Last-click is dead, long live the click; why last-click attribution modelling is so 2011

At our recent Bright Sparks event, Rachel Waller casually mentioned the death of last-click attribution models. It got me thinking. What is wrong with last-click attribution, and what’s come along to take its place? The majority of companies only started doing attribution in the last 4 years, so for a model that took the world by storm, it’s been a short run.

   

What is attribution?

Attribution is a method of consumer insight; it means assigning value to each of your acquisition channels according to how much you believe they contribute to each sale. It’s like figuring out which of your friends you have the most fun with, and using this to decide whose party to go to. I don’t know about you, but if I was awarded not according to my efforts but due to an arbitrary decision, in any field, I’d be put out.

Last-click is dead because small inefficiencies in the short term lead to long term misallocations of resources. In 2011, first or last-click attribution was used by over 80% of marketers, according to Econsultancy.

But the world has moved on.

Consider this journey: a customer Googles your brand and lands on your site through an affiliate link or an industry blog. A couple of days later, they’re on their phone and click through to your website. Later that evening, they’re on their laptop, they see your adwords ad and click on it. Once they’ve made a purchase from your site, which channel do you reward between these three? Businesses need some form of marketing attribution so they can figure out how to allocate their marketing spend. Businesses use last-click because it offers a simple solution to a complex problem, but if you’re looking to get involved with personalization, it’s a suboptimal solution.

So what is last-click?

Last-click is one of a variety of attribution models companies are using to attribute value to their different channels.It’s pretty much exactly what it says on the tin. It means the last channel a visitor arrives from before they convert gets assigned all the value of the sale.

Let’s imagine a customer, Maddie, who works for a multi-channel retailer and has two children. Maddie is shopping during the day, and arrives at your site from Facebook, having seen a product and liked it. She puts it in their basket and then closes the browser window; no time to be inputting credit card details when the kids have got football practice and piano lessons to attend.

Instead of abandoning the basket, Maddie decides to come back later to purchase. Next time Maddie gets to a computer, she fires up Chrome and searches the exact product name in the URL bar, knowing this is the easiest way to access the page. Would you say the value was driven by Google or Facebook? A tough question. It probably wouldn’t help if you asked Maddie, either.

Last-click attribution models get rid of this difficulty by assuming that all the value is driven by the last-click.At the end of the month, when it comes to assigning marketing spend, the proportion of money given to each channel is directly proportional to the number of conversions. Last-click does away with much of the complexity of assigning value. It makes a degree of sense: if I went to buy a pair of trainers from a shoe shop, you’d count it under the revenue of the shoe shop, so why wouldn’t you assign value in driving the sale to it also? Physical shops also have advertising, and despite its intuitive strength and ease of implementation, this lack of subtlety is what’s killed last-click.

Why is last click dead?

Last-click attribution can lead to wildly under or over valuation of different channels. For example, even if Facebook was involved in 99% of transactions but only 10% of last-clicks, this channel would be severely undervalued. Like claiming the winger who sent the perfect ball into the area had nothing to do with the striker’s header. Or that the health of a flower has nothing to do with the soil in which it’s grown. This negatively impacts revenue and gives you a false impression of the true value of your channels.

The flaws of last-click seem obvious when described from this angle, but up until very recently it was the go-to model for the majority of marketers. Recent attention from marketers has meant attribution has progressed. The previous industry standard has been overtaken by new, superior models.

What is it? Last-click attributes all the value of a sale to a single touchpoint

Benefits? It is easy to implement and update each month.

Drawbacks? Stops giving you the full picture as soon as customers visit more than once.

So what are your other options?

Attribution is an important part of the marketer’s toolkit. Don’t let a lack of ambition or no knowledge hold your marketing team back. Internal justification of KPIs and allocation of marketing spend is impossible to do correctly without sophisticated attribution.

With a proper behavioral model, your bid management systems will update on the fly, and you’ll never overpay an affiliate again. Qubit’s sophisticated attribution tool uses machine learning and Bayesian analysis to assess the probability at any given moment that the next conversion is going to come from a specific channel. Click here to find out more!