Follow the Crowd or Avoid the Crowd: Holiday Edition

Last week, Amazon announced its latest initiative: Prime Early Access Sale, a shopping event with deals exclusively for Prime members. 

The event occurred on October 11th and 12th, and if you’re wondering, “early access to what?” the answer is: to Amazon’s stockpile of unwanted inventory. Amazon is basically calling on you to help provide some much-needed cash relief in times of soaring interest rates. 

Lucky you. 

Lucky us. 

Of course, this is not exactly how Amazon pitched it. It presents it as a way to jump-start the holiday season, and tries to hint that they are doing this more for you than for them —the deals are described as “some of the season’s most popular and giftable items.” 

Why wait for Black Friday or Cyber Monday, when you can buy in salmon-colored October?

And indeed, Amazon is not alone in trying to convince consumers to buy early:

“‘The shape of the holiday season will look different this year, with early discounting in October pulling up spend that would have occurred around Cyber Week,’ said Patrick Brown, Adobe VP of growth marketing and insights, in a statement. ‘Even though we expect to see single-digit growth online this season, it is notable that consumers have already spent over $590 billion online this year at 8.9% growth, highlighting the resiliency of e-commerce demand.’”

These issues are going to plague the entire sector, as seen in the graph below:

So we should expect to see more announcements coaxing us to “buy now so we have it for the holidays.” 

But, how should we react to such an announcement? Should we believe that this early access is really an opportunity, like Amazon claims, or should we just wait?

An announcement like this has two sides: the firm’s side, with its “private information” on how much inventory it carries, and the customer side, with customers trying to interpret the information and decide on how to act. 

Let’s start with the customer.

The Customer Tradeoff

A customer presented with “announcements” on “early access” to deals, contemplates buying now vs. buying later. Buying now, means they are buying from a limited set of products at a known price, but they are guaranteed to get the products. 

Why wait? 

Because there may be more products later on, and their prices may be even lower at other retail chains. 

Why not wait? 

Because prices may increase, and the products you want may not be available. It is important to note that, while we no longer have major supply chain issues, that doesn't mean that there will be no scarcity. So there is a chance that you will not find the product you want when you want it. 

I know you will use the term FOMO (fear of missing out), and I know that FOMO may seem like an irrational phenomenon (and in many cases, it is), but there are times when it may actually be a very rational perspective. 

The term the operations literature uses to describe this mentality, is “Follow the Crowd” vs. “Avoid the Crowd.” Follow-the-Crowd systems are those in which if you think everybody is going to take a certain action, you should follow. Inventory systems are of this type. If a product has limited inventory, and you think everyone else is going to buy early, you should buy early too. Since everyone is going to make the same calculation, the “rational” behavior here is to buy early. 

Note, this is not the same as “herding.” Herding usually means you observe what other people are doing and decide to follow their actions based on the belief that they had more information than you (e.g., on the quality of a restaurant). A Follow-the-Crowd situation is not herding since everyone has the same information, and you can’t really observe what other customers are doing. You merely assume that if other customers buy the product, you will be left with nothing. 

Now, among other uncertainties, it’s also unclear whether these items are indeed limited, but the idea is clear: Amazon is trying to create a (rational) FOMO, and essentially change the equilibrium from one where everyone is waiting, to one where everyone is buying. 

Note that not all systems are Follow-the-Crowd type. Congestion-based systems are Avoid-the-Crowd type. For example, if everyone goes to lunch at noon, you should avoid it. You can go earlier or later, but you shouldn’t go at the same time. 

The Language of Announcing Deals

But let’s delve deeper into the “language” that evolves between a retailer and their customer. I already discussed some of it in my post on shelf and stockout signs, but this is a slightly different angle. 

Imagine a world where a firm has private information (in this case, on how much inventory it has), and the customer is the one acting (buying or not). Both face some residual uncertainty. 

For this language to be viable, there needs to be some mapping from the state of the system to the “message space,” and respective mapping from the set of announcements to the set of actions. 

This sounds a little too mathematical, so here’s what it actually means: there must be two different signals that elicit two different behaviors. For example, when a firm mentions an early access promotion, consumers should buy immediately in fear of running out. The signal is “Early Access,” the action is “Buy.” 

This becomes a language when a firm uses each signal when its best for itself, and the consumer buys when they infer that that is indeed the state of the system, and believes that that state (in this case, the amount of inventory left) merits an action (i.e., to buy). A language develops if the same signals elicit the same actions. 

It’s important to note that a firm also faces uncertainty on how much demand it will see when using different signals, but under any demand scenario, interest rate scenario, or inflation, a firm is always better off having customers buy early. 

But regardless of the true underlying state of inventory, Amazon specifically wants to “train” its customers to react by “buying” every time they see the signal “Early Access.” 

So if the question is: “If you think others are going to buy, should you do the same?” Then the answer is yes. If the question is: “If Amazon says you should buy now, should you?” Then the answer is absolutely not!

Why?

Once firms like Amazon establish that you buy in response to these “early access deals,” they will always use them (or some version of them), regardless of whether there is a shortage or not, or whether these are indeed the most giftable products. They will just use them to appeal to your consumerism and convince you to buy more. 

So please resist this early access holiday!

The argument I’m giving here may seem a bit hand-wavy, but in my paper with Achal Basamboo, Buying from the Babbling Newsvendor, we show that in a retail setting, you cannot create a credible language that isn’t just babbling —a language that doesn’t elicit consistent behavior. 

And here is an example of such an attempt, to train customers to always respond to such signals: 

“On April 5, 2012, two consumers filed a proposed class action in federal court in New Jersey against Jos. A. Bank, alleging that the men’s clothing retailer perpetually misrepresented its merchandise as being ‘on sale.’ The lawsuit claims that the advertised ‘sale prices’ were in fact Jos. A. Bank’s ‘regular prices’ and that ‘nearly all’ merchandise was sold at ‘sale price.’” 

If you’re wondering whether language is possible, there is an interesting contrast between Avoid-the-Crowd and Follow-the-Crowd systems.

In an Avoid-the-Crowd system, the firm can share credible information on the length of the queue. Because if the queue is too long, no customer should join, and usually, the firm would like you to avoid joining. When the queue is short, everyone would like to join, and the firm would like you to join as well. 

In a Follow-the-Crowd system, this is not the case since the retailer always wants to create FOMO —they always want to get rid of inventory and sell more. Not a single retailer will say no to more demand and less inventory. 

Do you want another hint as to why this sale was more about selling existing inventory and less about selling the “most giftable products?” 

Amazon didn’t start working with brands on this until July. For most brands, 90 days is just not enough to order products and have them available at the scale needed for an Amazon promotional event. 

In other words, you should do your homework on whether to buy or not. But you should absolutely not follow the recommendations of firms to buy early just because “they said so” through an announcement. 

And how can you be sure that I’m credible? Well, I’m selling you nothing!