Behavioral Science and Pricing - Article 3

This is the third article on Behavioral Science and Pricing. The prior article was dedicated solely to expectation bias, as it is one of the most powerful aspects of Behavioral Economics, a subcategory of Behavioral Science , which examines how buyers make decisions and how that affects willingness to buy and willingness to pay. 

But in this article, I will cover additional aspects of Behavioral Science and Pricing.

Compartmental Theory: It has been found that buyers place the memory of their purchase in distinct mental compartments. Furthermore, buyers do not combine the prices they paid across multiple compartments, even if the purchases are entirely related. 

To illustrate this, let me use ultra-low-cost airlines as an example. We are all familiar with ultra-low-cost carriers that nickel-and-dime their customers to the nth degree, while legacy carriers do so to a lesser extent. I checked a couple of routes by comparing the ticket prices of ultra-low-cost carriers. I examined what is included in the ticket price with legacy carriers and the add-ons with ultra-low-cost airlines. Depending on the time of day, in almost all cases, ultra-low-cost airlines, when you factor in realistic add-ons, end up being more expensive than legacy carriers, whose tickets include the same add-ons. 

But here is the kicker. Those who purchase tickets on ultra-low-cost airlines do so because the ticket price is the lowest, but they rarely add all the add-ons. To understand the total price of a trip. Much of this is because the ticket price is remembered in one mental compartment and add-ons in another. Furthermore, timing here also matters greatly. When a ticket is purchased, most buyers only buy the ticket. Add-ons are then added when the traveler checks in, or even at the gate when the passengers are about to board. (I had the misfortune of having several flights in a terminal shared with Spirit Airlines. The amount of aggressive up-sell and promotions by the Spirit gate agents just blew me away. Passengers lined up to buy early boarding, to bring a bag, to get a seat upgrade to an exit row, etc. None of those passengers realized they ended up paying more for their tickets than they would have if they had selected a legacy carrier. I’m sure some went telling their friends, family, or colleagues how cheap a ticket they got…) Despite all these upsells, it did not help the company avoid bankruptcy.) 

So what does this mean if your business is not an airline? Well, it means bundling, unbundling, and considering the purchase journey a buyer goes through when buying a product or service.  

Unbundling: Some buyers always want to pick and choose from a variety of options. But there is something called the Paradox of Choice. This means that if a buyer has too many options, the confusion makes it impossible to decide, and the buyer will not buy. 

In a famous experiment conducted in 2000, researchers set up a table with a large assortment of jams in a California grocery store. 24 varieties to be precise. On alternating days, they set up a table with only 6 different jams. Shoppers could sample the jams and use a coupon for $1 off any jar if they chose to buy.

* The large display attracted more visitors — about 60% of shoppers stopped to look, compared to 40% for the small display.

* But when it came to actual purchases, the small display converted far better: about 30% of those who stopped bought a jar, versus only about 3% for the large display.

The conclusion is that too many options can lead to decision paralysis, increased anxiety about making the "wrong" choice, and lower satisfaction even after a choice is made, whereas a curated, smaller set of options often makes decisions easier and more likely to lead to action.

Other examples:

P&G reduced the number of Head & Shoulders shampoo varieties from 25 to 15 and saw a 10% increase in overall sales volume. 

A cat litter manufacturer eliminated 10 varieties, increasing sales volume by 12% while cutting distribution costs in half, resulting in an 87% increase in profit.

Trader Joe’s is the most profitable supermarket in the US (by sales per square foot of retail space) and carries about 4,000 SKUs, compared to a regular grocery store’s 30,000-50,000 SKUs. 

BMW reduced the number of feature descriptions in an option package from 9 to 5, and sales increased. (The actual package stayed the same.)

So why is this? Behavior science tells us that humans can effectively juggle up to 5 items in our minds. Understand each item and compare them. If we are asked to juggle more than those items, it creates confusion. I already mentioned, in earlier articles on the subject of behavioral science and price, that every purchase is made with reference. If we are given too many choices, we cannot decide which choice is relevant as a reference, and, for that reason, we cannot compare the choices or make a purchase decision. 

Another aspect of too many choices is that, because we cannot make that comparison, we develop a fear of missing out. We think: “Is there a choice I missed that is the best for me at this point?” So we don’t buy.

The seller lost a customer. 

The end result of this unbundling strategy is that companies should unbundle when they can and when it makes common sense. But common sense is not so common, as Voltaire remarked already in 1764 in his  Dictionnaire Philosophique. 

The most effective way to create a unbundling strategy is to understand which features and functions of a product or service can be unbundled, and specifically how these affect willingness to buy (i.e., sales volume) and willingness to pay (i.e., the ability to take price). This is done in willingness-to-pay research and will also tell you what buyers are willing to pay for each of the unbundled features and functions, ie the specific price you can set to maximize sales volume or revenue. These are different price points. Furthermore, you will know the most effective customer targeting and marketing. 

Alternatively, if the products or services are already unbundled, AI, when used by an expert, can find what bundles are actually sold. This, however, does not provide any insights about what prices should be on the components of the unbundle or how marketing or customer targeting may affect sales volume, revenue, or profits. Analyzing sales data requires thousands of transactions within a fairly short period of time, as every market evolves with them.  

Bundling: Some buyers always want a complete solution to their problem. 

So, if unbundling is a way of splitting a product or service into some of its components, thus expanding the addressable market by allowing buyers to select what components are included in a product or service, bundling is obviously the opposite. 

Bundles are created to simplify the purchase decision. To reduce sales friction. See the discussion about the Paradox of Choice above. Many times, the bundle adds a level of convenience to the buyer that they are willing to pay higher prices for. A bundle does not have to be cheaper than the components that go into the bundle. This is a common mistake: many companies assume prices need to be reduced when a bundle is created. But a bundle adds value. Convenience, no need to select from options the buyer has to understand, and often a bundle solves the “whole product” problem.  

Best way to create a bundle:

Bundles can be created from an unbundling exercise, from adding products to other products, services to other services, adding services to a product, or products to a service.

But how to decide what should go into that bundle?

You can use experience and what the corporate memory says makes sense to add to a product or service. This can work but is an imprecise method. Buyers of a product or service may not always be aware of other products or services that could be bundled; thus, their buying choices may not reflect what the market at large likes to see bundled. The corporate memory and experiences are also not necessarily accurate, as staff in a company have certain biases and preferences that will affect what they recall. 

Another method is to use a data mining tool (that may or may not be AI-based) and discover from sales transaction data what products or services are bought simultaneously. This has some of the same limitations as just mentioned above - it is not necessary that a buyer knows all the options available, so they cannot buy multiple products or services to create a bundle for themselves. This method also requires a relatively high number of transactions to create dependable results. 

Finally, bundles can be created by online willingness-to-pay research. (Full disclosure: this is the business of my company.) In such research, all various purchase options can be presented to respondents, and how each option affects willingness-to-buy and willingness-to-pay. It is then possible to create bundles that generate the higher sales volume and support the most profitable prices. Simultaneously, as buyers buy bundles that best fit their needs, customer satisfaction increases too!

Hope this all helps!

Best regards,
Per Sjofors
Founder and CEO