Perhaps the greatest unbundle of all time has been the dismemberment of craigslist. Since its inception, startups have unbundled craigslist niches to create a combined valuation far bigger than craigslist itself. Think Airbnb, Carvana, Tinder, and more.
Unbundling doesn't stop, either. Startups like Sonder aim to unbundle Airbnb, and apps such as Lolly aim to unbundle incumbents such as Tinder. As far as I can tell, it's unbundling all the way down. Greg Isenberg says, "everything that is big enough is falling apart."
Both bundling and unbundling are compelling strategies—but for different companies. Bundling can be viewed as a distribution-driven strategy. For instance, Microsoft shipping its Office software pre-installed on their new machines is a bundle that makes perfect sense. Google bundling a browser with it's search product is a no-brainer. But a startup without established distribution channels lacks this advantage.
Fortunately for disruptors, unbundling is difficult for an incumbent to directly combat. It is often not smart for an incumbent to unbundle themself in response. Craigslist, for example has not unbundled in response to cannibalism-by-unbundling, and neither has Reddit in response to startups such as Goodreads and Discord stealing market share.
AOL began as an all-in-one internet portal. Yahoo then unbundled the content from the internet access. Then Google unbundled the search part of Yahoo. Marc Andreessen points out that you could even argue that Facebook then unbundled "search for people."
Eventually, the unbundler becomes the bundler: Airbnb has added "experiences" to their offerings, and Uber has added Uber Eats to the app.
When evaluating an attempt at unbundling, Marc Andreessen looks for a shift in the underlying technology. He asks himself, "if you were to sit down today with a clean sheet of paper, and you knew that the technology was changing, what would be the proper form of the product, if you were starting from scratch?"
Bundle, or unbundle. There is no alternative.