Use Attribution Modeling to Improve Your Conversion Rate
Data is supposed to be concrete. Numbers don’t lie, and the results you see in your analytics tools show exactly what is working and what isn’t. However, when it comes to attribution and measuring the value of your digital marketing efforts, there are shades of gray. Over the past few years, Google has released more attribution modeling tools within its platform. These allow marketers to understand where their efforts fall on the sales funnel and what is really driving sales.
If you’re not analyzing your attribution models, then you can’t fully understand how your marketing efforts affect customer behavior. This guide will explore attribution modeling and give you the tools to use it for your brand.
What Are Attribution Models?
We all like to believe that a customer encounters our brand, is wowed by our marketing efforts and immediately visits our website to convert. The whole process takes only a few minutes, and the results are clearly reported in Google Analytics. This isn’t actually the case.
Salesforce reports that it takes an average of six to eight touches to generate a viable sales lead. This statistic focuses on B2B companies and lead gen brands, but can also apply to B2C eCommerce companies as well. In fact, Kissmetrics reports that 98% of customers won’t convert the first time they visit your website.
Your customers will read a blog post and then bounce. They will read a social media update and scroll past it. Finally, they will click through an email or search directly and convert.
If you want to see this process in action, look at the Top Conversion Paths tab in Google Analytics. You can expand the timeline back to see the most common paths customers take before they buy, and filter by various marketing efforts. This shows just how convoluted the journey is from when your customers first visit your website to when they make a purchase.
Attribution modeling reviews the paths that customers take and assigns a value to them. For example, if a customer follows the path of paid search > social media > organic search before they buy, how can a marketer decide which channel gets credit for the sale? Paid search introduced them to the brand, but your SEO efforts led to the conversion. Depending on the attribution model you use, the answer varies.
What Are the Different Types of Attribution Models?
Marketers use attribution modeling for a variety of reasons. Some people want to see which marketing channels support their top-of-funnel efforts, while others want to improve their mid-funnel and bottom-funnel results. Depending on a brand’s goals, they might look at multiple attribution models and use them for different reasons.
There are five common attribution models you can use when reviewing your sales performance in Google Analytics.
First-Click
First-click attribution assigns 100% of the credit to the first point of contact a customer has with your brand. The number of touches they have before they convert, and the number of times they touch a specific channel doesn’t matter. All that matters is what was clicked first.
First-click attribution favors top-funnel marketing channels. For example, when you reach new audiences through social media promotions and drive them to your website, they likely aren’t ready to buy — especially if this is the first time they have seen your brand. However, this model credits that social click as the first domino that falls and leads the customers to eventually buy.
Last-Click
Last-click attribution assigns 100% of the credit to the final point of contact a customer has with your brand before they buy. Regardless of how the customer learned about your company or what motivated them to finally make the purchase, this channel gets the credit.
Last-click attribution favors bottom-funnel marketing efforts. After days, or even months, of research and engagement with your brand, the last piece is the straw that breaks the camel’s back and convinces them to make a purchase. Direct traffic and branded SEO efforts often receive higher credit in last-click attribution modeling.
Linear
Linear models follow the rule that all clicks and channels are created equal. If a customer has four touchpoints before they make a purchase, then each channel receives 25% of the credit for the sale. This method is preferred by brands with long purchase funnels that want to see how their mid-funnel efforts affect traffic and sales.
Time Decay
Time decay attribution is similar to last-click, in that the final touch before a purchase gets the majority of the credit. However, the last touch doesn’t get all of the credit. This attribution model assigns credit on a sliding scale moving backward from the last touch to the first. For example, the last touch might get 40% of the credit, while the second-to-last gets 30%, and the one before that receives 20%.
Custom
The first four attribution models are standard on Google Analytics, with other options if you use Google Adwords. They are relatively basic and you don’t need to be a data expert to use them. However, there is one more advanced option that some brands prefer when calculating attribution.
Custom attribution modeling allows brands to assign a value to sales through a variety of factors. One brand might give the first and last click 40% of the credit each, and then divide the remaining credit across the middle channels. Others vary the weight of the marketing channel, giving a paid search click more value in a sale than a social media click or branded organic search.
Custom models help analysts tailor their sales efforts to their industry, path length (pictured below), and customer behavior. The goal is to have a clear view of what marketing channels are driving the most-qualified leads and providing the highest ROI for the investment.
How Can Attribution Modeling Increase Your Conversions?
Like all data analytics insights, attribution modeling is only valuable if you take action from the information. While this data is useful for getting a feel for how your customers behave and view your brand, there are a few steps you can take to improve your sales funnel and increase the chances that your customers will convert.
Look for Holes in Your Marketing Strategy
Instead of viewing all of your marketing efforts as direct paths to purchase, use attribution models to consider your overall customer journey. Look for last-touch channels where customers don’t convert. What was missing? Why didn’t they come back?
When you start filling in these gaps with mid-funnel content and bottom-funnel sales items, you can reduce your lead abandonment rate and increase the number of people who convert.
Set Goals Based on the Sales Funnel
Attribution modeling will likely change how you set your marketing and ROI goals. Instead of expecting every marketing channel to pull its weight and drive sales, you might set lower ROI targets for top-funnel and mid-funnel channels and higher ROI goals for your bottom-funnel efforts.
Your paid search campaigns are easy examples of this. By separating your brand terms from your non-brand terms, you can set target goals for traffic and conversions based on customer behavior. Brand terms have a naturally high ROI because people are actively searching for your name. Meanwhile, non-brand terms can drive new traffic to your website and move people to the top of your sales funnel, even if the people searching for those keywords don’t immediately buy.
Better Allocate Your Marketing Budget
Once you set your target goals and discover gaps in your overall promotion efforts, you can adjust your marketing budget to increase conversions. At first glance, you might not think blog content or influencer marketing leads to sales and therefore would avoid investing more of the budget into those channels. However, your attribution modeling tools might show how these channels increase the chances that customers will buy later.
Another tool you can use to understand the value of your mid-funnel efforts is the Google Analytics Assisted Conversion report. You can see the top channels that move customers closer to buying, and sort the results by the path position and the number of days before a conversion.