How much can we trust the last click?
Industry standards say it takes multiple visits to a website and multiple touches to convert a user. Through search ads, banner ads, affiliate programs and email campaigns, people have different kinds of scenarios to surf on the internet and spend their money. But which activity gets credit for a multi-touch conversion?
Imagine you saw a homepage promotion of NIKE, but you left the page for some reason. On the next day, you performed an organic search on google for NIKE shoes and then you received an email from the brand but with no click through. A week later, when you were reading news on mobile, you saw a NIKE banner ad which was too intriguing to ignore, so you clicked the banner and made a conversion on the landing page. In this case, many would regard the banner ad as a vital element that drives revenue. Well, yes and no. It depends on what attribution model you employ.
There are several attribution models companies often use: first click, last click, linear credit, split credit, and decay model. Some consider first click is important because ‘if it were not for the very first campaign touch, the consumer might never have known we exist’. Others may not sure how to give credit to their various campaigns so they just give credit evenly (linear credit). You can also give 50% of credit to last campaign and split the rest between the remaining (split credit). Or, you give the last click more credit and then decrease in value each previous campaign (decay model).
As for the example of NIKE, every step counts for the final conversion. If you haven’t seen a homepage promotion, you may not know NIKE was on sale. If you didn’t google the brand the next day, NIKE may not know you have the intent to buy something then send you an email. Even if you didn’t click the email, you still have been interacting with the brand and finally made a purchase after clicking an intriguing banner ad. From my point of view, last click do matter but previous activities in this scenario also counts and affect the conversion. Hence, I would prefer decay model applied in this case.
Last click is the standard attribution model in all web analytics tools. It is applied to all the standard reports but there are still some problems: Shouldn’t other campaigns get credit if they touched the customer within a reasonable time period before conversion? What about products with long purchase cycles? To solve these problems, other attribution models should be taken into considerations.