Last click: why you should ditch this noughties attribution metric
Last week, consumers worldwide directly searched for products they had no prior knowledge of, imagined up the names of brands they had never heard of before, and immediately made a purchase without considering other options.
Think that sounds wrong? Well, that’s what last click is telling you.
Obviously, there are other stages in the customer journey. But last click attribution gives 100% of the credit to the final click before the purchase was made. Not only is this misleading, it could be actively damaging for your business.
A business’s growth happens at mid/top funnel, so it’s absolutely critical to long-term growth to understand where these new customers are coming from.
Time and time again we speak to companies who are measuring channel performance with last click data, which can seriously under-value discovery channels such as Meta and TikTok and over-value direct channels such as Search. This can severely limit a company’s ability to scale.
With data visibility decreasing year on year – and with no sign of stopping – it’s never been so important for businesses to move away from using last click as the sole metric for understanding channel performance.
Old habits die hard
Back in the noughties, Analytics helped solidify Google as king of the internet. Digital marketing was dominated by PPC and affiliates, so tracking performance with last click on Google Analytics made perfect sense.
Lots of people still love last click, particularly today’s decision-makers who forged their careers in these early days of the internet.
Last click is great because it makes sense. Last click adds up nicely. Last click is easy.
Last click is really effective in the short term. If something’s doing well, whack up the budgets and you’ll reap the rewards soon after.
But sadly it’s not the noughties anymore, and relying solely on last click is one of the easiest ways to mess up your business long-term. Don’t just take our word for it – even Google is turning its back on last click.
What last click isn’t telling you
User journeys in 2022 are more complex than ever, and a direct response measurement tool from twenty years ago is too blunt an instrument for understanding what initially drove the purchase.
If you’re selling high-end fashion, for example, your potential customers will probably need a few days to consider the purchase before making it. Or maybe someone sees an ad on Facebook, searches the product on Instagram and clicks through via an influencer. Last click will give the wrong channel for these two very common user journeys.
According to research from Meta, last click undervalues Facebook by 47% on average, and Facebook was undervalued in 66% of cases in the study. The study also showed that inaccuracy of last click data increases as you move up the funnel into prospecting activity. This is particularly important if you’re looking to grow your business quickly with new customers – like the clients we work with.
How to measure actual marketing performance
So how do we measure actual performance in 2022?
To some extent, you can’t. Perfect measurement is like Santa Claus: it doesn’t exist (sorry to any young children reading this blog).
And measurement is only going to become harder over the next few years. We’re now past the point of peak digital tracking. Changing expectations on personal data privacy has led to GDPR, cookie consent and iOS 14.5, and Google’s FLoC announcement gives us a hint of a cookie-less anonymised future.
In order to invest in your business’ long-term success, you have to embrace the anxiety of not quite knowing what’s going on.
What are the alternatives to last click?
Most importantly, you need to use several different models to try and understand channel performance. Last click can be a fantastic tool, but only when used in conjunction with others. Measurement is now an art, not a science.
Here are some things you can try:
- Understanding of channel value change? To grow your business, you need to understand how people are discovering your brand, not where they end up converting. There are lots more attribution models to try out, such as Time Decay and Position Based. These are the six main ones:
- Look into the correlation between internal KPIs and Facebook in-platform performance. Can you create a proxy between something like internal CAC and in-platform ROAS? Lots of our clients measure paid social performance this way.
- Do a stress test, where you significantly increase paid social spend for a period of time. What happens to overall performance? Do you see other channels start performing better?
- You could work with a measurement partner (e.g. Fospha) for a really detailed data-science approach.
- Come to us and we can help you run a conversion lift study to prove the incremental value of Facebook.
There has never been such a wide range of channels to help grow your business, and the sudden rise of Instagram influencers and TikTok have shown just how quickly consumer shopping habits can change.
Consumer journeys have become more and more complex, and it’s absolutely essential for the long-term success of your business to understand channel value in a much more nuanced way than using last click alone. It might just save your business in the long run.