Since the launch of Disney+, I’ve been thinking about the limits of the “unbundle.” Disney+ joins the rank of other OTT providers including Hulu, Netflix, and HBO, which individually average about $10.00 - $15.00 per month. The cost of several OTT providers plus internet sold separately through an ISP can easily add up to a standard cable bill. What is the limit of an average OTT consumer's willingness to pay? To fully answer that question, one must first understand the mechanics of bundling and unbundling. This post will focus on the basics of the bundle.

A bundle is the combination of several products combined & sold as a single unit. The key is that the items are usually sold at a lower price when compared to the a la carte prices. Read more on the formal economic theory behind bundles.

  • Producers bundle in order to capture share of wallet

  • Consumers prefer bundling will choose a bundle because of lower prices and/or lower discovery/search costs

There are many types of bundle formats. This post outlines 9 major bundle types:

  • Pure

  • Mix & Match

  • New or Less Known Product Bundling

  • Up-sell

  • Cross-sell

  • Necessity or Occasional Bundles

  • Old or Excess Inventory

  • Gift Sets

  • BOGO (buy Y quantity get Y)

Real world examples of a bundle:

Bundles are present in every kind of market structure (monopoly, perfect competition, oligopoly, etc) , but manifest in unique ways.

  • Home & Auto insurance - Insurance is a commodity product in a near perfectly competitive market. Home & auto is sold as a bundle so that insurers can reduce customer acquisition cost. A share of that CAC savings is passed on to the consumer. Consumers realize price savings & reduce discovery (who wants to shop for insurance).

  • Traditional Cable Television - The consumer is sold a bundle of cable channels versus purchasing individual channels and paying per channel such as NBC, MTV, ESPN, Lifetime Network a la carte.

  • Microsoft Office - The most basic home version includes the lesser known, lesser used Access and Publisher

Microsoft's Office for home pricing page.

A "good, better, best" pricing strategy is often using in software. Microsoft leads with the "better" product in hopes of maximizing revenue based on a typical bundle of user preferences. In this case, the difference between the good ($69.99/year) and better (99.99/year) is storage. In either case, you are purchasing 7 apps. By bundling 7 apps into the product, Microsoft can justify a higher selling price. Importantly, marginal costs of including those extraneous apps (let's say Access and Publisher) is low as the R&D costs have already been fully realized. In this case, Microsoft has a dominant share of office suite products and has the pricing power to sell customers apps that they don't need.

Industries as Bundles

Over the past ten years, retail has been unbundled into a long tail of unique, highly specialized D2C brands. Traditionally, in the fashion industry, buyers (retailers) would pre-buy for Spring & Fall seasons. The twin catastrophes of the GFC of 08/09 & COVID-19 along with the advent of social media chipped away at the incumbent model. Read more about this in the New York Times Magazine piece Sweatpants Forever: How the Fashion Industry Collapsed

COVID-19 upended human interaction in 2020. Most if not all in-person events scheduled after COVID-19 took hold in North American were cancelled. As the world grappled with the deadly disease, technologists tinkered with platforms that augmented in-person experiences.

In his June 2020 post titled Solving Online Events, Benedict Evans says, “No-one has ever really managed to take a networking event and put it online.” In his usual clinical way, Evans lays out a framework to understand the product-market fit dilemma. First, we must understand the benefit of the displaced technology. Despite being costly and time consuming, events endure because they foster serendipity and facilitate encounters that may lead to new business opportunities or ideas for the attendees. Further, attendees are centralized in one place so everyone can get a lot of business done all at once.

To summarize the key points of the post:

  • Most people attend events for the serendipity of meeting people & having everyone in the same place. Conferences make for an efficient way of doing business. Taking the content & pushing it online is easy, but creating the special, serendipitous moments online is an unmet need.

  • Physical events are time-bound. Essentially, a catalyst for all of the benefits of attending an event. Evans explains how you could attempt to re-create CES, “You can probably convert those scheduled meetings in hotel rooms to video calls - but if you’re going to do a video call, it doesn’t matter where you are or when it is.” While you can build online networking and collaboration software tools, it’s hard to productize a tool that captures the magic of an annual event.

  • Online event platforms look like the virtual malls of the 1990s. “Going online breaks the bundle, and conferences will be the same.” In other words, one cannot pick up one business model and drop it into a different format. Today, ecommerce is driven by a long-tail of diverse brands: cooky, cool, utilitarian, mom & pop. The coolest items purchased online aren’t coming from a “Hot Topic.” We don’t visit an online mall for as we would in real life. The same may be true for the online event model. Evans is doubtful that one platform can effectively reproduce an event online.

In my opinion, events are also an important shared experience that can bond disparate groups of customers and vendors together. Online events shouldn’t even wield the term “event” as they are nothing of the sort. As Evans posits, events may return, albeit in a reduced capacity format, which could reduce demand for these tools. I’m almost certain that there will not be a product that can recreate the magic of a day-long company gathering or an inspiring conference. Nonetheless, the limitations and possibilities for online events are important to discuss as COVID-19 accelerated the trend towards remote and distributed work.

Evan takes an anthropological oriented perspective when attempting to layout the future possibilities. I find that his frameworks are instructive, even for technologies where the writing is written on the wall (ahem, VR).

Another take on bridging the physical to virtual divide:

-a16z podcast: Remote Work and Our New Reality

As you can imagine, a16z is bullish on tools that bridge the vast chasm between physical and virtual:

  • Mentions Tandem (?) - an app that attempts to replicate real time “water cooler” style conversations online. Note how this is being described as a stand-alone app, proving the point of Evans

Discuss virtual events @ 5 minutes:

  • Networking at events can be inefficient and awkward and are biased towards extroverts. Online platforms can serve up information on participants and facilitate efficient matching.Further, online events are much more accessible to introverts & those with no to low budgets.

Before listening to this podcast, I was bearish on virtual events. However, the speakers provide compelling benefits for online events and I do believe there is a market for these tools, but it is hard to imagine a world where online events replicate live events.

The premise of the book, Good Calories, Bad Calories, is that most 20th century nutritional advice is based on "bad science." Science and food writer Gary Taubes, posits that "low fat" diets - en vogue in the latter half of the 20th century - is based on deeply flawed assumptions of data. In the first chapter of the book, he recounts how researchers of the day confused correlation with causation exacerbated by data cherry-picking. Although the concept of correlation is a simple concept, confusing correlation with causation trips up even the brightest minds with the best of intentions.

Here is a useful primer on the Fallacy of Correlation versus Causation

Bad Science

“In 1913, the Russian pathologist Nikolaj Anitschkow reported that he could induce atherosclerotic-type lesions in rabbits by feeding them olive oil and cholesterol. Rabbits, though, are herbivores and would never consume such high-cholesterol diets naturally.”

Rabbits developed the buildup of cholesterol in their tendons and connective tissues, suggesting it was a storage disease - meaning they had no way to metabolize. Despite this obvious flaw in methodology, the authors of the study concluded that high-fat diets caused a 'cholesterol disease of rabbits.'

Life Expectancy - 1900 versus 1950

Source: Our World in Data

In 1900, the average lifespan in the United States ranged from forty to fifty years. By 1950, lifespans had extended on average between sixty and seventy years. While infectious diseases were still (and still are) a major concern, extending lifespans revealed an entirely new set of healthcare challenges.

The first chapter reveals many more egregious errors. Heart disease was declared a major health issue in mid 20th century. Eisenhower's heart disease helped to catapult the disease to the forefront of America's concern. Yes, more Americans were in fact dying from heart attacks. However, there was something else at play. More Americans were dying from chronic diseases, as compared to contagious diseases like cholera and the flu, because they were living longer. Antibiotics and advances in public health extended the average lifespan by decades. However, when culling through mortality data, scientists did not connect the second order effects of living longer.

As I read through this chapter, I was reminded of the incipient nature of science and medicine. In the grand scheme of human existence, medicine & biology are still relatively novel fields. My other blog post on the state of medicine in the early 20th century touches on the infantile nature of American medicine in the early 19th and 20th centuries.