Selling a product or services on the internet using an online platform is similar to in-person sales. Somehow, it can be argued, especially when it comes to determining how a conversion happened. Let’s think of this way.
Imagine that you are running Google Ads, Facebook Ads, Email campaigns, SEO, Snapchat Ads, and more paid campaigns for your E-Commerce store on multiple platforms. You use Google analytics as your central analytics tool to track the online data coming from various sources. On average, you are getting 100 purchases per day. You went to Google Analytics to find out from where they came. You discovered that most of them came from Google Ads & Direct traffic. In context to the same, you further decided to switch off Facebook and Snapchat ads as they were negligible purchases from both the ends.
But suddenly your purchases deflated around 20-30 per day. You are wondering what happened, have tried every possible solution to get back to the purchase volume. But, everything failed. The best possible thing that you missed is that your google analytics is set up on the “LAST CLICK” attribution model.
What are an Attribution Model & Types
We know the question you’re thinking about – what is this attribution model & what does it have to do with the purchases/conversions? Attribution Model is a set of rules that determines which touchpoints (marketing channels) receive credit for a purchase/sale & how the credit is distributed among them. In other words, it is a process of determining how the credit for sales or conversion is assigned to different touchpoints when it comes to customer journeys.
Last Click Attribution Model
The word explains itself very precisely. The Last Click attribution model is the default model in most of the Google proprietary tools such as Google Adwords or Google analytics. This model credits 100% value to the last clicked marketing channel in the conversion path. In the example discussed above, it is either Direct or Google Ads.
As you can see, last-click attribution ignores the efforts you’ve added to the other channels and can make you spend more on the wrong marketing channels.
This attribution model is only beneficial if your customer’s journey cycle is small & doesn’t comprise of many touchpoints. But, we recommend you not to use the last-click model for my accounts to get a better insight into the customer journey & to know which marketing channel you need to invest in more.
First Click Attribution Model
The First Click attribution model is like the far-end brother of the Last Click. This attribution model also gives the entire credit to a single touchpoint in the conversion cycle. In this model, the entire credit is assigned to the first marketing channel from which buyers came to your website. In the above example, it is Facebook or Snapchat ads, it doesn’t matter if the customer found you on Facebook & watched your youtube ad, and then a fortnight later made the purchase by going directly to your website. The 100% credit will go to Facebook/Snapchat ads.
This attribution model is very useful for knowing which platform is responsible for bringing customers into your website. You may find that Facebook ads are performing very well. In that case, you can invest more budget & content into it. On the other side, this model neglects the efforts of other assisting channels like Youtube, e-mail, organic & Google ads. If your main goal is to acquire new top-of-the-funnel customers to scale your business & your customer cycle is small, the First Click attribution model can be your star model.
Linear Attribution Model
This is where interesting things start. Linear Attribution model credits to every touchpoint in the customer’s journey equally. This model simply assigns the same amount of credit to every touchpoint. In the above example, we have 4 touchpoints. Each touchpoint gets 25% of the total credit. Just like, a mother loves all her children equally.
This is the most used attribution model among marketers & should be used in the scenario where you want to know about all the marketing channels that are contributing to the purchase funnel. But, this attribution model also overvalues minor touchpoints and undervalue major touchpoints in the funnel. Do you think all touchpoints play an equal role in making a purchase? It’s unlikely the case. Thus, sometimes it’s tough to identify the best performing marketing channel in the conversion cycle.
Time Decay Attribution Model
Time Decay attribution model works exponentially. The touchpoint closest to the purchase will get the highest credit. As you go back to the farthest touchpoint, the credit decreases. Unlike the linear model, this attribution model gives more weight to the touchpoints in the middle where a user starts getting convinced about making the purchase. The closer to conversion, the more weight it gets.
If you want to focus on marketing channels that are closer to the conversion, this attribution model is the best for you, especially if you have a complex and long purchase cycle. Is this model perfect? Not quite. Despite providing excellent information about the purchase cycle, it highly undervalues the credibility of the first touchpoint.
Position-based Attribution Model
The amalgamation of the Linear & Time Decay attribution model. The position-based attribution model credits 40% value to the first and last touchpoints in the purchase cycle and the remaining 20% to all the other touchpoints equally. As discussed earlier, 40% credit goes to Facebook/Snapchat and Direct/Google Ads channels & 10% to both Email/Youtube and Organic channels. This attribution model can be insightful if you are interested in knowing about the first & last touchpoints in the conversion cycle. You can also change the credit percentage as per your business goals.
Position Based attribution model fits the best if your business brand value is low or it probably has low-ticket products where decision making doesn’t need to be very complex. But, if you have a pretty big e-commerce website & your products are of high-ticket, this attribution model can be non-informative, as it will give around 80% credit to the first & last touchpoint, leaving the middle, decision making touchpoints with low credit.
Data-Driven Attribution Model
The mother of all attribution models – Data-Driven. The data-driven attribution model uses machine learning to evaluate the conversion that is impacting touchpoints in the conversion cycle. This model differs from all the above-mentioned attribution models as it assigns credit to the touchpoint based on how much it actually contributed towards the conversion. Like in the example above, it was decided that email campaigns contributed the least & Google Ads contributed the most, while organic search also contributed a lot to the conversion cycle.
If you ever fall under the requirements, you should definitely go for this attribution model. But, the requirements to enable this attribution is massive. You need to have at least 15,000 clicks & 600 conversions, for example in Google Ads, in the last 30 days to be able to use the data-driven attribution model.
We would suggest going for the linear attribution model if most of your conversion cycle has less than 3 to 4 touchpoints, or else, you can go with the position-based attribution model. On rare occasions, we go with the first-click or last-click attribution model.
- The smart-display campaign is not performing at all in any attribution model. I am going to pause it and shift the budget to other campaigns.
- DSA-category-Gesicht – In terms of last-click attribution, it’s ROAS is 1.46 but when we see the same campaign from the first click attribution point of view, this campaign is doing great in terms of bringing in a new audience.
In fact, in last 30 days, ROAS is 2.68. So, where is this difference going? This difference is being attributed to the direct channel where returning users are making a purchase directly from the website. The same applies to the DSA-category campaign.
If you see at the direct channel data, the difference in Conv. value is around 6000. That’s so because the first attribution is taking away the last-click attributed data from the direct campaign.