In the follow-up to the sales event, be sure to steer clear of these too-common measurement mistakes.
Amazon Prime Big Deals Days are finished up. Now, there’s a good chance you – and your CFO – are in the thick of trying to understand not just how your brand performed, but how much marketing investments contributed to that performance. Of course, this information isn’t just important for understanding what drove traffic to the Amazon event alone. As a moment often viewed as a kickoff to the Q4 shopping season generally, understanding what’s driving the Amazon Day sales can provide crucial insights into your end-of-year strategy overall.
That said, there’s a best way and suboptimal ways to evaluate what worked and how. As you measure the impact your marketing had at driving conversions through Amazon, be sure to avoid these three mistakes. (These lessons apply equally to other retail “events” too, such as Target Circle Week.)
Mistake 1: Focus on Last-Click Results
The biggest mistake many brands make – even very savvy ones – is to focus only on the immediate last click prior to a conversion. Did a banner ad drive customers to your Amazon store? Maybe the final leg of the journey was an in-Amazon sponsorship? These last touches should clearly get some of the credit, to be sure. At the same time, it’s crucial not to ignore the impact of top-funnel initiatives that are creating interest upstream. Yes, sometimes customers see one ad or one deal and immediately buy. But even if the “Rule of Seven” – that it often takes a customer seven brand encounters to decide to buy – isn’t iron clad, it is true that the last click is often the outcome of many marketing initiatives, not just one. Be sure your measurement is set up to credit as many of those initiatives as possible.
(For a longer explanation of the flawed thinking of last-click measures, check out this thought experiment involving pizza.)
Mistake 2: Rely too Heavily on Amazon’s Measurement
Like all major platforms, Amazon offers analytics with insight into what drove customers to your products. This includes Amazon Attribution, which proposes to help brands “measure and optimize how non-Amazon marketing strategies help customers discover, consider, and purchase your products on Amazon.”
While all platform measurement tools can be useful, it’s critical to keep in mind that they’re not coming from an objective point of view. After all, as the Wall Street Journal reports, Amazon is now the third-biggest seller of digital ads, with ad revenue passing $12 billion in Q3 — “up 26% from the period a year earlier”—and plans to only keep growing.
Even with the best of intentions on the part of the platform, it’s impossible to fully separate any platform’s measurement and recommendations from its underlying business, which is getting brands like yours to spend more money with them. Amazon is no exception. For this reason, it’s important to not rely on Amazon’s own measurement and reporting tools exclusively, or even primarily. Be sure you have objective third party measurement in the picture as well.
Mistake 3: Don’t Account for Organic Upticks
To be sure, most marketers understand that seasonal moments can have an outsized impact on organic sales. They know, for instance, that Marketing can only take so much credit for sales spikes when the world’s biggest ecommerce brand drives customers their way.
But when it comes to marketing measurement, many marketers fail to account for these moments. Instead, they register sales activity on special event days as they would activity from any other day of the year; as a result, measurement systems can mistakenly read organic growth as a sign of marketing impact. That risk is especially high with the common measurement approaches that look only at overall average returns, and so aren’t set up to identify brief spikes as deviations from the norm.
In the short term, failing to account for seasonal organic upticks can lead to severely over-crediting marketing for incremental sales. Longer-term, this confusion can create models with the wrong presumptions of what an “average day” should look like – potentially cascading into a whole host of inaccurate measurements and faulty predictions.
In short: Don’t make the mistake of neglecting to denote Amazon event days – and all outlier days – within your datasets. Be sure to note special days as special.
(For more tips on handling seasonalities, read: “What Drove That Sales Spike: The Season, the Promotion, or the Ad?”)
To be sure, these aren’t the only traps we see marketers falling into when it comes to media measurement – on Amazon or elsewhere. Want to be sure you’re measuring media investments right? Talk to LiftLab.
