Attribution Modeling For Newbies

First-click, last-click, u-shaped? Attribution isn’t getting any simpler. Unless you’ve been in the digital space for a decade or more, it’s hard to know which model will be right for your campaigns or your organization’s goals. 

You can spend days digging through articles and graphics trying to figure it all out, but let’s make it simple. The team at Beeswax is always here to help, and we’ve created this handy, dandy guide to help you determine which form of attribution modeling is right for you. 

Before you even start to consider attribution models, let’s start with this: you can do this! 

Attribution modeling is complicated and confusing. The best way to figure out how to tackle it is to think about your goals and what you’re trying to learn. If you just need to show some charts to the C-suite, single-touch attribution may be all you need. That said, once you start really analyzing your campaigns with an eye to attribution, you may discover that you want to learn more. Understanding where people click, why, and how it impacts conversions can be fascinating stuff - if you’re nerdy like that. (We are!) 

The more you understand about what motivates your prospects and what drives them to click that orange button, the more you’ll want to experiment to see what drives increased leads and sales. That’s the time to start looking at multi-touch attribution models, and eventually at algorithmic or data-driven models. 

But before we dive into that, let’s start with the basics.

For the beginner - or the marketer who likes to keep attribution simple:

Let’s start with the basics. These are single-touch attribution models, and they’re a great jumping-off point as you begin to explore attribution modeling: 

• Last-Click Attribution: Newer to digital? Just need to report on conversions? This is the simplest way to track success. With last-click attribution, 100% of the credit for the sale or form completion (or whatever constitutes a conversion in your campaign) goes to the last campaign interaction. The great thing is that you can accurately track which creatives led directly to conversion, which is really helpful. The downside of this attribution model is that it ignores all previous interactions. That means that a prospect may have found your site through search and may have been seeing your ads or reading your blog posts for months - but the only interaction you’ll be able to track is that single click. Again, if you’re laser-focused on what directly persuaded your prospect to convert, last-click will work well for you. 

First-Click Attribution: If you’re more concerned with awareness than conversion, first-click attribution will tell you almost everything you need to know. This model puts all the conversion emphasis on the first interaction - even if the prospect engages with seven additional touch points before they convert. That puts the focus on the top of the funnel, or the interaction which introduced that lead to your company or offer. If that’s the kind of data you’re looking for or what your C-suite wants to see, this is the attribution model you may want to use.

Attribution modeling for the experienced digital marketer, or marketers who want more data to guide their optimization.

If you’re a more seasoned marketer running more complex campaigns - or if you’re just ready for more data about the interactions that led to conversion, you may want to consider one of these attribution models. They’re more complex and maybe not really for “dummies,” but they’ll give you a better picture of your campaign performance than the single touch models.

• Linear Attribution: The simplest of all the multi-touch attribution models, linear attribution assigns equal credit to every touchpoint along the customer’s journey. The downside is that you won’t know if one interaction had a greater impact than another - but at least you’re able to see which ones are contributing to your success. This is helpful if you’re running a lot of ads in a lot of places and you’re not sure which channels are worth a continued investment of your budget or your energy.
• Time Decay Attribution: Similar to linear attribution, the time decay model considers each touchpoint, but gives more weight to the interactions that occur closest to conversion. As Facebook puts it, “Compared to a last touch or last click model, time decay better reflects how all touchpoints helped lead to a conversion in a way that may more realistically represent how customers interact with and consider ads as they get closer to converting.” It’s not perfect, but at least it considers every touchpoint.
• Position-based (aka U-Shaped) Attribution: This brings us a little closer to a true picture. While it’s not based on true engagement, this attribution model serves 40% of credit to the first and last ad interactions, and then divides the remaining 20% across the interactions in between. Again, it doesn’t deliver the most accurate picture as far as which touchpoint contributed most to conversion, but it does look at your campaigns holistically. 
• W-shaped Attribution: The W-shaped model is similar to the U-shaped one, except that it takes five touch points into consideration. This model assigns 30% of the credit to each of the first, third and final marketing touch points, then splits the remaining 10 percent between the second and fourth touch points. This gets you closer to a 360-degree view of campaign performance, but it may undervalue those interactions between the points of the W. 

For the really seasoned marketer - the one who knows what the execs want to see AND knows that they want to get the best out of every campaign, there’s data-driven attribution. This attribution model uses algorithms to deliver an honest, unbiased attribution for every campaign, so you’ll see which touch points truly delivered the greatest bang for your buck. This is really important if you’ve got several creatives appearing across multiple networks and devices. If you’re relying on a W- or U-shaped model, the most successful ads could be at some point toward the beginning or end of the buyer’s journey and not getting the credit they deserve. That could lead you to funneling budget into lower-performing ads, but you’d never know it. 

Data-driven attribution isn’t for everyone. To qualify for use of the attribution model in Google, you need a lot of data and hundreds of conversions over a period of weeks. A marketing professional would have to have many campaigns under their belt to meet those requirements. However, once you’ve reached that threshold, this type of attribution modeling is definitely the best if you want to truly understand how well your campaigns are working. You’ll gain a clearer view into what’s working and what isn’t, and have the opportunity to optimize for greater success.

Hopefully, this guide to attribution modeling has been helpful. While you’re most certainly not a dummy, attribution - and analytics in general - can certainly make you feel like one sometimes! Attribution modeling is tricky business, and even the most experienced marketers can get tripped up. 

So give yourself a break. Spend some time figuring out what your KPIs are and what you need to report to the business’s stakeholders. From there, you should be able to figure out your course. Consider starting with single-touch attribution, and then work your way up to the more complex models. You’ll figure things out as you go. 

You’ve got this!