· Ryan Hoover

Ryan Hoover: Growth Flywheels, Building Communities, and LP Investing

Durable products are human-interaction flywheels: build a system where each participant''s engagement recruits and re-engages the next, and defensibility compounds from network and brand long after feature work stops.

growth-flywheelscommunity-buildingproduct-huntlp-investingemerging-manageraudience-firstdistributionai-softwareconsumer-health0% confidence

Why this is in the corpus

Ryan Hoover built the canonical maker-launch community (Product Hunt, from an email list to 250k+ makers/year) and runs an intentionally-small emerging-manager fund (Weekend Fund). The episode is a dense source on growth-flywheel design, audience-vs-community distinction, distribution-as-product in the AI era, and emerging-manager LP strategy.

Summary for skimmers

Product Hunt started as a sub-1000-subscriber collaborative email list built in 20-30 minutes; the growth engine was a maker-launch flywheel amplified by Twitter-graph replication and founder-authored launch retrospectives. Hoover argues community (people connecting with each other) beats audience (a megaphone), that people/network products are more defensible than pure software, and that a small collaborative fund beats a big lead fund for both the manager''s enjoyment and the early founder''s value.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

Direct episode extraction

Best used for

Decision-grade retrieval metadata not yet added for this episode.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

A community is members connecting with each other; an audience is a megaphone

Community is defined by members connecting with and helping each other, not by follower count.

Hoover draws a hard line: Twitter followers are an audience; a community exists when people in the ecosystem connect with and provide value to each other. Communities raise word-of-mouth and retention because people want to stay associated with them (his CrossFit example).

Audit your "community": if the value only flows from you outward, it''s an audience.

“there''s a big difference between an audience and a community and people like mix them up. Like your Twitter followers are probably not a community”Ryan Hoover

Principle

A people/network layer makes a product far more defensible than pure software

Adding a human-network layer to a product creates durability that pure software cannot match.

Hoover notes many built clones of Product Hunt and none worked because it has a brand, a flywheel, and a network that just keeps spinning. He loves multi-party systems (Product Hunt, AngelList) precisely because the people layer resists cloning.

If your moat is only features, assume it''s copyable; a live human network is what endures.

“if you have a people aspect to it, like a network aspect to it, it''s just far more defensible and sustainable I think from like competitors”Ryan Hoover

Principle

Human motivation is the invariant; bet products on what technology won''t change

Anchor product design to durable human motivations, not the current technology wave.

Having worked with Nir Eyal on Hooked, Hoover fixates on psychology because it is the thing that won''t change. The Product Hunt flywheel stays relevant into the AI future because it is ultimately about people gaining status and supporting each other.

When designing for a new tech wave, ask which unchanging human motivation you''re actually serving.

“technology and everything is gonna change dramatically, but like how people think and what they''re motivated by is not gonna change”Ryan Hoover

Principle

Nobody wakes up wanting to join a community — lead with utility

Start a community from a concrete utility people actually want, not from "community" itself.

Hoover compares it to podcasts — nobody wakes up wanting to listen to more podcasts or join another community. You need a hook and a reason to exist; on Product Hunt it was getting users, feedback, and status.

If you can''t name the utility a member gets on day one, you don''t have a community yet.

“people don''t wanna join a community. They''re not like, oh, I''m waking up in the morning. Like, I need to be part of a community”Ryan Hoover

Principle

Reuse behaviors and language users already understand — don''t invent a new verb

Borrow the interaction patterns and vocabulary users already know instead of inventing new ones.

Product Hunt deliberately looked at Reddit and adopted upvotes because people already understood them, and chose "makers" as a flat, inclusive label. Hoover warns people get "too cute" with novel language and UX that fails to communicate.

Before inventing new UX language, check what convention your users already fluently understand.

“what behaviors and actions and language do people already understand? Like up votes. People understand what that is like, especially people in tech. So we''re not gonna invent like a new verb, you know?”Ryan Hoover

Principle

Status is a real, exploitable hook — use it as the entry motivation into a community

Status-seeking is a legitimate design hook that draws members into a community before deeper engagement forms.

Hoover is blunt that makers come to Product Hunt to get users, get feedback, and earn status; status is the initial hook, and higher engagement (events, writing) follows once they''re pulled in. He notes high-status people deny caring, but everyone does.

Give new members a visible status win early; deeper community bonds form after the hook lands.

“you''re trying to earn some sort of status, you know, in the status game”Ryan Hoover

Principle

Know your circle of competence and route everything outside it to specialists

Be explicit about your narrow zone of real value and connect people to specialists for the rest.

Hoover raised from 380 mostly-operator LPs precisely so that a founder with a data-science or enterprise-sales question gets connected to an LP who actually knows, rather than getting Hoover''s guess. Self-awareness about competence is the operating model.

Map your genuine competence narrowly, then treat your network as the coverage for everything else.

“my background''s like product and community building and like marketing and like anything outside of that, I''m probably not the best person to ask”Ryan Hoover

Principle

At entry, valuation is the one variable you actually control

Since outcomes are unknowable early, discipline on entry price is the primary lever on returns.

Hoover argues that this early, you don''t really know who''s going to be successful, so entry price is the controllable variable. He cites a friend''s CPG investment at $1M post as a case where a low entry makes even a modest exit a strong return.

When the outcome is a coin flip, entry valuation is where your edge actually lives.

“the only thing you can really control is where you come in at”Ryan Hoover

Principle

Compete against bad, misaligned competitors

Pick categories where the incumbents are structurally misaligned and complacent.

Hoover''s CPG example: the stereotypical incumbent is a big corporation owned by passive index funds and run by managers with little stock who care about hitting the next earnings target for their bonus — versus a founder who owns 60% and wants a $10B company. Misaligned incentives make incumbents beatable.

Score a market by how misaligned the incumbents'' incentives are with building long-term value.

“you wanna compete against bad competitors”Ryan Hoover

Principle

Optimize for what''s sustainable and what you''re suited for, not fund size or status

Choose the game that fits you and is sustainable over the one that maximizes size or status.

Hoover could raise a $200M fund but doesn''t want to, because it would force him to lead deals, use sharp elbows, and travel to schmooze — a game he''s not suited for. He deliberately stays small to play the collaborative game he enjoys and can sustain.

Before scaling into a bigger, higher-status game, check whether it''s one you''re actually suited to sustain.

“I want to do what I want to do. I know it sounds like super selfish, but like also I think it''s sustainable and I also wanna do what I think I''m better suited for”Ryan Hoover

Principle

Design social products as flywheels where each participant''s engagement recruits the next

Build products where more engagement mechanically drives more growth or re-engagement back into the product.

Hoover''s core lens on Product Hunt: makers launch, share with their audience, some visitors sign up, some of those are makers who launch too. Each turn recruits the next turn. He treats every social product as a flywheel and then asks where he can speed the spin or lift conversion.

Ask of any social product: does each engagement recruit or re-engage the next participant?

“how do you build these flywheels where the more people engage, the more it either grows or drives more engagement back to the site”Ryan Hoover

Principle

Only raise venture with a secret or traction — wait for one to be true

Don''t raise venture capital unless you have a secret or traction; wait until one exists.

Hoover, having built and raised before, says he sees many founders chasing spaces with no secret and no traction who "might do fine" but shouldn''t take the venture path. He''d wait for one of the two to be true before raising.

Before raising, ask honestly: do I have a real secret, or real traction? If neither, wait.

“I would only do it if I had like a, a secret, like a secret or traction. And I would wait for one of those to be true first”Ryan Hoover

Principle

Build products, not just meetings — leverage that helps while you sleep

Adding value through software scales beyond your time; meetings do not.

Hoover increasingly invests his time in products (Rolodex, and a new tool) rather than meetings because software helps the portfolio in ways that go well beyond his own hours — value delivered "while you sleep."

Whenever you find yourself giving the same help repeatedly, ask whether software could deliver it while you sleep.

“you can make a piece of software and it can exist while you sleep and help people while you sleep”Ryan Hoover

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

The AI-collapsed build loop: idea to software to feedback to iterate

The build loop (idea to code to feedback to iterate) is collapsing because AI writes software faster and removes the translation step between idea-owner and builder.

Hoover''s diagnostic distinguishes near-term (MVPs and tests, not production-ready, missing security) from long-term (once codegen matches a senior engineer, building this way is smarter). The immediate unlock: a PM can build and test an idea in an afternoon instead of a clumsy Figma prototype.

Ask which of your build steps is pure translation loss — that''s the first thing AI removes.

“I have an idea now I have to translate this idea into a piece of software and then have to get feedback and then iterate on that. And now we''ve collapsed that”Ryan Hoover

Framework

The Product Hunt maker-launch flywheel

A named, replicable loop: maker launches, shares to their audience, some convert to users, some of those are new makers who launch again.

Hoover breaks the loop into a flywheel plus "accoutrements." Once you have the loop you can optimize two levers: speed the spin, or increase conversion at each step (e.g. Twitter-graph emails that re-engage the right people). The diagnostic: at each hop, what fraction advances, and how fast?

Diagnose your loop hop-by-hop: for each step, measure conversion and cycle time, then attack the weakest.

“a maker launches on Product Hunt founder company, they share it with the audience. Some of those people are like coming into the site, they sign up and some of those people are makers too”Ryan Hoover

Framework

The party-host model of community leadership

Lead a community like a party host: curate the environment and welcome people, then let members connect with each other.

Hoover proves the model with Product Hunt''s hundreds of community events — most led by members, he attended only four or five. The diagnostic: is the leader the bottleneck for every interaction (audience), or are members generating value with each other (community)?

If every valuable interaction routes through you, you''re hosting a broadcast, not a community.

“you as a leader, you''re like a really good party host and you welcome everyone in”Ryan Hoover

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

AI in 2025 is "early 2011 in mobile" — the worst it will ever be

AI is at an early-exploration inflection comparable to 2011 mobile, and today is the floor of model capability.

Hoover''s StarCraft "fog of war" metaphor: we''ve discovered new land but can''t see what''s over the ridge. He pairs it with the 2011-mobile analogy (subscription payments unlocking categories) to argue the exploratory phase alone, at today''s capability, is a large opportunity.

If a technology''s capability only improves, the current, undervalued moment is the time to explore.

“we''re still like early 2011 in mobile”Ryan Hoover

“we''re at the worst point that it will ever be right now”Ryan Hoover

Signal

GLP-1s are opening consumers'' eyes to a whole peptide category and self-injection

GLP-1s'' scale is normalizing self-injection and priming consumers for a broader peptide/self-directed-health market.

Hoover, self-experimenting with peptides after knee issues, notes the behavioral unlock — Mind Bloom found patients even prefer self-injected ketamine over lozenges. GLP-1s are the wedge acclimating the mainstream to needles and off-insurance biological "instructions to your body."

A breakout product can normalize a behavior (self-injection) that unlocks an entire adjacent category.

“GLP ones of course are massive and that''s also opening people''s eyes now to like this whole other world of, of Peptides for other types of improvements”Ryan Hoover

Signal

Twitter has become the launchpad for startups, displacing press and even Product Hunt

The startup launch surface has moved TechCrunch to Product Hunt to Twitter, driven by algorithmic reach and video.

Hoover traces the lineage: press was scarce and high-status, then Product Hunt offered a community-led launchpad, and now Twitter''s algorithmic + video shift plus mindshare make it the launch surface. Investor herding on high-like demos accelerates rounds.

Where founders launch shifts with the platforms — right now the demo-with-a-thousand-likes on Twitter is the signal investors chase.

“Twitter in some ways has become the launchpad for companies and startups because it has the tech town hall like Mindshare”Ryan Hoover

Signal

Inference cost is heading from ~$20 toward $1 or less per heavy user

The consensus and near-certain bet is that per-user inference cost falls roughly an order of magnitude.

Hoover calls the cost decline consensus (so "not the most novel insight"), and pushes founders to instead ask what else gets unlocked for free at the foundational layer — e.g. deep-research-quality diligence that was useless a year ago becoming "incredibly useful." Caveat: newest frontier models stay expensive, so cost can bounce back up at the top.

Bake the falling cost curve into your model, and watch what new capability the base layer hands you for free.

“it''s gonna get cheaper and cheaper and cheaper and it''s gonna decrease from like $20 to $1 or less”Ryan Hoover

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Personalized, remixable software — everyone builds their own Evernote

Natural-language coding enables a category where users remix apps into personalized versions instead of buying generic ones.

Hoover is an investor in Adaptive, which lets people remix and change apps with natural language, and flags personalization-of-software as a thesis he hasn''t fully resolved but finds compelling — "if you can speak English then you can code."

When coding becomes speaking, "build your own version" becomes a mass-market product category.

“why do I just use Evernote when like, I want Evernote, but I want it to have these features and so I''m just gonna like build it myself and I''m gonna remix it”Ryan Hoover

Opportunity

CPG is an underpriced entry with weak, misaligned incumbents

Out-of-favor CPG offers low entry valuations and beatable incumbents — a strong risk-adjusted entry.

Hoover contrasts AI (quick adoption but negative gross margins and heavy competition) with CPG (less attractive margins but "nobody else is there"). He cites a friend''s $1M-post CPG entry as a best-in-portfolio bet; a $100M exit there dwarfs hot-category returns.

When everyone crowds the hot category, the underpriced entry is in the one nobody wants.

“CPG maybe like margins less attractive, but like it''s a bad in like nobody else is there”Ryan Hoover

Opportunity

A productized reverse-CRM: "who knows you," not "who do you know"

There''s an unbuilt, more valuable CRM keyed to who follows you rather than who you know.

Hoover built a janky internal tool (Rolodex) that searches his ~200k Twitter followers for portfolio B2B outreach and diligence, and finds it "extremely valuable" despite barely working — evidence a properly productized version is a real, underserved opportunity.

"Who knows you" is a larger, underexploited graph than "who you know."

“a better angle of CRMs is like, who knows you not, who do you know?”Ryan Hoover

Opportunity

Consumer health is a ridiculously large, high-share-of-wallet market

Consumer health is an enormous, uncapped-share-of-wallet market newly workable with cheap data plus AI interpretation.

Hoover cites healthcare at ~20% of GDP, premiums at 15-30% of comp, and health spend as more variable than housing/food. He points to Function Health-style blood panels plus AI interpretation, TrueMed (FSA/HSA for Peloton/AG1), and Mind Bloom as pieces of the opening.

Uncapped, essential, data-rich categories like health reward founders despite the consumer-is-hard heuristic.

“health is like, if you''re not healthy, nothing else matters. It''s like a huge ridiculously large market”Ryan Hoover

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

Selling 15% of the Deel position returned Fund I — and left 2x on the table

One 15% secondary sale of the Deel stake returned the whole fund, crossing into carry but forgoing significant upside.

Hoover framed it as a deliberate trade: LPs weren''t getting distributions, the venture market had shifted since 2021, and returning the fund got them "in the clear" and into carry. He acknowledges it would have been worth at least 2x more held — but felt good about de-risking LPs.

A partial secondary in a breakout can lock LP returns and carry — weigh that certainty against forgone upside explicitly.

“We sold 15% of our position in deal to return the first fund”Ryan Hoover

Lesson

Twitter-graph auth made "just launched" the highest-engagement email

Replicating the Twitter follow-graph let Product Hunt send launch-notification emails that outperformed every other email on open and click rate.

The mechanism: everyone in tech was already on Twitter, so Twitter-only auth reconstructed the social graph and captured emails. "Turner just launched this new thing" emails converted highest because relevance was maximal and launches infrequent.

The best re-engagement email is the one keyed to a relationship the user already has.

“that was one of the highest engaging emails from like an open and click through rate that we ever introduced”Ryan Hoover

Lesson

Weekend Fund I targeted $1M and closed $3M, 3x oversubscribed

Weekend Fund I was 3x oversubscribed against Hoover''s own $1M target, closing at $3M.

He started it in mid-2017 right after Product Hunt was acquired by AngelList, partly to dogfood AngelList''s new fund product. He took no management fees on the $3M (a rounding error) and later was glad, since it left more capital deployed into the portfolio.

Your existing audience and operator credibility are fundraising assets — size the raise to demand, not to your prior.

“I ended up raising 3 million”Ryan Hoover

“deployed it over two years into like 40 something companies”Ryan Hoover

Lesson

Founding through Product Hunt meant "crying inside while smiling" and raising money

The isolation of founding — projecting confidence while privately struggling — is a concrete, recurring cost, not an abstraction.

Hoover says you can''t be honest with your board or team, only maybe a partner, so you''re isolated. He got emotional recounting days of raising money while "crying inside while smiling." His coping: really good friends, and knowing the worst case (getting a job) isn''t that bad.

The confidence you owe your team creates isolation — build a private support system to survive it.

“I''ve had moments in during Product Hunt where I was like, just, just crying inside while smiling and trying to like raise money”Ryan Hoover

Lesson

Rosebud''s power users cost more than their $5/month subscription

Rosebud''s heaviest users were unprofitable at a $5/month price because each session incurred model-inference cost.

Hoover uses it to illustrate that AI founders must bet on where the technology and cost are heading. The consensus (and his) bet: per-call cost drops from ~$20 toward $1 or less, so today''s underwater power users become the profitable base tomorrow.

Model your AI unit economics against the projected cost curve, not today''s inference price.

“their power users were costing them money”Ryan Hoover

Lesson

Product Hunt''s MVP was a 20-30 minute collaborative email list

Product Hunt began as a near-zero-cost collaborative email list that validated demand before any real product was built.

The list never had more than a few hundred subscribers, and Hoover didn''t even curate it — friends co-added links via LinkyDink. Yet random people saying "I read it every morning" gave him enough confidence to build the website with Nathan Bashaw over a Thanksgiving week. It grew into 250,000+ makers launching per year.

Before writing code, test the demand with the cheapest possible artifact — even a shared newsletter.

“that was like the MVP took 20, 30 minutes to set up, shared it on Twitter”Ryan Hoover

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Search your own follower graph to warm-intro portfolio outreach

Outcome: Turn "who knows you" into a searchable database for warm outreach and diligence.

Context: Hoover built Rolodex to search his ~200k Twitter followers. Portfolio companies send a list of target accounts; he searches, finds followers, and DMs warm intros. Also used to find niche experts (e.g. accountants) for diligence. The reverse-CRM angle: who knows you, not who you know.

“build a tool to be able to, to search through all of my like almost 200,000 Twitter followers. Let's say half of them are fake, but still it's like a ton of people who are in tech. And now we can like search and we use it for like portfolio, like B2B outreach.”
— Ryan Hoover

On demand per outreach request per

Scripts

Before you start

· A large, relevant follower base
· A tool to search it
· Portfolio target lists

Ask top makers to write their own launch story and cross-promote it

Outcome: Give makers permission to promote themselves promoting you — authentic, free distribution plus backlinks.

Context: Product Hunt reached out to the daily top ~10 makers, congratulated them, and invited them to write their launch story on their own blog or Medium, cross-promoting on the newsletter and social. Hundreds obliged, spending hours writing, generating authentic promotion and SEO backlinks.

“Do you wanna write about your experience launching and building your product? Like we'll cross promote it on the newsletter. We'll cross promote it on social.”
— Ryan Hoover

Ongoing, per launch day per

Scripts

Before you start

· A leaderboard of top makers
· A newsletter and social channel to offer cross-promotion

Lead with a status/utility hook, then ladder members into higher engagement

Outcome: Use a concrete self-interest hook (users, feedback, status) as the door into a community, then escalate engagement.

Context: Product Hunt's entry hook for makers was utility and status — get users, get feedback, earn standing in the status game. That pulls people in, after which higher-engagement layers like community events (hundreds of them) become viable.

“that's like the hook, let's say that's like, maybe that's how you start being a part of a committee because you're like, oh, I can get users here, let me go check it out. And then that pulls you into hopefully a higher engagement.”
— Ryan Hoover

Hook is immediate; ladder plays out over months per

Before you start

· A real utility or status payoff to offer at entry

Give founders public micro-support (a reply, a like) as low-cost distribution and goodwill

Outcome: Spend your platform on tiny public boosts to founders — near-zero cost, unbounded upside.

Context: Hoover argues that even a single supportive tweet or reply can meaningfully nudge a founder to keep going, while boosting their post via the algorithm. With a platform, the cheapest high-leverage move is public micro-support.

“even if it's just like literally a tweet. Like those can actually, I think, be quite meaningful and give people like a nudge to continue to build or explore or share what they're working on.”
— Ryan Hoover

Ongoing per

Scripts

Before you start

· Any platform/following
· Genuine attention to what people are building

Use Twitter as the only signup method to replicate the follow graph

Outcome: Make signup ride an existing social graph instead of building your own from scratch.

Context: Because the tech audience already lived on Twitter, requiring Twitter auth let Product Hunt know who each maker followed and who followed them, plus their email — the raw material for graph-based notifications that drive the flywheel.

“we use Twitter 'cause we're like, okay, we can replicate the graph. Like everybody on tech is already on Twitter.”
— Ryan Hoover

From launch; sticky once chosen per

Before you start

· A target audience concentrated on one social platform
· OAuth access to that platform's graph

Trigger a "someone you follow just launched" email to spin the engagement flywheel

Outcome: Notify users when someone in their graph launches — it is the highest-engagement email you can send.

Context: Because the Twitter graph was cloned into the product, Product Hunt could email you the moment someone you follow launched. Scarcity of launches plus genuine interest in specific people made it the top-performing email by open and click-through.

“I can get an email when you launch a product to say, Hey Turner, just launch this new thing. And I wanna see that. Like, that was one of the highest engaging emails from like an open and click through rate that we ever introduced.”
— Ryan Hoover

Real-time on launch; launches are ~once a year per maker per

Scripts

Before you start

· A cloned social graph
· Launch/ship detection
· Captured emails

Ship a 20-30 minute collaborative email list as the MVP before building product

Outcome: Test demand with a near-zero-cost email-list MVP before writing any software.

Context: Hoover assembled Product Hunt's first version in Linky Dink, invited a few friends to co-submit links, auto-sent a daily digest, and shared it on Twitter. Sub-1000 subscribers, but organic resharing plus strangers saying they read it every morning gave enough confidence to build the website.

“So that was like the MVP took 20, 30 minutes to set up, shared it on Twitter and then, and that gave me like enough confidence to realize, okay, there's something here to, to explore.”
— Ryan Hoover

20-30 minutes to set up; run for weeks before deciding to build per

Scripts

Before you start

· An off-the-shelf list/newsletter tool
· A few contributor friends
· An existing audience channel (Twitter) to seed it

Let community members run the events, not the founder

Outcome: Scale community events by having members lead them, not the founder.

Context: Product Hunt ran hundreds of community events, most led by members — Hoover attended only four or five. This is the party-host model in practice: set the environment, then let members connect with each other rather than centering yourself.

“we, we hosted some like tons and tons, hundreds of, of community events, which by the way, most of them were led by the committee. Like I didn't go to most of them. I went to like four of them, maybe five.”
— Ryan Hoover

Ongoing once the model is seeded per

Before you start

· A core of members motivated to host
· Light-touch support infrastructure

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

In 2013 Hoover had gone part-time at his job to tinker, wanted a running list of cool new products, and was unsure whether enough demand existed — skeptics said not enough interesting products launched to sustain it.

Did: Instead of building a product, he stood up Product Hunt as a collaborative daily email list in Linky Dink in 20-30 minutes, invited friends to co-submit links, and shared it on Twitter to test demand cheaply.Outcome: Fewer than a thousand subscribers, but organic resharing plus strangers unpromptedly saying they read it every morning gave enough confidence to build the website with Nathan Bashaw over a Thanksgiving week.

Test demand with the cheapest possible artifact — an off-the-shelf email list — before committing to build; watch for organic resharing and unprompted praise as the go signal.

Part of an emerging decision pattern across multiple episodes

Right after AngelList acquired Product Hunt (end of 2016), AngelList was introducing funds and Hoover was still CEO of Product Hunt. There was cultural stigma that a sitting founder-CEO raising a fund was not seriously focused on their company.

Did: Started Weekend Fund in mid-2017 to dogfood AngelList's new fund product while still CEO, targeting ~$1M, taking no management fees, and instituting conflict rules (cannot post companies you have a stake in on Product Hunt).Outcome: Raised $3M, 3x oversubscribed, and deployed into ~40 companies over two years. Fund I ended up with a ~10x fund-returner on paper (Deel).

Dogfooding your own ecosystem and aligning with its culture can overcome the stigma around a side bet; forgoing management fees on a tiny fund keeps more capital deployed and avoids signaling misalignment.

Part of an emerging decision pattern across multiple episodes

About 18-24 months before the interview, the venture market had shifted hard from its 2021 peak. Weekend Fund I LPs had received no distributions, and Hoover had an opportunity to sell part of the breakout Deel position.

Did: Sold 15% of the Deel position to return all of Fund I's capital to LPs, moving the fund into carry rather than holding the full position for more upside.Outcome: Returned the entire first fund and got into carry, so all further gains are paid on carry. In hindsight the held position would be worth at least 2x more today, but the early return de-risked LP relationships and locked in carry.

Returning the fund early versus holding a breakout for more upside is a real tension; certainty of distributions plus getting into carry can be worth leaving upside on the table.

Part of an emerging decision pattern across multiple episodes

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Self-directed health data: scary false alarms vs flying blind

Self-directed AI health tools both catch missed conditions and spiral users into scary false alarms — neither extreme is clearly better.

Hoover recounts asking ChatGPT about eye floaters and haze; it surfaced four possibilities, one of which was see-a-doctor-today-or-go-blind. The tension: the anxiety spiral is real, but so is the risk of having no information at all.

When information both saves and scares, the product win is calibrating how it''s delivered, not whether to deliver it.

“there is an element of like the self-directed hair health stuff, which spirals people into like scary potential issues, but Well if they don''t have that then like the reverse, which is like you have no information then all of a sudden you''re blind”Ryan Hoover

Tension

Return the fund early vs hold the breakout for more upside

Selling to return the fund secures DPI and carry; holding maximizes multiple — the trade turns on certainty vs upside.

Hoover sold 15% of the Deel position ~18-24 months ago to return Fund I. In hindsight the stake would be worth 2x+ more, but he felt good: LPs got distributions, the fund got "in the clear," and everything beyond finally accrues to carry.

Name explicitly what certainty (DPI, carry, de-risked LPs) you''re buying with a partial sale, and price it against the upside you forgo.

“if I held, you know, be worth even more”Ryan Hoover

Tension

Small collaborative fund vs large lead fund

Bigger lead funds win on economics and suit some founders; small collaborative funds often deliver more value to first-round founders.

Hoover concedes a $200M fund is appealing (instant wealth on fees) and that great founders take lead checks, but he''d personally, as a founder, only take a majority lead check from someone exceptional in an area he deeply needed. Otherwise he''d want 20-40 angels/small funds. So he stays small by choice.

Resolve the size question by which game you can sustain and where you truly add value, not by fee maximization.

“a founder in their very first round, theoretically get a lot more value by having, let''s say 20 investors on their cap table for the same amount of equity as one lead investor”Ryan Hoover

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

• growth-strategy
• fund-strategy
• product-design
• market-timing