“Well, Gentlemen, there’s only two ways I know of to make money: bundling and unbundling…Now excuse us, we’ve got an airplane to catch.” -Jim Barksdale

In the Innovator’s Dilemma, Clay Christensen writes about the cycle of disruptive innovation, in which new technologies or business models enter the market at the low end and eventually displace market incumbents as their offering improves. In time, these disrupters are disrupted by new firms with new architectures.

This article articulates some ideas around disruptive unbundling, and how disruptive innovations can drive a rise and fall in product bundles. We’ll cover:

• What bundles are and when they make sense
• Why bundles, especially bundles of information goods, grow to remain competitive—and how this is the bundle equivalent of sustaining innovation
• What causes bundles to no longer make sense—and the factors that disrupt bundling and drive unbundling
• A few unbundling cycles playing out today

But first: what is a bundle?

A bundle is a collection of goods, services, or features that are sold together as a package. The components of the package could be simple goods—a burger and fries, for instance—or features and benefits like the luxury and transportation features in a car.

There are three bundling strategies: Pure, mixed, and components-only.

Pure bundles are offerings only available as a bundle—you can’t buy the components individually. A buffet is a pure bundle; to get the prime rib at the buffet, you have to pay for the whole buffet—no picking and choosing.

McDonald’s has a mixed bundle: you can buy your burger and fries a la carte, or you can bundle them into a value meal for a discount.

Finally, most restaurants are components only, meaning that there is no bundle. If you buy an appetizer and a pasta, you pay for an appetizer and a pasta. If you just want the pasta, you just pay for the pasta.

But Why Bundle?

If you’re selling something, there are plenty of reasons to bundle.

• Bundling makes it easier to offer “discounts” for goods that are substitutes for one another. “Buy one, get one 50% basketball tickets” is priced that way because the second ticket is worth less to the buyer.
• For certain goods, like entertainment or consumer tech, bundling can create a better product experience since you don’t have to make a purchase decision every time you want to watch something.

The most important reason to bundle, however, is that under the right circumstances, bundling is legal price discrimination. A well-designed bundle allows the seller to extract more profits from the buyer.

How do bundles increase profits?

Bundling Smooths Out Price Differences Between Customers

Suppose Microsoft sells two Office products, Word and Excel, to two kinds of customers, bankers and lawyers. For the sake of this toy example, lets assume that lawyers are willing to pay $100 for Word and $20 for Excel, and that bankers are willing to pay $100 for Excel and $20 for Word.

Willingness to Pay	Lawyer	Banker
Word	$100	$20
Excel	$20	$100

If Microsoft could, it would sell Excel to bankers for $99 and lawyers for $19, and vice versa for Word. It would extract as much possible value from each customer based on their willingness to pay.

This kind of price discrimination—called first-degree price discrimination—would be an excellent profit-maximizer for Microsoft. Unfortunately:

1. Microsoft can’t read minds. Probably.
2. Government authorities may not like companies selling the same products to different people for different prices, especially if lawyers couldn’t arbitrage by buying Excel and selling it to bankers.

But what if Microsoft bundled?

If Microsoft could then sell Word and Excel together in a bundle for $119, they would be taking two kinds of customers with different willingness to pay and turning them into one kind of customers with one bundled willingness to pay. Under the right circumstances, bundling can take people with different price preferences and make them more like each other. And by pricing at $119, Microsoft is able to take the deadweight loss—the wasted opportunity of needing to stick with one price—and capture all of the gains for itself.

Bigger Bundles are Usually More Profitable, So There’s Pressure to Grow

But the real world is messier than two products and two kinds of consumers.

For information products like Microsoft Office or subscription television, as the bundle gets larger and larger, the deadweight loss gets smaller and smaller, and the bundle is able to capture more and more profits. So, a bigger bundle is better able to extract value from the marginal product added than a smaller bundle.

That means:

• Bundles have an incentive to grow and increase price accordingly,
• In fact, that means that bundlers in competitive markets need to become larger because the largest bundlers are best able to extract profits from incremental new additions to the library.

This is the bundler’s version of a sustaining innovation. It exploits an existing market and product architecture in service of higher profits, and will continue as long as the market supports it.

So in theory, bundles grow bigger and bigger, like black holes, consuming everything in their path and increasing prices forever…right?

How Bundles Get Unbundled

Some bundles grow to a certain size, and then rise or fall in an incremental way. Other bundles get disrupted. Let’s explore a few ways this can happen:

A Tentpole Loses Value or Is Disrupted

AOL was how a lot of Americans first got onto the Internet. When you purchased dial-up access from AOL, you also got free access to a basket of other platforms, like email, instant messaging (AIM), chat, news, and over 10,000 “keywords.”

But AOL’s bundle was still anchored by dial-up access to the Internet, and dial-up wasn’t the future. With an irrelevant architecture, AOL had three bundling choices:

• Unbundle the offering, making useful aspects of their offering like AIM or email free with ads,
• Create a mixed bundle, with some features available only with a paid subscription,
• Keep their offering purely bundled, and count on the platform effects from services like AIM, chat, and keywords to sustain their pricing power.

AOL tried to make their bundle black hole even larger by buying Time Warner for $165B. Time Warner’s broadband infrastructure and original content would combine with AOL’s Internet platform to reach into the American home to make their bundled value even more vast. But as dial-up collapsed, no such synergies emerged, and AOL was eventually spun out.

AOL’s noncore services—AIM, AOL Mail, and some of its other offerings—only sustained value because they were available unbundled. Many of these, like AIM and AOL Mail, were eventually outcompeted by Gmail‘s disruptive bundle (gigabytes of free email storage with Google Chat); other offerings were outcompeted by the open Internet.

In media, the loss of a tentpole might be the loss of a core franchise. For programmers like HBO or AMC, hits like Game of Thrones or Mad Men not only create value in their bundles, but it also gives the more opportunities to create new hits.

Richard Plepler, the beloved former head of HBO, said that “our model is building addicts… Whether its Girls, boxing, [Game of Thrones], or Silicon Valley, that’s their obsession.” “Obsession” can be thought of as “a deep willingness to pay for part of the bundle.” If a wildly popular show like Game of Thrones ends, it puts pressure on the rest of the bundle to sustain that “obsession.”

Disruption at the Edges of the Bundle

What if there is no tentpole in the bundle? Let’s go back to our Microsoft Office example.

Our lawyer is considering Excel, and had previously valued it at $20. Now, they’ve discovered Google Sheets, a disruptive substitute that offers the essential features of Excel but without the bells and whistles. Faced with a free, low-end option, our lawyer’s willingness to pay might drop precipitously, perhaps even to $0. If so, the bundle power goes away.

When Platform Effects Go Away

If Word and Excel “play nice” together, then they may be worth more together than as individual parts. If they do, Microsoft has even more ability to capture this “platform value” created for the customer with higher bundle pricing.

But these kinds of platform effects can dissipate and threaten the components of the bundle.

One real-world example: Windows was an indomitable platform. A Windows computer gave you access to a whole world of applications, from productivity to games to Internet browsing. A non-Windows computer didn’t come with as broad of an app ecosystem and was therefore more niche. By developing for non-Windows, you were doing work that could only service a fraction of the market.

When the Internet happened, those constraints were lifted. With HTML and JavaScript, you could write application code once and run it on any device with a modern browser. By 2004, emergent platforms like Facebook didn’t need a beachhead on the operating system at all.

These platform effects haven’t totally gone away—PC gamers still don’t have as many options on macOS, for instance—but Internet-native apps have commoditized other core parts of the bundle. If you can run my most needed apps on any device or in the browser, you have less added bundle value with Windows over any other device or OS.

Demand Becomes Correlated

Bundles can lose value if groups start to value each component the same way. If both lawyers and bankers value Excel and Word at $30 each, there is no profit motive to bundle because there is no deadweight loss.

What Disrupted Bundlers Do

If this sounds similar to the Innovator’s Dilemma, it’s because it is. Successful bundlers have two choices:

• Admit defeat, and launch a lower-margin set of products that disrupt their own businesses. This can mean unbundling gradually. In TV, this includes “skinny bundles” with more limited channel offerings. Dish is doing this with SlingTV.
• Stall for time by launching sustaining innovations that focus on their highest-margin customers, and focus on controlling cost and maximizing revenue. AT&T is doing this with DirecTV.

Platform bundles can also focus on platform lock-in via friendly (creating more value) or unfriendly (increase switching costs) means.

Some Live Examples of Bundling and Unbundling

Today,

• The cable industry has thrived on the “quad-play bundle” of Internet, voice, TV content, and home security. By owning “the pipes” coming into the home, big cable could monopolize video and broadband distribution, and use the bundle to compete for home security, landline voice, and other services like mobile broadband by promising lower prices, better service, and one bill. But:

  • Video began unbundling with services like Netflix and Hulu, and the traditional cable bundle has unbundled via services like Philo and YouTube TV.
  • 5G and satellite Internet may offer a “good enough” solution for home broadband. This disrupts cable’s monopoly and collapses the Internet tentpole that holds up the rest of the bundle.
• Google is vertically integrating more and more of its search. For certain queries associated with (for example) travel, restaurants, certain questions, Google shows its own offerings at the top of the search results.
• Disney bought Fox, and AT&T bought Time Warner, in hopes of launching new bundles. Disney’s control of Disney and Fox assets allows them to do things like launch Disney+ and launch FX on Hulu. Time Warner’s control of CNN, Warner, and HBO allows them to launch HBO Max.
• Netflix is spending ~$15B per year on content to build their bundle. Netflix risks building a vast and expensive network of content, and eventually being disrupted and replaced by something else.
• Spotify is offering exclusive podcast content on top of its commodity music offerings. They’ve paid fortunes to secure exclusive audio content from Joe Rogan, the Obamas, and others, on top of purchasing Gimlet. Apple, its rival in music and the podcast incumbent, has decided to offer premium podcasts available a la carte and not bundle it into Apple Music.
• Facebook made Messenger its own app in 2011. 10 years later, they started rebundling their

Conclusion

Bundles are the ultimate “sustaining innovation,” allowing firms to extract more profit out of an existing basket of products and customers. Bundlers do this successfully because they “smooth out” different customers’ pricing preferences for products in the bundle and capture deadweight loss. Just as many industries face pressure to launch sustaining innovations, bundles will face competitive pressure to grow if they want to remain competitive with other bundlers. These bundles cannot grow forever, and disruptive forces can freeze or reduce bundling. These force can include a collapse in value of parts of the bundle, or a change in differences in how customers value products. Unbundled offerings—or smaller or lower-end bundles—can quickly unravel incumbent bundles. These new products eventually develop into their own ballooning bundles, and the cycle begins anew.

Thanks to Pablo Azar, Ben Edelman, Erik Krasney, Patrick Perini, Chloe Popescu, and Eric Ruiz for reading early drafts and giving helpful notes.