Understanding Attribution Modelling on Google Ads

Google Ads isn’t for sissies. It takes a heck of a lot of work and practice to get the hang of things and earn yourself some ROI. And we’re only talking about the basics here – creating an account, campaign, ad group, keywords, and the ad itself.

But if you want to make your business successful, then you need to learn more than just the basics. You need to have a more in-depth understanding of your business model, your industry, and your target market. And to do that, you need to learn all about attribution modeling.

What is attribution modeling?

We all take that journey from browser to buyer. The steps we take before making a purchase can give marketers insight on how, when, and why we bought a product.

For example, you see an ad on Google. You click on it and buy the product being advertised. Obviously, it was the ad that converted the browser into a buyer. Of course, we don’t always operate that way. The customer journey is almost never straightforward. According to the Online Marketing Institute, it takes 7 to 13 touches (also known as engagements, and also depends on the product) with your business before that lead becomes a sale. This means that they may have seen your ad on Google, checked out your Facebook and Instagram page, visited your website, downloaded your app, and finally bought something in your physical store. So, how do you figure out what, where, and when the customer converted from a lead to a sale? This is where attribution modeling comes in.

An attribution model is the rule, or set of rules, that determines how credit for sales and conversions is assigned to touch points in conversion paths.

Here’s an even simpler explanation from Brad Smith, a software engineer in Google: attribution modeling is telling your analytics which channel or keyword to give credit for the sale/conversion. In short, an attribution model makes sense of all the data gathered and gives you a clearer picture of what channels and keywords contributed most to conversion events. With this data, you get to know what are the keyword groups, ad groups, and campaigns that will yield you the most ROI for future budget allocations. After all, you don’t want to keep spending on marketing efforts that are not bearing any fruit at all. In addition, you don’t want to overlook channels that are actually performing.

More importantly, you don’t want to spend time guessing instead of actually knowing what to do.

Google Ads Attribution Models:

Google Ads has several attribution models that you can choose from. Take note that there is no right or wrong attribution model. And no single attribution model can give you the whole story. It is important that you not only know what each attribution model is but when you should use them.

Last click:

This is the default model on Google Ads. With this model, credit for the conversion is given to the last-clicked ad and corresponding keyword. As we’ve already mentioned, it’s pretty rare that the customer simply saw your ad and clicked or visited your website and clicked purchase. What’s more likely is that they saw your ad, probably visited your social media accounts to see what people were saying, saw another ad again, and then visited your website directly and made the purchase.

That’s not to say that you shouldn’t use the last-click attribution model. It has its place. In fact, if you’ve already generated quite a few leads through your various campaigns already, you might want to learn which ones are ultimately causing your leads to convert. In short, you want to know which of your keywords, ads, and even campaigns are your closers. Another situation that calls for this type of attribution model is for companies that already have an established brand and pipeline and need to keep spending within a limited budget. Focusing your money on just the channels that will have a higher rate of conversion is key. This model lets you know which ones will work best for you.

First click:

Similar to the attribution model above, this type only relies on a single touchpoint. The difference is that this model gives credit to the original source – the channel or keyword that first brought the customer’s attention to your brand. And like what we’ve already mentioned, this rarely paints you the full picture. It ignores other subsequent engagements that may have occurred which may have contributed to the conversion event. Yes, your ad may have caught the attention of the customer initially but was it really the one that “closed” the deal?

While it isn’t the best attribution model for determining which keyword group is driving your conversion rate, it is a great one to use if you want to determine which campaigns are best for brand awareness and top-of-the-funnel engagement. It helps you figure out which channels are best for generating new prospects, particularly ones that are most likely to buy.

Linear:

This attribution model divides the credit equally among all the touchpoints on the user’s journey towards conversion. It doesn’t matter how much value each touchpoint provided in the sales cycle, every one of them gets equal weight. The benefit of using this attribution model is that it helps you determine which channels and keywords are contributing and which ones aren’t. Another instance where this works best is if your campaign is designed to maintain contact and awareness with the customer throughout his/her journey.

It also is best used if you are running a complex, long-winded campaign that starts with a broad search term and leads the customer to conversion. For example, you can use a broad keyword in a blog post and then remarket the readers to learn more on your website. You can later on also remarket them with RLSAs (remarketing lists for search ads). Their journey then goes from brand awareness to conversion, with each touchpoint playing an important role. If this is what you’re doing, then you should use the linear model.

Time decay:

Like the linear model, the time decay model gives credit to all touchpoints along the user’s path. However, it does not distribute credit equally. Interactions that occurred closer to the sale are given more value than those that were taken further back in time. Credit is distributed using a 7-day half-life. In other words, a click 8 days before a conversion gets half as much credit as a click 1 day before a conversion.

This type of attribution model works best if you have a long sales cycle and you want to determine which channels work best in moving leads down your sales funnel towards conversion. This isn’t an ideal option if you’re running a simple Google Ads campaign or selling eCommerce products online as your sales cycle is pretty short. That being said, this is a good tool for assessing promos.

Position-based:

This is another multi-touch attribution model. Like the others, it also gives credit to all touchpoints or clicks on the customer’s journey. However, the distribution of value is different. The first and last click both receive 40% of the conversion credit while the rest of the remaining clicks divide the 20% among them.

Compared to the first-click and last-click model, the position-based model (a.k.a. U-shaped model) gives you more concrete data and a clearer picture of your customer’s journey. You get insight on what keyword group or ad caught their attention, which one closed the deal, and what keywords helped them move along the funnel. When should you use this? If you want to see which combination of brand awareness and closing work best, then this is the model for you.

Data-driven:

This attribution model is a new addition to Google Ads. Instead of a set of rules assigning credit to certain touchpoints along the customer journey, this uses an algorithm (the Google machine learning technology) and accumulated data to determine which touchpoints have the most impact. This data-driven approach makes choosing a model moot because it does it for you, based on your account performance.

Google explains how their data-driven model functions here:

“Sophisticated algorithms evaluate all the different paths in your account (both converting and non-converting) to determine which touchpoints are the most influential. Factors such as the number of ad interactions, the order of exposure and the creative assets used in each conversion path are all incorporated into results. Using a counterfactual approach, the algorithms contrast what actually happened with what could have happened to determine which ad clicks are most critical for a conversion.”

Sounds good, right? No need to guess which model works best for your business goals. Everything is fully automated. However, there is a catch. Only certain accounts can avail of this type of attribution model. To be eligible, you need to have a minimum of 15,000 clicks and 600 conversions in 30 days. To be able to continue using the model, you will need to be able to sustain a performance of 10,000 clicks and 400 conversions. In short, you need to be spending a lot (and I mean a lot) of money per month on Google Ads to be getting this many conversions and clicks. If you are eligible, it’s best to go with this attribution model as Google has already proven that this works very well through testing.

Choosing the Right Attribution Model for Your Business:

As we’ve already mentioned, no attribution model will truly give you the full picture. Each attribution model provides you with a certain perspective and insight into your sales cycle. To find out which one to use, you’ll need to think about what question you are trying to answer.

For example, if you want to know which channels generate new prospects, then you need to use the first-click attribution model. If you want to know which interaction led to closing the deal, then you need to use the last-click attribution model. That being said, you can’t generate two reports and combine them, expecting to get the full picture. Doing so will cause you to double-count your conversions.

As we’ve already mentioned, all attribution models have their pros and cons, and work best in certain situations, campaigns, and types of businesses. Because there’s no right or wrong model, you will need to play around with the different options to see which suits your needs best.

Here’s a pro-tip: Google Ads actually lets you compare different attribution models side-by-side. Click on “Search Attribution” in your tool dashboard, under the Measurement category, then choose  “Attribution Modeling” to run attribution reports and see how each model impacts your data. If you want to determine which keywords, ad groups, or campaigns that are undervalued on a last-click basis, you can start with the by comparing the first-click model and the linear model.

Menu