Commoditize your complements: Lower the all-in cost of your product by making its adjacent requirements cheaper. Think Microsoft in relation to PC hardware; Apple in relation to cell carriers; Google in relation to Android.
Come for the tool, stay for the network: On way to build a defensible position is to use a superior tool to acquire customers, then leverage the customer base to offer services no one else can match. Medium is doing this by acquiring writers with a simple tool, which in turn builds an audience, which can then be used to lure more writers. (They've been extremely explicit about this strategy.) Square is doing by getting to iPhone point of sale first, which allowed them to acquire customers and then offer services that are only possible with scale (short-term loans, etc.).
Aggregation theory: Markets have three primary components: Suppliers, distributors, and consumers. The most value will accumulate to companies that can monopolize two of the three pieces, and increasingly that is starting with monopolizing the relationship between distribution and consumers due to superior user experience. Think Netflix, Amazon, Airbnb, etc.
Bundling and unbundling: Waves of disruption in technology can often be explained by this concept: An upstart unbundles an established business by doing a single part of the bundle really well (Google unbundles the Yahoo! homepage by launching superior search), then rebundles many of the same services around this new starting point (Gmail, Adwords, Google News, etc.). The opening to unbundle an established business is usually created due to changes in the technology landscape (discovery became increasingly important as the internet scaled, and Google’s PageRank approach was far superior to the manually curated Yahoo! index—“Yet Another Hierarchically Organized Oracle”).
A dozen things I’ve learned from Charlie Munger about mental models and worldly wisdom: Any one model of how the world works is insufficient. Instead, a heterodox approach that mixes different mental models can product truly sophisticated insights.
Superforecasting: An argument for structured, analytically rigorous thinking about the future without relying heavily on mathematical models. You can get a long way by just starting with simple baseline assumptions about the event you're trying to predict, and then cautiously accommodating specifics. (Excerpt here.)