In an era where digital marketing is holding more weight in a CMO’s budget, it confuses me as to why so many marketers still measure effectiveness based on last-click attribution. Multi-touch attribution is no longer a concept or technology that is only for those with deep pockets.
Attributing the success of your business to the last touch point that delivered your paying customer is as effective as valuing a football team purely based on their strikers. Sure, Player X may be striking the ball into the back of the net at least once per game, and it seems like it’s the star player we should be keeping an eye on. But what about Player Y, in defense, who manages to steal the ball away from the opposition in the first place? Or how about Player Z, who skillfully weaves the ball up the field in search of X? Is your perfect football team one that is made up purely of Xs or one that has the right mix of Xs, Ys, and Zs? That is exactly the type of question multi-touch attribution aims to answer.
Now imagine that:
- The manager of the football team is you
- The ball is your potential customer (but please don’t kick them in real life)
- Goals are sales (or leads or brand engagement)
- The opposition is your competition
- Player X is Google Search
- Player Y is Facebook
- Player Z is the Google Display Network
Traditional last-click attribution-based planning tells you that you should invest all your marketing dollars into Google Search, because that is what is bringing you sales. However, it completely ignores the fact that your activity on Facebook may have convinced your customer to consider you over your competition (i.e., stealing the ball from the opposition), and that branding efforts on the Google Display Network drive brand-related searches on Google (i.e., passing the ball to the striker).
Overcoming Last-Click Attribution Bias
Having a bias toward last-click attribution is inevitable. The digital marketing ecosystem evolved with this form of attribution at the forefront because of the technological limitations that existed 10 years ago. Fast-forward to today, however, and there is no excuse not to look beyond last-click.
First, most publishers and platforms provide some form of metric that allows you to evaluate their effectiveness beyond the last-click, such as view-through conversions, AdWords’ search funnels, and so on. Making use of these is a first step to understanding the effectiveness of marketing channels before the final conversion.
Second, affordable technologies now exist that allow marketers to effectively evaluate performance beyond the last-click. An affordable – i.e., free – solution is Google Analytics’ Attribution Modeling Tool. The Attribution Modeling Tool allows you to evaluate the effectiveness of your digital channels, across multiple sessions to your website, using weighted models of your choice.
A Practical Example: Amari Hotels
Amari Hotels were able to increase their online bookings by 47 percent simply by reallocating their marketing budgets so that activity on the Google Display Network went from constituting 7 percent to 17 percent. No extra resources required; only a reallocation of existing resources.
Identifying how to reallocate their resources was achieved thanks to the insights derived from Google Analytics’ Attribution Modeling Tool. After experimenting with various standard and custom models, Amari’s marketing team determined that the linear and time-decay models best reflected the reality of their campaigns.
The linear model allocates the value of a transaction equally to all channels that contributed to it. Whereas the time-decay model weighs more in favor of channels that drove traffic more recently.
Applying their chosen models to their conversion data yielded a report that looked like the below.
The report illustrated how the value of their marketing channels changed when they applied a different attribution model to their conversions, and how much that value differs from a last-click point of view. This allowed Amari to evaluate by how much they were under- and over-valuing channels.
What stood out for Amari was by how much they were undervaluing their efforts on the Google Display Network. Both models showed that by taking a last-click attribution approach to their display efforts, they were undervaluing the impact of this channel by a whopping 500 percent!
Through a series of experiments, Amari was able to determine that increasing their Google Display Network activities to constitute 17 percent of their digital marketing mix would allow them to increase online sales dramatically. Amari could not have yielded such a result if they had continued evaluating their marketing efforts on a last-click basis.
This has now become such an important area, that Amari has institutionalized the value of multi-channel attribution analysis by integrating the above metrics into their reporting frameworks.
Next Steps for You
Achieving the same success as Amari is not beyond any digital marketer’s reach. If you’re not sure where to begin, I suggest doing the following:
- Implement Google Analytics on your website.
- Define goals in your Google Analytics account. If you sell online, you should also integrate e-commerce tracking.
- Navigate to the Attribution Modeling Tool.
- Select two standard attribution models and compare it to the last-click attribution model.
- Which channels are undervalued? And which are overvalued?
- Consider redistributing your resources from overvalued channels to undervalued channels.
- Test, observe, and refine your resource allocation strategy.
- As you become more comfortable with multi-touch attribution, start to experiment with your own custom models.
- Last, but definitely not least, integrate attribution model metrics into your reporting framework.
The Attribution Modeling Tool opens up powerful insights for digital marketers – insights that were previously locked behind expensive or difficult to implement technologies. Get started with it today to gain a new perspective on your players on your digital marketing football field.
Marketers need to know what’s in their data and trim out the filler to provide continuous, data-driven ROI for their brands.
All top Chinese retailers, banks and internet companies share mobile data in earning releases. None of the top 10 US retailers do, nor does Google. US banks and Facebook are better.
A new starter in Team SaleCycle recently asked me the following question… “Wouldn't they just come back anyway?”
American Apparel's chief digital officer discussed the future of retail, the importance of delivering value to the consumer, and strategies for an IoT and omnichannel world.