Special: 2016 Review and 2017 Predictions

The episode opens not with a company but with a method. Gilbert and Rosenthal announce that instead of the usual single-acquisition format they have gone back through the archive, pulled out the technology themes named in each installment, and tallied them, so that the year-end review rests on their own corpus rather than on recollection. Rosenthal supplies the sample size and the timing: twenty-three full episodes recorded in 2016, this being the twenty-fourth, taped on December 30 with what he calls borrowed editing heroics, and therefore likely published after the calendar has already turned.

The cold open is a small joke about the state of Twitter — Dorsey reportedly thinking hard about an edit button, which the hosts treat as a symptom rather than a feature — and it establishes the tone for the rest: skeptical, informal, and willing to editorialize. That the first substantive claim of the episode is a tally, and the first prediction is framed as something the audience can check later, is the structural promise the rest of the hour tries to keep.

Rosenthal is explicit that the predictions carry no guarantee, and both hosts frame the review as an attempt to establish a baseline before moving to 2017. Gilbert adds the practical reason for the compression: the 2016 themes are handled quickly so that the bulk of the episode can go to forward-looking material, with an extended Carve Out section at the end for the year's best purchases, reads, listens and viewing.

The 2016 Theme of the Year is Ben Thompson's Aggregation Theory, which Gilbert says came up in six separate episodes — Acompli, Snapchat, Jet, Android, Skype and Marvel. He states the theory in one sentence: in a world of infinitely accessible, zero-cost distribution, building a superior user experience is the winning strategy. The interesting claim is not the definition but the epistemics: once the lens is applied, it cannot be switched off, and every subsequent company analysis becomes an exercise in spotting the same mechanism.

Rosenthal credits Thompson directly and adds a nuance about the idea's lineage — the canonical post appeared in 2015, but the concept keeps resurfacing and being refined across Thompson's earlier work. His restatement sharpens the mechanism: when distribution costs fall to zero in digital marketplaces, the addressable market is the entire world, competition becomes global and near-perfect, and the best experience rises to the top.

The most concrete evidence offered is professional rather than analytical. Rosenthal says the lens has changed how he evaluates startups as an investor, and Gilbert describes a talk he gave at a product conference called Industry, where he realized mid-recording that his prepared material had to become an argument about network effects. Gilbert's account of the audience is a useful detail: taking a venture lens to product managers, and arguing that virality and dependencies on the rest of the user base belong in the core product rather than in the afterthought category.

A short exchange about whether the talk is online — it is not yet, though the conference was about to start releasing recordings — functions as a small illustration of the show's own distribution problem, and the hosts promise to tell the audience when it is available.

The four-time theme is network effects, raised in the Virgin America, PayPal, LinkedIn and Adobe episodes. Rosenthal's contribution is a scarcity claim: true strong-form network effects are rare, and among all the companies covered only Facebook, Snapchat and LinkedIn clearly qualify, with others sitting at varying distances from that bar.

Gilbert then builds a spectrum rather than a binary. At one end is an app like MyFitnessPal, where the value is almost entirely utility — tracking macronutrients, calories and a food catalogue — and the social layer is a secondary motivator. Instagram is the crossover case: filters were powerful utility independent of any network, and the network effect arrived later. At the far end sits the telephone, which has no utility except through the network, and which has proved correspondingly durable. His conclusion is a rule of thumb: the closer the network effect sits to the core of the product, the more staying power the technology has.