What they actually built
a16z describes its new media operation as go-direct as a service, and it runs on four motions. In-house creative, producing video and essays on a turnaround measured in weeks.
Owned channels, which now carry a million followers on X, a daily newsletter with a quarter of a million subscribers, a podcast doing a million monthly downloads and a growing Instagram.
A high-touch service team that runs launches end to end and folds traditional press into a direct-to-audience plan. And the network itself, which they describe as a database of who is everybody in tech, activated through a fellowship and an events programme.
The service team is the part worth studying, because it is the one being productised. Portfolio companies get launch support as a defined offering with tiers, from baseline social and copy support up through in-house video, partnered production crews and studio-grade commercial work. They report tens of millions in contract value, doubled hiring pipelines and waitlists coming out of it.
None of this is a side project. a16z frames it as an extension of the bet the firm was founded on in 2009, which was to build a venture firm as a network rather than as a group of solo investors. The media arm is how the network is fed.
The claim about agencies
Alongside all of that, a16z makes a specific competitive claim, and it is worth quoting rather than paraphrasing: they can guarantee distribution, and when you work with other agencies, those agencies cannot promise reach in any real sense.
We are on the receiving end of that sentence and we think it is largely correct. Most agencies in this category cannot promise reach, for a structural reason. An agency that sources creators after the brief arrives is negotiating with strangers against a deadline.
It has no history with those accounts, no measured medians, and therefore no basis for a number. What it offers instead is a plan and an invoice, and the reach is discovered afterwards rather than modelled in advance.
Anyone who quotes you a guaranteed impression figure without showing you the accounts and their median performance is giving you a sales number. That is true of us as well, which is why our proposals name the creators and show the medians before anything is signed.
Two different guarantees
Where the claim gets interesting is what is being guaranteed. a16z can promise distribution because it owns the channels: a million followers on X, a quarter of a million newsletter subscribers, a podcast with real download numbers. That is a genuine asset and it took fifteen years to build.
It reaches a16z's audience. That audience is overwhelmingly founders, investors, operators and people who follow venture as a subject. For a company selling developer tooling to other founders, that is close to a perfect match, and the guarantee is worth a great deal.
For a company selling agent infrastructure to a head of AI at an insurer, the overlap is partial. The buyer may follow a16z, and they certainly follow several accounts that go deep on their specific problem, and those are different accounts.
| Asset | What it guarantees | Whose audience |
|---|---|---|
| Owned channels | Reach on demand, no negotiation | The publisher's, assembled around the publisher |
| Standing creator network | A modelled range from measured medians | Many, each assembled around a subject |
| Per-campaign sourcing | Nothing in advance | Unknown until people reply |
Both of the first two are real. They guarantee reach into different rooms, and the right one depends entirely on who has to see the thing. The third is what a16z is describing, and they are right about it.
Which motions to build and which to rent
The useful question for a company without a16z on its cap table is which of the four motions to assemble internally and which to bring in. They separate cleanly.
Build the narrative
Nobody outside the company can produce its story faster or more credibly than the people who built the product. A team that has lived inside a problem for two years understands it better than any agency will after a kickoff call.
This is why we ask clients for a brief rather than finished copy, and the reasoning is in launch narrative design.
Build the founder channel
Personal accounts outperform brand accounts on reach and on trust, and the person has to be someone who will still be there in five years. This is the motion that compounds and it cannot be outsourced without becoming obvious. Founder-led marketing covers where it breaks.
Rent the launch amplification
Launch weeks are episodic. Staffing for the peak means carrying the cost through every quiet month, which is why most companies under two hundred people end up with three people owning pieces and nobody owning the outcome. Who owns launch distribution works through the headcount maths.
Rent the network
This is the one that takes longest and is most available. Relationships with several hundred creators, with measured performance across many launches, is a multi-year asset. It is also the one thing on this list you can access immediately without owning it, which is the entire argument for creator amplification.
People confer signal
The best line in a16z's own account of this is the one that explains why the whole model works: signal is what matters, and people confer signal, not brands.
That sentence is the case for a creator network stated by somebody with no reason to make it. A brand account announcing its own product is discounted automatically by everyone reading it, which is why a16z put fifteen years into building personal reach for its partners rather than corporate reach for the firm.
The same logic scales past one firm's partners to a network of independent accounts, each carrying their own credibility with their own audience.
a16z built the version that serves the companies it invests in. That is a good model and the portfolio benefits from it enormously. What it leaves open is the same capability for every company that a16z does not back, which is nearly all of them, and which is the business we are in.
Questions we get asked
What is launch as a service?
A productised version of launch support, covering narrative and messaging, creative production, amplification and press, delivered as a repeatable service rather than as ad hoc advice. a16z runs one for its portfolio companies, structured in tiers from baseline social support up to studio-grade production. The same functions exist outside a portfolio, assembled from separate suppliers rather than bundled.
Can a distribution agency actually guarantee reach?
It depends entirely on whether the roster is standing or assembled per campaign. An agency sourcing creators after a brief arrives is negotiating with strangers on a deadline and has no basis for a number.
A standing network that has measured the same accounts across many launches can model a range in advance, because the medians are known before anyone is briefed. The honest form of the promise is a modelled range with the assumptions shown, rather than a single figure.
What should a company build in-house rather than rent?
Build the narrative and the founder channel, because nobody outside the company can produce its story faster or more credibly than the people who built the product. Rent launch amplification and network reach, because both are episodic and both take years to assemble from a standing start. The owned channel sits in between, worth building slowly and continuously rather than around any single launch.