketing funnel, and is generally used by marketers who are focused on lead and demand generations.
Pros and Cons of First-Touch Model
Pros: This model is also relatively easy to implement. Google Analytics can easily build this model.
Cons: This model doesn’t give much room for optimizations, since it doesn’t really reveal the factors that drove a customer for conversion. As with the last-touch model, the first-touch model suffers from the over-emphasis issue: putting too much focus on one end over the other.
3. Lead Conversion Touch Attribution Model
The lead-conversion touch attribution model, as the name suggests, focuses on the user actions where lead was generated.
Pros and Cons of Lead Conversion Touch Model
Pros: You can get a clearer picture on which actions/touchpoints generate leads.
Cons: The lead generation process is a relatively small part of the overall marketing process and the buyer’s journey. This model misses out on the other details.
3. Time Decay Attribution Model
The time-decay model gives the most credit on user actions that is responsible for conversion. The further away a user action/touchpoint is from purchase, the less important it is for this model, hence the name “time-decay”.
Pros and Cons of Time Decay Model
Pros: With this model, marketers can easily optimize user actions that drive conversions and those that assisted in driving conversions.
Cons: With the time-decay model, the user actions that produce lead and drove customers to the business in the first place will receive less credit. With this model, there’s a risk that we might ignore important touch points that happened early in the marketing funnel.
4. Linear Attribution Model
The linear attribution model assigns credits linearly/evenly to all touch points of the marketing funnel. For example, if there are 20 steps in the buyer’s journey, each user action will receive 5% of the credit.
Pros and Cons of Linear Model
Pros: Marketers can optimize the entire marketing funnel, not just one or parts of it.
Cons: Not all user actions/touch points are created equal, and the fact that this model assigns the same value for all can create an issue: we might give too low for high-performing user action, and vice versa. It can be difficult if we only want to optimize high-performance actions.
5. Position-Based Attribution Model
Sometimes called the U-shaped model, this attribution model gives 40% of the total value to the very first and very last user actions (80% in total). The remaining 20% is divided evenly to the rest of the user actions/touch points)
Pros and Cons of Position-Based Model
Pros: Every single user action in the marketing is valued, while the first and last touchpoints are optimized. With this model, marketers can optimize the user action that introduced customers to the brand in the first place, while we can also optimize actions that drove conversions.
Cons: Since this model assigns fixed 40% values to the first and last user actions, that number might be inaccurate and might affect the optimization.
7. Algorithmic Attribution Model
Sometimes called custom attribution model, and here the attribution model is custom-tailored by a data scientist to meet the specific needs of the client, to accurately match the buyer’s journey of the business.
Pros and Cons of The Algorithmic Model
Pros: Since this model is custom-built, this attribution model offers the most accurate representation of the contribution of each user action on the marketing funnel.
Cons: This model can be very difficult and time-consuming to build.
In general, a custom-built algorithm is the best choice for every business. Yet, it can be very expensive, and not everyone has the luxury of time to implement it.
Key Things Every Marketer Should Know About Attribution
Here, we will discuss several things every marketer should know about attribution. These things can make a significant difference in how you approach marketing budget, enabling you to optimize ROI and achieve better results.
1. Communication Is The Main Factor In Implementing Attribution Model
The modern marketing teams often suffers from the issue of separation, where different sub-teams are working in different channel silos. It is worth noting that implementing marketing attribution will remove these silos altogether. Yet, the requirement of implementing attribution is for all the different parties to agree on the same goals, specifically the goals of the attribution program, including the key metrics and the specific attribution model.
The goal of marketing attribution is to optimize marketing spending, with the goal to produce more revenue by a more effective spending. However, without a clear communication within your team, some sub-teams might far that they will lose their usual budget -and status-, and won’t have the same vision with the attribution project.
2. Marketing Attribution Has A Low Entry Barrier
The marketing attribution might seem like difficult concept at first glance. Yet, if you have understood the ways to analyze specific marketing channels like Google Analytics, Facebook, Twitter, and other common ones, actually you’ve already got the necessary skills to implement basic attribution.
Especially if you are already using Google Analytics to track various data within all the different channels, you are, in fact, already doing attribution practice. You can start with analyzing how your audience engage with your site. Specifically, from where did they enter your site, where they bounce, and how long does it take until the conversion (if any).
If you are a B2B business, you can tie that information into your CRM platform or other database, and then you can analyze the impact against revenue.
Using just the data in your CRM and Google Analytics, you can understand how each channel falls in the sales cycle. For example, do clicks from Google search directly lead to conversion, or is it more effective as a lead generation device?
With this simple data, you can set up basic attribution models like first-touch and last-touch models.
3. Attribution Can Be Done With Free
Above, we have mentioned the benefits of implementing marketing attribution. With all those benefits, you may suspect that implementing marketing attribution will be an expensive investment, and it’s true. Advanced, custom-tailored marketing attribution models can be extremely expensive and can take a really long time to build.
However, as we have discussed just above, there are free options where you can implement basic marketing attributions with relative ease. With the data you’ve collected from those free sources, you can slowly build a more advanced attribution model over time.
4. Marketing Attribution is Not Solely About The Marketing Team
Marketing attribution is not an issue exclusive for the marketing team. Implementing complex attribution models will require the cooperation of the tech team, while the finance team should be involved from the very beginning to align the goals.
This is due to the fact that marketing attribution, when implemented correctly, should affect budgeting. For example, over-crediting the bottom aspects of the marketing funnel may cause the unnecessarily tight budget for the upstream channels.
Proper implementation of the marketing attribution can show the real value of each channel, and the finance team can be persuaded to assign the proper budget for each. So, involve every team when implementing marketing attribution.
5. Consider Customer Lifetime Value
Arguably, understanding the lifetime value (LTV) of customers is the most important marketing metric for every organization.
If we are only focusing on channels that drove conversions, we might neglect the channels further above in the marketing funnel, and vice versa. By considering LTV, we can measure the end goal (conversion) over time as a process, rather than just one single purchase.
This is especially true in the B2B environment and e-commerce businesses, where improving customer retention is often more valuable than acquiring new customers.
6. Don’t Chase The Perfect Attribution Model
Let’s face it: we will never achieve such thing as a perfect marketing attribution model. It is very difficult to translate every single thing in your marketing activities into accurate values. No matter how advanced a marketing attribution model is, there will always be flaws.
So, part of the secret of implementing proper marketing attribution is acknowledging when good is good enough.
7. Test Everything
As mentioned above, every marketing attribution model has its flaws, so we should test the model with care, just like everything else in digital marketing.
Remember that the goal of marketing attribution is to correctly interpret each value of each marketing action/touchpoint so that you can spend your budget more effectively. Yet, this doesn’t directly increase the value of those actions/touchpoints, and we still need to do further testing to optimize those that lacking.
While the term “marketing attribution” might seem overwhelming at first, the idea behind it is actually quite simple: assigning value to every aspect of your marketing activities. Hopefully, after this guide, you’ve got a clearer picture about the concept of marketing attribution and how to implement it.
Just as everything else in digital marketing, or marketing in general, marketing attribution is constantly evolving virtually every other week. However, the things we have discussed above can give you a good basis on the attribution concept and its implementation.