The Beginner's Guide to Marketing Attribution
As a former writing major, my mind would drift to attributing authors or other content creators for their work within a bibliography or perhaps a brief annotation when the word attribution came to mind. Today, as a digital marketer, attribution can take on many different forms of the same meaning. At the base of the meaning though, attribution in marketing allows marketers to attribute traffic sources to their website or place of business with specific marketing strategies that got them there in the first place. Attribution is all about understanding the customer journey, measuring how individuals learn about your organization and making strategic decisions about how to attract more leads or customers.
There are a plethora of strategies that marketers attempt to bring traffic to their sites from pay-per-click and search engine optimization to affiliate marketing and TV advertising campaigns, and it’s critical for your marketing success to understand how they interact with your marketing website and what brings in the best results because ultimately that’s where you will end up spending most of your marketing budget. In this blog post, we’re going to review the most common forms of marketing attribution and how they line up with the most common forms of marketing strategies.
Per usual, before we start, I must give Russell McAthy from Ringside Data a quick shout out as I gleaned information from his course at CXL Institute, and his course goes into even greater detail on the subject if you’re interested. Let’s get started!
Attribution Models for Dummies
Now before you start reading this entire blog, let’s establish the obvious question here: do you need to worry about marketing attribution? In my opinion, all businesses who are deploying marketing campaigns should have a general idea of attribution to understand what’s working vs. what’s not working, but my opinion doesn’t apply to all contexts. If a majority of your business comes from referrals, your website is barely used by customers and prospects alike, or you have an incredibly short buying cycle that’s not an e-commerce site, perhaps you don’t have to prioritize attribution from the start. However, if any of those businesses begin to scale, then marketing attribution will quickly become important as the marketing budget increases. For those of you on the fence, here are a couple questions to ask yourself:
Do I want to know what my customer’s lifetime value is for my business?
Am I spending an exorbitant amount of money on marketing with little insight into ROI?
Do I have a buying cycle with numerous touchpoints that might influence the buyer?
Am I using multiple marketing channels because it sounds like the right thing to do?
If you answered yes to any of those questions, then it might be time to think about attribution models. What’s interesting about attribution models is that it’s commonly used by a lot of larger companies with expansive marketing teams, which is a great practice. The problem though is that most teams don’t align on a proper attribution model to use across the team. Instead, teams will opt to utilize the attribution that highlights their team the most without directly stating the attribution model being used. Additionally, while the default attribution lookback time is 30 days, there’s typically not a consistent lookback time for attribution either. A 30-day window to attribute credit for a sale might not be the most accurate if the buying cycle is long. Let’s take the last-click attribution model for example. Last-click gives 100% of the credit to the last interaction made with your site and basically erases the other strategies that might have contributed to the sale.
Let’s look at this from the perspective of an auto-buying site such as Carvana where you can basically order a car online. Most people – unless you’re a millionaire – don’t wake up out of the blue and say, “I’m going to buy a car today.” Buying a vehicle is typically a several week to several month long process of narrowing down your vehicle hunt to a few choice cars and haggling with the seller for the best price. During that researching phase, car dealerships often send retargeting ads, email campaigns and TV advertisements your way with the hopes of convincing you, the consumer, to buy from them. In Carvana’s case, that convincing stage might not even happen as customers simply click a button to have the car delivered to their front door. Imagine the following customer journey to Carvana’s website:
Ad click on a PPC campaign
Direct/none to website
Ad click on a PPC campaign
Entry from an email campaign
Google search organic
If we’re using a last click attribution model, then the last click on an organic search engine result page would be the resulting attribution source. That doesn’t seem right does it? It’s because it’s not. You probably spent $50 to $100 on your PPC campaign and lost sweat equity on formulating your email campaign. However, the wrong attribution model discredits that work forces you to think organic (SEO) is the way to go when in reality you might want to focus on PPC. Choosing your attribution model can be the key to success for your organization, so choose wisely.
First click attribution
First click attribution is exactly what it sounds like. The first-click that led to a user entering the site will become the accredited source. For sites where a conversion is likely to occur quickly without further thought, first-click attribution might be a safe method.
Last non-direct click attribution
Last non-direct click attribution is the default attribution model in Google Analytics. If the Analytics user has not actively changed the attribution model, then this is what is being used. Last non-direct click attribute 100% of the credit to the last click that was not a direct entry onto the website (they did not type in your URL and press enter). Unfortunately, as stated with last click attribution, this often skips a large part of the buying cycle.
Last interaction attribution
Last interaction attribution attributes 100% of the credit to the last interaction to get to the site. The difference between last non-direct and last interaction is that direct traffic can gain attribution.
Linear attribution attributes each touchpoint with an equal percentage of credit. If you’ve ever looked at the multi-channel funnel within Google Analytics, this is the model that is used. It gives you an equal view of each channel to see what is working vs. what is not. This only works if you believe each touchpoint holds an equal amount of value.
Decay model attribution
Decay model attribution attributes most of the credit to the last interaction, decaying credit from the first interaction and more interactions occur. This is a good model to understand what the final decision-maker channel might be for your prospects.
Position-based attribution (Bathtub model)
Position-based attribution is often how most people logically picture attribution. In this model, the most credit is given to the first and last interactions. A good way of thinking about this, as stated in Google Analytics attribution model page, is 40% is given to both the first and last interactions while 20% is given to all of the interactions in the middle.
Understanding the best model for your organization takes some time and becomes increasingly important from a reporting standpoint to show marketing results. However, if you can’t figure out what attribution means for each marketing strategy, then your model isn’t really going to help you out. For SEO, what keywords get your customer to the website? For PPC, where are your ads taking the buyer in the buyer journey? How in the world do you attribute campaigns that aren’t website based like TV or direct mail? Let’s talk about it
Search engine optimization is all about prepping your website to perform well when individuals search for it on Google or other popular search engine platforms. We should think of SEO as three tiers: technical, on-site and off-site. Keywords (which can be optimized via all three tiers) come into the focus here as they’re the method people use to search for your site in the first place. It’s critical to understand what keywords push users to your site and at what stage in the journey they are when they arrive at your site. For example, brand queries that include the name of your company are often people who are either just beginning their search inquiry about your product or actual users of your product. Long-tail keywords (queries that are often four or more words) can often indicate that people are in the middle of their customer journey with your brand. Optimizing your website with tailored content for individuals based on their place in the customer journey often helps get what they want they want to see, further convincing them on the way to a purchase.
Pay-per-click campaigns typically use the last-click or last non-direct click attribution model, but attribution can vary here based on the length of the customer journey. For shorter buyer journeys (i.e. 1-3 days), PPC can be an ROI goldmine as it can convince users to buy quickly and easily through search ads on SERPs and retargeting ads via display advertising. However, the longer the buying cycle, the less likely customers are willing to convert because the decision making process might be a long one. This is why we see PPC campaigns that include a quick short landing page asking users to sign up or gain access to a piece of content for quick conversion as well as campaigns that lead users to a purchase page now. Figuring out your attribution model for PPC relies almost entirely on the behavior of your customers.
A good question to ask here is are there macro conversions or micro conversions occurring? A macro conversion can be defined as the end all be all action you would like a user to take on your site. A few examples of this could be purchasing an item on an ecommerce website or filling out a demo request for SaaS companies. A micro conversion categorizes all the mini actions taken on your site before the purchase that you would like users to take as well. Micro Conversions aren’t the final goal you’d like taken on your website, but you understand that they must be taken to get to a macro conversion. A few good examples of this would be webinar signups, content downloads or video views. Your PPC campaigns can lead to either conversion type, but it’s important to attribute correctly based upon importance to your business.
Affiliates & Email campaigns
In the case of affiliate marketing, content marketing is the most often form utilized by companies. Since affiliate marketing can get expensive quickly, it’s important to understand how that traffic performs on your site once they arrive. This is where UTM codes come into play. Having a consistent and easy to use UTM code system for your team can either credit or discredit your affiliate marketing plan quickly. This allows you to choose and strategize in the future to see which affiliates should be partnered with vs. those who might not provide the best results.
UTM codes are important for email campaigns as well. By default (unless UTM codes are used or your email marketing platform automatically categorizes source/medium), Google points all email traffic to the direct/none traffic source. That’s no good! UTM codes will make or break your email campaigns and allow you to pinpoint which campaigns exceed expectations and which campaigns you should go back to drawing board on.
TV & Direct mail
In order to attribute TV and direct mail, Google Analytics users have to get a bit strategic. There is no way marketers can know how many mail recipients saw their offer or how many TV watchers understood and watched their advertisement. However, if a special landing page with an offer is only used for each channel, it’s easy to monitor a URL for a certain amount of time. If your organization knows that there will be offer fliers sent to homes over the next week, it’s easy to look at the URL being sent on the flier for overall performance. The same goes for TV. If you know your advertisement shows five times on Tuesday from 1pm to 8pm, it’s easy to check for traffic updates and conversion on Google Analytics.
In the meantime, what attribution models are you using? How are you tracking traffic to your website? We’d love to hear how you’re using attribution models for reporting and strategy.
This post was created in an effort to complete my CXL Institute Mindegree Scholarship obligation and speak to the materials reviewed in the course. The information is a combination of my previous knowledge and excellent insights from a phenomenal program.