Understanding shopper behavior during Prime Day via AMC insights

Prime Day 2024 has come and gone. Now that the data is fully in, it’s time to dig into what we observed and if this is reflected in insights from AMC.

Overall performance & onsite observations

Prime Day returned for the 10th year, running across 23 different countries. According to Amazon, Prime Day 2024 was the largest Prime Day shopping event ever with millions of shoppers and endless deals. But did Prime Day truly live up to its hype for advertisers?

Globally we saw lower ASPs than in years past, indicating consumers were hunting for big discounts on lower priced items. Many advertisers felt Prime Day was softer than expected. However, Prime Day across the EU performed better than anticipated due to easier comps from last year.

This leads us to talk about what we observed onsite, which we believe contributed to the softer than expected holistic performance. The Amazon homepage was highly personalized this year.  Shoppers were shown deals for single products that were often products previously purchased or viewed, driving less opportunity for deal discovery. 

With ASIN-led deal tiles vs. brand-led deal tiles, significant personalization, and a reduction in SOV packages, we were curious to see how these UX changes impacted key metrics such as new-to-brand (NTB), add-to-cart rates, reach, and category-level data.‍

The importance of utilizing AMC

Before we get to the good stuff - a note on the importance of utilizing AMC insights when looking at performance from a deal event. 

When brands analyze their media investments with AMC, they receive a fuller picture of consumers’ paths to purchase leading up to the event. Before AMC, it was at best a guess. AMC allows us to look at consumer behavior on a more granular level, determining which consumers are most valuable during deal events and how brands can leverage these insights to drive more NTB consumers with higher long term values.

AMC findings were largely in line with onsite observations

New to brand increased slightly vs prior 30 days, but this increase was smaller than last year 

In the 30 days leading up to Prime Day 2024, brands saw ~76% of total product sales being NTB, meaning customers had not purchased from the brand on Amazon during the prior 12 months. On Prime Day, NTB % increased by 7%, which was slightly below the 8% increase from Prime Day 2023 vs its P30D average NTB %.

We would have expected to see higher NTB increases during a major deal event, but the personalization of homepage deals selected from past browsing behavior could be a factor. 

A high % of consumers added to cart before the event, but waited to purchase on Prime Day 

For most categories, exposure to brands that were purchased on Prime Day happened before Prime Day. On day 1 of the event, 25% of orders had items added to the cart prior to Prime Day, with a similar trend following on day 2. When looking at items added to cart on day 1, 16% of those were not purchased until day 2, showing some customers waited to see if deals would improve. Only 5% of orders had items added to the cart more than 15 days before the event, indicating that most of the pre-event add to carts occurred within 14 days of Prime Day.

For advertisers, these AMC findings mean it is critical to reach consumers specifically in the 2 weeks leading up to Prime Day, not just during the event - especially if Amazon continues with a personalized homepage approach.

AMC makes it clear why impressions are not a proper indicator of reach and traffic

Many brands use impressions as a proxy to measure Prime Day traffic, but with AMC we can compare impressions against reach to determine how accurate it is as an indicator of how many consumers saw an ad. Its obvious impressions will increase significantly over Prime Day, but brands should ultimately look to see what role frequency plays. Let’s see what we found when comparing reach vs. impressions:

Unsurprisingly impressions increased significantly across the board, but the percentage increase of impressions is significantly higher than reach, meaning many of the same eyeballs were hit more than once with ads. It highlights the shortfalls of using impressions to estimate reach. This also ties in with what we’d expect from personalization with a 36% increase in impressions vs 2023 despite reach only increasing 13%. This means consumers in 2024 were more likely to be hit with a higher frequency of the same ads and less overall ad variety.

However, it is encouraging to see increases in reach YoY as this means that traffic during the event continues to grow and there are more new consumers out there for brands to acquire.

AMC can help brands determine the viability of deal seekers

This leads us to talk about the cohort of shoppers that we call deal seekers. Deal seekers are shoppers who show up during a deal event to make a purchase who then never return or they only return to make purchases during future deal events. The trouble with deal seekers is it costs a significant amount to reach and capture these shoppers, and that effort might not be worth the investment if the cost of acquiring and retargeting them is higher than the return if / when these shoppers repeat.

Bringing this scenario to life with AMC, we’ve found that NTB customers acquired on Prime Day do spend more during the event than customers acquired the 30 days before or after Prime Day. However, these NTB customers spend less in the 270 days after Prime Day than customers who were acquired 30 days before or after Prime Day.

‍

With this information, brands need to weigh their short-term volume goals of a deal event against their long-term growth goals to determine if the juice of the deal seekers is worth the squeeze. Deal seekers greatly contribute to sales volume during a deal event, and if short term sales volume is the goal, it’s worth it to try and capture these customers during the event. If long-term growth is a larger priority, it makes sense to invest in retargeting these deal seekers after the event with custom AMC audiences to try and build up their LTV, transcending the deal seeker label.

When Amazon provides longer time periods for AMC datasets, brands will be even better equipped to evaluate the progression of deal events over time and if deal seekers truly pay out. Our AMC experts are consistently helping advertisers determine what audiences to invest in, how much to invest, and why. 

Our AMC experts are consistently helping advertisers determine what audiences to invest in, how much to invest, and why. 

What does this mean for brands for next Prime Day?

While we can’t directly connect changes in the UX to specific AMC metrics, there are clear correlations that help put consumer behavior and these results in context.

A Prime Day that was softer than expected can be concerning for brands, but the best preparation brands can do for these events is to invest in AMC audiences consisting of potential long-term customers during the lead-up to the event.

‍