This week, in data
This week’s data covers Amazon’s Prime Big Deal Days, their early-Black-Friday cannibalization strategy designed to capture holiday purchase intent long before the big weekend.
It worked: Americans spent nearly $2 billion dollars on Amazon during the two day event. That’s two billion that won’t get spent during Black Friday, and Amazon gets a healthy cut of that two billion.
To help compare: Americans spent about $35 billion over the Cyber Five in 2025.
This is the third big selling event of the year for Amazon, where it now pushes spring, summer and fall sales events without consideration of existing marketing tentpoles like Valentine’s Day or the Fourth of July.
It keeps getting more intense. Amazon captures about 40 cents of every dollar spent online. Amazon’s net sales in Q2 jumped 20% year over year. Advertising revenue jumped 26% over the same comparison period.
So this strategy of “early Amazon-controlled Black Friday sale” is working, clearly. Amazon won’t stop until it controls everything - there’s a great book about this.
It makes sense. If I was Amazon, I would want to train American consumers to move away from these frivolous and un-trademark-able shopping events like “Mother’s Day.”
(This is satire, obviously. I love you Mom, thanks for reading!)
But in all seriousness: it feels as though a century of American purchasing habits are starting to bend toward Amazon’s business goals. American Consumers, the bedrock of our CPG/DTC economy, love a discount, and Amazon loves giving it to them.
The prevailing opinion is that “American consumers won’t buy in late September or early October, and conversion lag is long, because they are waiting for a discount on BFCM.”
Amazon’s strategy? Just give them the discount now, then. Black Friday is a made-up holiday that started in Philly decades ago. Why can’t it be disrupted?
Which leads me finally to my question, dear reader:
If Amazon decides to reshape the entire rhythm of the American consumer’s purchasing habits, what are we supposed to do about it?
“Should we be on Amazon?”
You aren’t truly in the ecommerce game unless you have asked, been asked, or answered this question.
The answer is neither simple nor permanent. Amazon will continue to influence purchasing habits and seasonality in their pursuit of complete ownership of global commerce.
So, you have to evaluate (quarterly, in my opinion) WHY you are or are not selling on Amazon.
On one hand, selling on Amazon is a Faustian bargain. Amazon takes a cut, but you could potentially convert new customers who would never consider buying from your Shopify store.
On the other, the margins might not work for you. The halo effect or cannibalization effect of Amazon on your DTC sales are tough to predict. Are you finding incrementality? Or are you just acquiring the same customers, except now you are paying a tax to Bezos on top of it?
So how do we decide to '“be on Amazon” or not?
Here’s our experience from working with thousands of brands dealing with Amazon.
Follow the numbers
First, separate the conversation into two:
Should we list on Amazon?
Should we advertise on Amazon?
Each is a mathematical equation whose values are perpetually shifting.
Listing on Amazon
If the unit economics work for listing your products on Amazon, then you probably should. Customers will check Amazon first because they want the shipping, or they want to make multi-store carts.
Even if you don’t spend a dime advertising your Amazon store, just having inventory there will generate some level of incrementality.
Use correlation analyses to understand the interaction between your decision-making on Amazon and your Shopify stores. You can use Northbeam’s Metrics Explorer to do some of this.
Does your revenue increase when you begin listing on Amazon, even if all other variables remain equal? If so, that means Amazon is incremental to some degree.
Is there a correlation between decreasing DTC sales and increasing Amazon sales? That suggests that Amazon is cannibalizing your DTC purchases — but not necessarily that you should remove your Amazon listings.
To decide if you should list on Amazon, approach this like a math problem. Answer this question with data analysis: “Am I generating or capturing demand with my Amazon listings, and how is that affecting my profitability, revenue, and ROAS?”
Advertising on Amazon
Let’s say that just listing on Amazon gives your brand a halo effect. So how do we decide to put ad spend on Amazon or not?
Once again: through the glory of math.
Don’t allocate Amazon budget just by looking at platform ROAS. You need to separate Amazon, DTC, and retail (if you have it) and measure each channel’s performance independently.
You need to understand what is demand capture versus demand generation. Amazon’s search and branded campaigns are useful for defending your brand (like on Google) but these can have diminishing returns. Separate these strategies out into different budgets.
Amazon is best measured with correlation and incrementality. If you really want to understand halo versus cannibalization on Amazon, then you need to use appropriate measurement tools to identify those effects. Do NOT just look at revenue numbers and make a call. That’s just vibes, and as operators, we don’t act on those.
Speak with your operations team, as supply and customer economics are the biggest levers on Amazon. Having limited supply or rapid stock turns could mean it’s better to just NOT advertise there. Or perhaps you want to feel “upscale” as a brand and therefore not cheapen your brand on Amazon, but I feel as though this thinking is a little outdated.
Watch out!
Amazon is not all sunshine and rainbows. Stuff to look out for:
Amazon is watching you. It is well known that Amazon will take best-performing products from their marketplace and launch white-label versions to compete, but this has somewhat relaxed.
Paid performance can be hard to separate from organic sales. Make sure you establish a (believable) baseline of performance before you launch, so you can really see what’s incremental.
Amazon’s promo periods can eat away at your profitability. If they keep pulling demand forward and eating your sales away at lower margins, you’ll just be less profitable. Pay attention to this.
I want to hear from you. What are your opinions on Amazon these days? Reply to this email and let me know. Or tweet/chat/DM me here:
🚨 Meta’s new ad terms lets them test your ads without your notice. Everything you need to know, from our friends at Common Thread.
🌐 The TikTok Ad Network has arrived. We’ll see how this plays out.
🔥 JOB: Growth Planning & Analytics Associate // David Protein // NYC: join the world’s hottest protein company on-site in NYC.
Love this newsletter? Please forward it to your friends, coworkers, or Kelly Clarkson. Subscribe here to get this data in your inbox every week.