If you’re like most marketers, you’re likely running multiple campaigns across an array of different marketing platforms, putting time, energy, and plenty of ad spend into as many baskets as possible. You start seeing results— more traffic is coming to the site, and there are more conversions happening.
But how do you know exactly where that traffic is coming from and which campaigns or traffic sources successfully drive the high-value traffic leading to sales?
That’s where site attribution comes in. Site attribution allows you to gain insight into where your traffic is coming from and which platforms are sending users that are converting, opening the door for optimization potential moving forward.
Site attribution can be a tricky subject, but with some basic knowledge, the right tools, and an understanding of different attribution models, you’ll be able to interpret the data (and act on it!) in no time.
Ready to learn more about your site traffic and conversions? Let’s dive into this guide to site attribution.
Site attribution is the process of defining where your site traffic is coming from so you can determine which sources and channels are bringing in the most traffic. Most site attribution tools also have audience tracking features so you can track user behavior from different audience segments
Site attribution is essential to helping you understand which marketing campaigns and platforms are most impacting your specific business goals, including brand awareness, sales, and other actions like app-downloads. Only with this information can you optimize your campaigns moving forward.
This information is invaluable, and can keep your business on track.
You may realize that:
It’s not just about where the traffic is coming from, but what users are doing once they come to your site (and when they come back).
You can also better understand your brand’s specific customer journey, and how they’re moving through your site. In this process, you may discover “sticking points” that present obstacles to customers you can resolve.
You may, for example, notice a high bounce rate on a landing page, and after testing realize that there’s a glitch on that page for certain devices that prevents the lead form from loading correctly.
Site attribution is possible thanks to web analytics tools.
The most commonly-known web analytics tool is Google Analytics, which is free and easy to use, although much more limited than paid web analytics tools.
All web analytics tools utilize cookie-based tracking. You’re given a snippet of code by the tool of your choice, which you install into the code on every page of your website. This code will fire when users land on the page, and using the cookie-based tracking, they’re able to see where the traffic is coming from.
Most web analytics tools have different site attribution models that allow you to determine which touchpoints in the customer journey receive credit for value of traffic (or the value of sales).
Different attribution models have their own pros and cons.
The first-click attribution model, for example, helps you assess which touchpoint first introduced a buyer to your business.
The last-click attribution model helps you see what touchpoint users interacted with directly before taking a conversion action.
And other attribution models can help you to better understand the customer journey overall, and look for crucial touchpoints that seem to play a pivotal role in eventual conversion.
When it comes to attribution models, every web analytics tool may have its own unique terminology for the specific models in question. Many platforms, however, offer similar models (and may use the same terminology) as what’s available on Google Analytics.
Let’s take a look at the most commonly-used site attribution models.
First-Click attribution is a single-click model, meaning it only accounts for a single interaction. It grants 100% of a conversion’s value to the first-tracked click the user took on their journey (within the lookback window assigned).
The First-Click attribution model is exceptionally helpful when evaluating the effectiveness of customer acquisition channels. If you want to determine which channel is most effectively bringing you the largest quantity and the highest quality of traffic, this is the model to view.
Limits: While this model is great for showing customer acquisition channels, it doesn’t show anything else. And while acquisition and discovery is important, it’s far from the only thing that matters. This won’t tell you anything else about the customer journey— including what finally drove users to convert.
Example: You are the owner of an online cosmetics store. The client first learned about your store through advertising on Google Ads, and a couple of days later found the store in a search engine and made a purchase. The First-Click Attribution model attributes all the success of a purchase to that first contact through Google Ads.
The Last-Click attribution model is another single-click model, and it’s the exact opposite of its first-click counterpart.
Exactly as it sounds, the last-click attribution model gives 100% of attribution to the last touchpoint users interacted with before converting. Like with First-Click, nothing else in the entire customer journey is accounted for here. It’s useful when evaluating the effectiveness of different campaigns that trigger users to finally make a purchasing decision.
Limits: You can learn a lot about what touchpoints are finally driving users to convert, but the limitations are the same as with the first-touch model. This model does not account for the contribution of previous sources that may have influenced the purchase decision.
You’re getting an extremely limited look at the customer journey, and while that final touchpoint may have sealed the deal, there may have been several other essential touchpoints the user needed to interact with before being willing to convert.
Example: You are the owner of an online store of household appliances. The client first learned about your store through advertising on Gogole Ads, then visited your site through a search engine and made a purchase after receiving an email with a promotional code. The Last-Click Attribution model attributes all the success of a purchase to the last contact—an email with a promo code that convinced the user to finally make that purchase.
The Time Decay attribution model is our first multi-touchpoint model. With this model, attribution will be divided up amongst all touchpoints that occurred within the lookback window. The highest value is assigned to the traffic source that most recently occurred, with each touchpoint after receiving a decline in attribution.
This model is particularly useful for businesses with longer sales funnels, and it’s helpful because it can show you the exact path users took towards conversion. You may notice that certain mid-funnel actions pop up frequently, indicating it’s playing an important part in the consideration stage that pushes users closer to purchasing.
Limits: Time Decay does not take into account the degree of influence of different channels on decision making. Many experts argue that the first touchpoint should never be at the very bottom of the totem pole, because without a great first touchpoint users would have never converted, and it’s likely more valuable than some of the mid-funnel actions that may not have impacted conversion at all.
Example: You're running a real estate sales campaign where customers typically look through a lot of properties before making a purchase. At the beginning of the funnel, they can find your site through contextual advertising, then go to the site using the link from social networksher and, before buying, go to the site from the mailing list.
The Time Decay model takes into account all of these interactions, but gives more weight to later ones, such as the transition to the site from the mailing list that are more likely to lead to a purchase.
The linear attribution model is a multi-touchpoint model that distributes attribution evenly across all touchpoints that the user interacted with prior to conversion and within the lookback window. It’s helpful when you’re evaluating different channels’ contributions to both acquisition and retention.
Limits: While the Linear model does give you a full look at what’s happening in your customer journey, it does not account for the different degree of influence of different channels on decision making.
Example: You are the owner of an educational center that offers exam preparation courses. Clients usually study a variety of materials and attend various events before deciding to pay for a course.
In the process, they can go to the site via a link from social networks, read articles on the site, and watch webinars. The Linear Attribution model evenly distributes weight across all of these touch points, recognizing the equal importance of each step in the sales funnel.
Position-Based attribution (also known as “Positional attribution”) assigns credit for a conversion based on the position of different touchpoints.
The first click and last click receive the most value (divided equally between them), while all remaining sources divide up the remaining percentage between themselves.
This is a popular attribution model because it gives strong credit to the discovery channel and the final conversion touchpoint, both of which are essential in the customer journey, but it still awards some attribution to every touchpoint that played a part.
Limits: While this model does give plenty of credit to the first and last touches in a customer journey, it can underestimate sources that are between the first and last user interaction with your site.
Example: You are the owner of a travel agency. In your line of work, customers usually look for information about different offers before making a booking. At first, the client may see your offer on a travel blog,then go to the site using the link from the public in VK and finally go to the booking page. The Position-Based model takes into account all of these touchpoints, but gives more weight to the first and last interactions.
One important thing to note is that you don’t have to set your web analytics platform to a single attribution model and never change it.
It’s common for marketers to toggle between different models to gain new insights into the customer journey, such as to easily identify the most profitable discovery channels or which touchpoints show up most frequently in converting customer journeys.
That being said, many marketers prefer to have a single attribution model that they “leave on” for quick review. This is a little like their default settings.
So, let’s take a look at how to choose the right attribution model for your business.
What are you selling, and how long does it take (on average) for customers to convert? That’s an important first place to start.
If you’re selling $20 lipstick on Facebook Ads, sometimes a single-click attribution model is all you need. There’s not a long consideration process; it may be an impulse purchase in many cases.
If you’re selling SaaS software to enterprise-grade companies, however, you’ll want a more complex attribution model. You’ll also likely want a much longer lookback window, as large decisions like that are rarely made in the standard 30-day lookback window.
What model you’re using will depend on what questions you most need answered.
Are you trying to determine what touchpoints are most commonly interacted with? Or which are most influential, and appearing towards the end of the customer journey?
In many cases, marketers will choose to have their default as a multi-touch attribution model so they can get a basic idea of the overall sales funnel and customer path.
Some businesses may only use a few marketing channels. That $20 lipstick cosmetics company, for example, may be able to rely mostly on a combination of Facebook and email marketing.
Other brands, however, may use a wide number of marketing channels, including SEO, sponsored placements, affiliate marketing, social media marketing, and multiple types of pay-per-click ads.
If you have a diverse mix of marketing channels (and users that often interact with touchpoints across channels), opting for a multi-touch attribution model works well here, too.
Your goals— and the key performance indicators (KPIs) that you use to track them— will have an impact on which attribution models you focus on.
Apps that are currently implementing a go-to-market strategy, for example, are going to be looking a great deal at channels that are helping with acquisition. And businesses that are focused on refining their ideal customer profile (ICP) to sell to higher-value clients are likely going to look more at late-stage actions and touchpoints that may indicate higher conversion values.
If the main goal is to increase sales, then Last-Click may be the most suitable model. If your strategy is to increase brand awareness, then First-Click and Linear models can give a more accurate picture of the effectiveness of your marketing efforts.
One thing we want to note is that you may sometimes have attribution tracking set up across multiple platforms and you notice inconsistencies in the reporting.
This may happen for several reasons:
If your attribution is different across unique platforms, start by looking at your settings and searching for discrepancies from there.
When you’re ready to learn more about site attribution, we strongly recommend choosing a dedicated web analytics tool that offers the following features:
MyTracker offers all of the above and more. Take a look at how we can help you here.
Different site attribution models can help you better understand how users are interacting with every touchpoint in your digital sales funnel and how those touchpoints are impacting the customer journey. This is essential information that can help you not only invest more in channels that are successfully increasing acquisition with high-value customers, but to optimize the entire customer journey.
And remember that you don’t have to choose just a single attribution model permanently. You can view your data under multiple attribution models so you can get a better big-picture view of what’s actually happening, especially if you’re customizing your lookback windows to align with your standard customer journey.
