Erik Torenberg joins Andreessen Horowitz as GP to lead media strategy and podcast network expansion

Summary

Erik Torenberg is joining Andreessen Horowitz as a General Partner, taking on a dual role spanning early-stage investing and building out the firm's podcast and media network. The move grew out of ongoing conversations with Marc Andreessen, who suggested Torenberg bring to a16z what he had been building independently at Turpentine — a bootstrapped network of vertical tech podcasts modeled loosely on the niche B2B media playbook.

The role

Torenberg describes his mandate as a mix of investing and media network development, with podcasting as the immediate priority. A16z already runs flagship shows including the Ben and Marc show and vertical podcasts tied to its funds. Torenberg wants to extend that through an affiliate network — partnering with independent creators, helping them with distribution and monetization, and making it compelling enough that they'd join without needing to. He cites Lenny Richetti, who he says is approaching a million newsletter subscribers and "printing cash," as the model: a Lenny for HR, for CFOs, for every professional vertical.

The Turpentine constraint was always capital — a bootstrapped business has to monetize every show. A16z removes that ceiling, which changes what's worth building. The goal isn't ad revenue scale; it's having the most important conversations in, say, biotech happen on an a16z podcast, then investing in the companies that surface from those conversations.

Media business models

Torenberg is bearish on investing in media companies directly. The power-law outcomes in media went to platforms — YouTube, Spotify, TikTok — not the content businesses riding them. BuzzFeed, Vox, and Upworthy raised venture capital in the 2010s and crashed when the platforms captured the value.

The undertold counterexample is Industry Dive, a collection of niche B2B trade publications covering utilities, HR, pharma, and similar verticals. It sold to Informa for $500 million. The business logic: if you own the leading HR publication with even 5,000 decision-maker readers, advertisers trying to sell to that audience will pay exceptional CPMs because a single closed deal can be worth tens of thousands of dollars. Sean Griffey, Industry Dive's CEO, has been making this case publicly.

On the question of what business models actually work in media, Torenberg argues the best one for this particular context is investing — produce content, build relationships with founders, invest in the companies that come from that flywheel. That's distinct from the "solo GP with a podcast" version of the same idea; at a16z the investing infrastructure already exists.

A16z's media history

Torenberg's read on Future, a16z's earlier owned-media attempt, is that it was a reasonable idea that ran into org complexity as the firm decentralized into separate funds. He thinks there's more institutional buy-in now to try again, though that's not his immediate focus. He flags one format he found compelling: a16z's "everyday carry for founders" series, which simply asked portfolio founders what apps were on their phones. Niche, insider, but genuinely interesting to the right audience.

The No Clip / Secret Tape collaboration on an A16Z Games documentary about Baldur's Gate is the partnership model he points to as working — nearly 500,000 views, driven by matching the right specialist creator with the fund's content goals and letting them run.

Group chat as the next social layer

Torenberg sees group chats as where the most valuable conversations have moved over the past five years, driven by public social becoming too noisy and too politically charged. But existing platforms — WhatsApp, Signal — weren't built for communities of hundreds of people. Notifications fire on every message, identity management is clunky, there's no directory or cross-chat interoperability.

He was building a product in this space before the a16z discussions began — a text-based version of the Clubhouse model, with permission tiers separating speakers from audience. He stopped when his plans changed, but says he has since connected with a team building something in this space that he expects to announce later. His view is that a 10x better product is achievable and that whoever builds it will fill a gap that WhatsApp and Signal structurally cannot.

His broader observation is that social has balkanized — Bluesky, Threads, X — and as people become more comfortable speaking publicly again under a more permissive X, the intensity of small private group chats has dispersed into larger, more diverse ones. He describes a chat where Mark Cuban and others with opposing views debate topics like DEI, with those debates occasionally spilling into public podcasts. That cross-ideological dynamic, he argues, wouldn't have been attempted a few years ago.

Market maps

Torenberg makes a brief case that market maps are underrated despite having become associated with lazy content marketing. His argument: when uncertainty is high — AI at the application layer, for instance — a well-executed map of 50 or 70 companies taking different approaches to legal AI or another vertical creates genuine orientation value. Goldman and Morgan Stanley produce the equivalent in equity research without apology. The execution got sloppy; that doesn't mean the format is broken.