An effective marketing strategy usually spreads across multiple channels to reach a maximal audience and have an optimal impact. You compose engaging blog posts, invest in social media and Google ads, build email campaigns, and stretch resources to make yourself present everywhere you might be noticed by potential customers.
Not every platform, however, works equally well for every business or for different campaigns within the same company.
Depending on the goals of your strategy, you need to know how to distribute resources to achieve the best results and the highest return on investment or ROI. Most marketing channels have inbuilt analytics allowing you to monitor performance within the local campaign.
But how do you know which channel provides the most satisfying results for different aspects of your campaign? Which ones work best for brand awareness, lead generation, and, hopefully, conversions?
To obtain a clearer picture of your customer’s journey from top to bottom of the funnel, you need to compare channel performance. This way you will understand the role each one plays in the journey and will be able to determine its ROI. And there is nothing more important in marketing than knowing where to invest resources in order to make profits.
The process of comparing channel performance, based on the customer touchpoints, correlations, and results, is called Marketing Attribution Modeling.
It might sound a bit abstract at first but bear with us, by the end of the article, you’ll feel like a pro.
What Is Marketing Attribution Modeling?
In short, marketing attribution modeling is used to determine which are the main and which are the collateral channels that play a role in a customer’s conversion. Different models prioritize different touchpoints of interaction between your brand and your customer, and distribute ROI accordingly.
Building and comparing attribution models allows you to monitor which elements of a marketing strategy are actually effective and lead to conversions. You can estimate which of them help to deliver results for brand awareness, which are best for generating leads, and which to credit for the final sale.
This happens by pinning each touchpoint between your brand and the customer throughout the journey and estimating its value in ROI, based on its position in the chain of events.
In order to receive genuine measurable results, however, you need to include in your model all the channels used in your campaign. Otherwise, the information will be incomplete, the ROI will not be credited fairly, and your time will be wasted.
How Will Marketing Attribution Modeling Work Without Cookies?
Customer touchpoints are determined through different ways. It could be cookies, third-party data, sitewide tagging, log-in information or all the breadcrumbs a user leaves behind when browsing the internet.
With Google eliminating the cookie soon, there is a wave of uncertainty about exactly how the process will evolve and how marketers will be able to follow the customer journey.
Fortunately, we will not be left flying blind. Google has been testing an alternative to cookies, called Federated Learning of Cohorts (FLoC). The technology relies on machine learning and groups people with similar interests in cohorts. It will deliver reports with aggregated data that will provide accurate information about website visitors and conversions, but at the same time it will ensure the individual user’s privacy.
So, no worries, attribution modeling will still be possible in the post-cookie world and with the development of the FLoC technology, might even become more accurate.
Benefits of Attribution Modeling
When you are running an omnichannel strategy it is not always easy to determine which stream is responsible for the quality of your customer experience and which part of your efforts really convinced them that you are their best choice.
Attribution modeling allows you to break down every step of the journey to a comprehensive path of touchpoints leading to the sale. Analyzing different models can help you decipher the chain of events that drives prospects to conversions, eliminate the weak links and enforce the strong ones.
This enables you to clearly understand the ROI of every channel and measure the effectiveness of a strategy. Once you focus efforts on platforms that provide solid results for set goals, optimizing them should deliver even stronger performance.
Another great asset of attribution modeling is learning how different channels correlate in your strategy and how to strengthen those connections with quality content.
Attribution Tools and Software
Attribution modeling can be done manually or via software. The manual process is very complicated and requires advanced marketing knowledge and skills, usually familiar only to trained analysts. Here we will focus on the digital options that are easily accessible to businesses.
There are many free and paid tools to help you with attribution modeling. Google Attribution is the most popular free option available, so you might consider trying it out first. It will give you an insight into how attribution modeling works and what the different touchpoint models mean.
The benefits of Google Attribution often are overlooked. Even though it is free, it does cover all the general models and offers manual settings. Once you’ve familiarized yourself with the processes behind attribution modeling, you can build personalized models to serve your campaign needs.
Some of the paid tools, such as Bizible, TrackMaven, and Nielsen, have more advanced features, and if you are willing to make the investment and learn more, you can go wild with building sophisticated models.
But before you dive too deep, let’s focus on the basics.
Types of Attribution Models
Wouldn’t it be nice if people just landed on your site and started giving away their personal data and buying stuff without a second thought? Well, unfortunately, this isn’t the case. Most times, a customer’s journey consists of, at least, a few steps, before a purchase is made.
With attribution modeling, you break down and analyze the steps of a customer’s journey to find out which channels take the credit for the conversion. Every model prioritizes different touchpoints and attributes a varying value to their importance for the sale.
Let’s analyze the following example.
Someone, let’s call her Annie, finds a blog post of yours through organic search, reads it, and leaves your website. A few days later, she stumbles upon your Facebook ad and is reminded of your cool brand. She clicks the ad, browses your website again, becomes more interested. Annie is still not ready to buy but she fills in the opt-in form to receive updates from you, hoping discounts are coming her way soon. In a few days, you send Annie an email with a free shipping offer or a first-buy discount voucher. She’s excited, clicks the link with the intention to buy but gets distracted, and abandons her shopping cart. The next day she remembers about your brand, types in your website URL (or maybe clicks a link she saved), and finishes the purchase.
So which one of these channels gets the credit for Annie’s conversion? Your blog, the Facebook Ad, your awesome email, or direct traffic?
It depends on the attribution model you decide to rely on. Let’s have a look.
1. Last Interaction Attribution
As you can guess by the name, the Last Interaction model attributes all the value of a sale to a customer’s last touchpoint with your business.
This means in Annie’s case the credit goes to direct traffic. You can see why this model might be considered inaccurate.
In our example above, the direct traffic had little input in the customer’s journey. Annie discovered your brand through organic search and your blog. If it wasn’t for this original contact, she wouldn’t have been aware of your brand and your website URL, and the sale wouldn’t have taken place.
The last interaction, however, can be very useful in some cases. If you have a short customer life cycle, without complex touchpoints in between, this model will help you determine which of your channels is most persuasive sales-wise.
In Annie’s case, this would mean that your brand website is easy to remember and that your brand awareness channels are also delivering results.
As for saved links, you’ll find it easier to attribute value to them if you tag your URLs to follow channel performance in campaigns.
2. Last Non-Direct Click
The Last Non-Direct Click model is a variation of the Last Interaction model, designed to get around the exact issue we had with Annie’s case above.
Here the credit for the sale goes, as you can tell by the name, to the last non-direct click. In our example, this is your awesome discount email.
It is obvious in such a scenario that the email played a very important part in the sale and ignoring its value simply isn’t fair to the efforts you put in designing and targeting it.
Going with the last attribution model, in a chain of events like this one, might have led you to believe that sending targeted emails is not an important part of your strategy. And you might have misinterpreted the significance of this channel in your strategy’s ROI.
3. First Interaction Attribution
Although the First Interaction model might seem of little significance compared to the channel that really nailed the sale, it is not to be underestimated. Especially if your campaign, amongst other goals, also aims to raise brand awareness.
In Annie’s case, the organic search and your blog content were what introduced her to your brand and hooked her in. So this model will attribute all the credit for the sale to this first touchpoint.
The cons of such an approach are that it doesn’t take into account the time Annie spent on your website. She could have just bounced when she initially opened it and maybe she would never have looked back if it hadn’t been for the Facebook ad she later saw and the email you sent her.
But, as we already stated, this model is still a great choice if you want to be certain which marketing platforms deliver the best results for putting your name out there and introducing your brand to a broader audience.
4. Linear Attribution
A linear approach distributes the sale’s value equally between all channels involved in the customer’s journey.
In our example, this means that your blog, the Facebook Ad, your awesome email, and direct traffic will get the same amount of credit for Annie’s purchase.
The cons of the linear model are that it puts all channels on a par and fails to distinguish which one had the biggest input for the sale.
What it does, however, is show a clear picture of the whole customer journey and the importance of having an omnichannel marketing strategy.
It also serves as a reminder that every channel has its importance and impact. Skipping just one might result in breaking the chain of events that led to the sale and prevent it from ever happening.
5. Time Decay Attribution
The Time Decay model is very similar to the Linear one because it distributes the purchase value amongst all channels that had any part in the customer journey.
The difference here is that the closer the touchpoint is to the sale, the more credit it gets.
In Annie’s shopping journey, the ROI of the marketing channel she interacted with will gradually increase with the organic search getting the lowest percentage, and the direct traffic getting the highest.
The con of such an approach is that it doesn’t take into account how long the customer interacted with the website at any point of the journey, or how great the impact of first impressions was. Annie might have made up her mind to do business with you the first time she came across your brand, but this model would have totally ignored this.
The benefits here lie in determining which channels drive actual conversions and sales. The model can be a good choice with longer customer life cycles. Like, for example, with B2B interactions, or with expensive purchases that require more consideration and research such as in the automotive industry.
6. Position-Based Attribution, U-Shaped
The Position-Based Attribution Model, also known as U-shaped Attribution, gives an equal 40% credit to the first and last interaction of a client with your business and divides the remaining 20% amongst the other channels in between.
The first impression you give a prospect and the last one that makes the sale happen, are considered the most important in this model. All the while, the other steps in the process are not deemed completely unimportant.
This makes the U-shaped model a very common choice for marketers and a reasonable one indeed. It gives you a perspective of the strongest points of your strategy and is suitable for businesses with longer customer life cycles and multiple interactions along the way.
7. Position-Based Attribution, W-Shaped
The W-shaped Attribution model is a variation of the Position-Based one but gives extra credit to the touchpoint that converted the prospect into a lead.
In our example above, organic search and direct traffic will get equal credit again. But this time so will the awesome discount email you sent, which convinced Annie to share her details with you and made her a lead. The other touchpoints of her journey will split the rest.
This model attributes value to the most important touchpoints a client has with your brand. This makes it an even better option if you want to also recognize which are the channels generating the most leads for your business.
8. Lead Conversion Touch Attribution
The Lead Conversion Touch Attribution Model gives credit only to the touchpoint in the chain of events where the conversion happened.
In Annie’s case, this would be the awesome email she received from you.
As with the first two channels that focus only on one touchpoint of the journey, this one too has value for very specific campaigns mostly.
If the main purpose of your campaign is to generate more leads, this is a model that will give you a clear view of the channels that convert the most and will allow you to focus efforts there.
9. Custom Attribution
Custom Attribution Models, as the name indicates, let you build your own models depending on your specific marketing strategy and business priorities.
Building custom models requires a deep understanding of marketing attribution modeling and advanced knowledge of the tools you would be using. Which is why we advise you first familiarize yourself with the process before you jump into the depths.
In a custom model, you can incorporate data from online and offline strategies and gain a deeper understanding of your customer’s journey.
Nowadays few people are influenced solely by digital channels and even fewer rely only on offline sources. Managing to segment your data and build models based on your full marketing strategy will give you a bird’s eye view of the whole process and insight on which are the strongest touchpoints with your customers.
Attribution Modeling is one of the most complicated parts of marketing. It requires time, analytical skills, in-depth knowledge of the marketing processes, and insight into a business’s profile and goals.
Luckily, modern technology lends a hand with the tough parts such as calculations and pinning down customer touchpoints, and gives us a head start.
Building attribution models for your campaigns will help you identify the ones that deliver solid ROI and optimize your marketing team’s performance.