Last Updated on March 15, 2022 by Growth Hackers Sthlm
Different types of ads affect your potential customers during different parts of the customer journey. Because of this, you need to analyze the customer journey both in its entirety and per channel.
You need to understand how long the whole journey takes and how long individual channels affect. You also need to understand when during the customer journey the different channels affect the customer. Because it is often different. The time between an ad click and a conversion is called an attribution window.
Each individual channel has its own attribution window and should each individual ad lead directly to a conversion, it would be enough to watch there. But that is rarely the case. To be able to attribute the right value to the right channel, you need to analyze your advertising based on an attribution window that is as long as the customer journey itself.
Channels need different lengths of time, so the attribution window must cover the entire customer journey
If you use too short an attribution window, you risk believing that some channels will not work or underperform. In fact, it may be that those channels just need more time than others to influence. With an incorrect attribution window, you run the risk of turning off errors or prioritizing incorrect advertising channels.
The graph below shows a multi-touch attribution model, ie a model that takes into account and evaluates assisted conversions. This means that an ad click that occurred on day 1 will receive the value from a conversion on day 5
The importance of getting the right attribution
Look at the dark blue line in the graph. It shows how an advertising channel has performed. After 10 days, only about 80% of the value that will actually be attributed to the clicks from day 1 is included. Should we use an attribution window of 10 days and make a decision on the information we have then we will believe that we only get SEK 80 for our invested SEK 90. That is, we go minus on advertising. This is not true when we look over a longer period of time and then instead see that we get SEK 100 for our invested SEK 90. In this example, we have 111% ROI instead of 88% from this channel.