One slide per fund, eight rows in the same order on every slide so the themes read across. Every line comes from the Hyrax investor marketing playbooks dossier of 2026-09-11 and nowhere else. Greyed rows mean that fund published nothing on that row.
| Fund | Bracket | What they publish |
|---|---|---|
| Andreessen Horowitz | Undisputed core | Complete, usable playbook |
| Sequoia Capital | Undisputed core | Complete, partial playbook. The Arc go-to-market module is unpublished |
| Accel | Undisputed core | Complete, richest single fund in the set |
| Benchmark | Undisputed core | Publishes nothing. The entire website is two office addresses |
| Founders Fund | Undisputed core | Publishes nothing on marketing |
| First Round Capital | Early-stage specialist | Complete, date-verified, three items cut |
| Bessemer Venture Partners | Early-stage specialist | Complete, one item cut and one stage-flagged |
| Greylock Partners | Early-stage specialist | Complete, thinnest fund in the set, four usable items |
| General Catalyst | Early-stage specialist | Complete, one 2024 series, dense enough for a slide |
| Union Square Ventures | Early-stage specialist | One citable post, otherwise nothing |
| Kleiner Perkins | Historically top tier | Publishes nothing usable. Two 2016 panel videos, no transcript |
| Y Combinator | Dominant seed brand | Complete, date-verified, three items cut and replaced |
| Conviction | Sector specialist, AI | Complete, usable playbook, all sources post-2022 |
| Index Ventures | Not on the list | Included because it publishes the most data-backed marketing chapter found anywhere |
| Sequence | Simple use case first. Acquire users the way a consumer company would. Iterate until the engine works. Only then layer sales behind it. |
| Who to hire | A product marketing generalist, equally comfortable with product, pricing, placement and promotion as with events, social, content and sales enablement. |
| Channels | Rich content, a great product, and conversations direct from the company. Early pipeline comes from founders, open source communities, the founders' network, the investor network, events and prospecting. |
| Mechanics | Core messaging comes from the founders. Run a small number of paid tests and do not be discouraged if it lands flat. |
| Tools | Not published |
| Metrics in priority order | CARR first. Then live ARR, net new ARR, net dollar retention, gross dollar retention, net new logos, new logo ACV, CAC payback, quota attainment, net burn. CAC payback 12 to 18 months average, 18 to 24 enterprise, 6 to 12 small business. |
| Money | Paid marketing gets a relatively small budget for testing only. No dollar figure published. |
| What not to do | No key performance indicators on marketing qualified leads. Never outsource demand generation. Never hire a firm for messaging or positioning. Do not hire communications first. Do not lock in and scale a motion too early. |
| Sequence | Three archetypes. Hair on Fire means the demand is obvious and you win on being different. Hard Fact means you educate the market first, then capture it. Future Vision means finding commercial pit stops. |
| Who to hire | Marketing does the four Ps: positioning, promotion, price, place. Do not hire a sales leader before the model is set. |
| Channels | Not published. The Arc go-to-market module is described in one sentence. |
| Mechanics | Fifty or more customer conversations in a week. Demo early and demo often. Provide something of value at every meeting. HubSpot coined a term for the new way and wrote the book on it. |
| Tools | Not published |
| Metrics in priority order | Engaged and retained users with low churn. A high ratio of engaged free users upgrading to paid. A high hit rate on cold outreach, because warm leads are predisposed to be nice. |
| Money | Frugality. Focus spending on what is critical. Start with only a little money, because it forces discipline and focus. |
| What not to do | Do not be pushy or force a yes or no. Do not put up market-study numbers with no detail behind them. Self-serve needs an enormous base of dedicated users first, so it is not a day-one plan. |
| Sequence | Two phases. Phase one is the viral moment that draws interesting people in. Phase two is outbound and paid campaigns against the customer profile that phase one revealed. |
| Who to hire | One growth marketer running experiments before any sales or marketing leader. Screen on the message, not on tactics or growth hacks. Paid work trial on a real campaign prompt, judged on writing. |
| Channels | Word of mouth is the channel for this generation. Coding assistants are replacing search as the recommendation surface, and they favour the best developer experience. Launch weeks every three months. Community. Documentation. |
| Mechanics | Put something shareable in the product: accessible, fun, shareable. Ship what the community asked for, then say publicly that you did. Ask investors for named portfolio introductions. Put a captcha on the signup form. |
| Tools | G2 and Capterra to read the category. ZoomInfo, LinkedIn and SurveyMonkey for cheap discovery. GLG and Tegus for expert calls. Profound for visibility inside generated answers. |
| Metrics in priority order | Daily over monthly users above 40% is elite, 30% is the 90th percentile, 14% is the software average. Organic inbound above 30% of pipeline is excellent. |
| Money | $2,000 per engineer a year into open source, counted as brand marketing. Linear spent $35,000 across the company's entire life. $1.50 of spend per incremental $1 of ARR is achievable. |
| What not to do | Do not fixate on growth rate, because it can be bought. Do not scale revenue by scaling headcount. Do not run outbound before the target is known. Do not let large customers pull attention off the community. Do not optimise for repository stars. |
| Sequence | Company purpose, then product positioning, then personality. Positioning sits on an 18-month horizon, pre-launch to Series A. |
| Who to hire | A generalist with breadth across areas and depth in one, starting from strength in either demand generation or product marketing. Ship a real launch inside the first 30 days. |
| Channels | Under $1M of annual paid spend, all of it goes to performance and none to brand. |
| Mechanics | Start paid search at $100 a day. Run to 50 conversions so the platform has enough data. Then scale from that baseline. Write the positioning against five questions and take a position that repels some people. |
| Tools | Not published |
| Metrics in priority order | Fully loaded and correctly attributed cost to acquire a customer. |
| Money | Under $1M of paid spend: 100% performance, 0% brand. At $5M: $1.5M brand and $3.5M performance. The only dollar-level split published by any fund. |
| What not to do | Do not ship the pitch deck to the website. AI-powered is not a position. Never blend paid and organic acquisition into one cost figure, because it destroys credibility. |
| Sequence | Build the middle of the funnel before the top. Early-stage companies have essentially no middle of the funnel, and traffic with nothing to convert into is waste. |
| Who to hire | One strong generalist growth marketer, not specialists. A marketing expert in place by $4M of revenue. |
| Channels | Six middle-of-funnel channels: the website, long-form content, events, LinkedIn, earned media and communities, and email. |
| Mechanics | One content capsule: two to three pieces in each of five categories, each mapped to a format. Repurpose every asset seven to twelve times. Build the moat from proprietary data. Make the target customer so narrow it almost feels too small, defined on pains, gains, shifts, blockers and motivators. |
| Tools | Category reference only. No named products. |
| Metrics in priority order | Revenue growth 75% good, 100% better, 125% best. CAC payback 12 to 18 months good, 6 to 12 better, 0 to 6 best. Net revenue retention 100, 110, 120%. Logo retention 85, 90, 95%. Shooting Star profile: $3M year-one ARR, 60% gross margin, $164K ARR per employee. |
| Money | Judge spend against $8 to $10 of pipeline per $1. Their own subtitle puts that at $10M to $25M of ARR, so it is the standard to build toward and not the one we are held to now. |
| What not to do | Do not drive traffic into a funnel with no middle. Do not chase the fastest growth path, because it comes with fragile retention and 25% margins. |
| Sequence | Customers one to three come from the personal network, with basically no counter examples. Four to ten come from things that do not scale. Ten to fifty is where prospecting tools and sequences start to work. |
| Who to hire | Not published |
| Channels | The personal network, direct messages on forums, flying out to customers, a dinner for six prospects, free consulting sessions. |
| Mechanics | Keep outreach under 75 words. One explicit call to action. Read it aloud and cut any line that sounds machine-written. Follow up three or four times across two weeks. Trade 20 minutes of your own preparation for 30 minutes of their time. |
| Tools | Apollo, whose free plan is enough to build a first list. Clay once you need to qualify on a specific signal. LinkedIn Premium for current professional data. |
| Metrics in priority order | About 50% of connection requests accepted and 20% of those converted to calls. About 30% acceptance when paying professionals $100 to $200 an hour for product feedback. Around 30% of seed companies raise a Series A. |
| Money | Not published |
| What not to do | Do not offer a free trial in B2B. Use a money-back guarantee at 30 or 60 days, or let them exit an annual contract by paying one month. |
| Sequence | Start the self-serve motion and the sales motion at the same time rather than in sequence, because timing always matters. |
| Who to hire | Reject the profile that does months of pre-work and organisation building before doing work with impact, and stalls waiting for direction. |
| Channels | Community as a churn intervention rather than a brand exercise. Discord and X. Partner marketing, treated as a sales problem rather than a marketing one. |
| Mechanics | Name the 50 accounts leading the category and recruit them as investors or customers. Bring power users onto live streams to demonstrate their workflows, then push what they do back into the product. |
| Tools | Not published |
| Metrics in priority order | Trial revenue churning at 30% a month is novelty, not demand. Churn of 60 to 70% for AI products that skip community education. |
| Money | Not published |
| What not to do | Do not build another generic chat interface, because the prompt is a bug and not a feature. Hiring outbound reps and account executives is the least efficient distribution method available. |
| Sequence | Early bets are the self-serve product, pricing and packaging, and viral sharing mechanisms. Optimising signup and onboarding flows comes later. |
| Who to hire | A pod of designers, product managers, analysts and growth marketers, once the motion is running rather than before. |
| Channels | Developer tools suit a bottoms-up motion, because developers want to explore on their own. Invest in developer communities for word of mouth. |
| Mechanics | Six experiments with measured results. A contextual signup button gave a 2.5x lift in conversions. Highlighting gated premium features improved retention by over 20%. Lifting the one-member limit on the free plan more than doubled collaborative workspaces. |
| Tools | Not published |
| Metrics in priority order | Five categories in order: conversion, distribution, virality, customer value, efficiency. Efficiency comes last and gets added as the go-to-market strategy matures. |
| Money | Not published |
| What not to do | Do not copy another company's tactics as a prescription. Do not choose the motion because investors mentioned it or because it worked elsewhere. |
| Sequence | The website is the forcing function for positioning and brand. Close at least five customers who are not design partners before hiring a seller. |
| Who to hire | Do not accumulate specialists. Hire generalists who understand how the work ties into the product and contract the specialists. First communications hire post-Series B at the earliest. |
| Channels | Earned, owned and paid. Pick one owned channel and hold a real cadence. Comparison pages rather than explainer pages. Reddit, run by real disclosed humans. |
| Mechanics | Build one free diagnostic that shows the buyer their own problem. Qualify budget with a small number, because if they cannot approve $3,000 they will not buy a $100,000 product. Design partners who almost sell the product back to you are the validation signal. |
| Tools | Do not buy software that scores your own visibility inside generated answers. Use brand trackers instead. |
| Metrics in priority order | Net promoter score is a better indicator than domain authority. A healthy business of this kind keeps 95% or more of its customers through the year. They publish no numeric Series A bar and refuse to. |
| Money | Do not trade organic against paid as though they compete for the same money. Spend as much on paid as is affordable while payback works. |
| What not to do | Do not cold pitch reporters, especially alongside 50 others. Do not announce the first funding round by default, because funding is not the news. Do not base positioning on a competitive analysis. Do not build a three to six month growth roadmap. |
| Sequence | Pick one core channel, not several. Three months maximum to find scalable traction. Do not diversify until the core channel plateaus. |
| Who to hire | Someone in a growth or marketing role inside the first ten employees. 51% of highly successful startups do this, 41% in software. Early-to-mid career, creative plus analytical, working alongside engineers. |
| Channels | At the earliest stage the realistic choice is content or pay-per-click. Their own examples of single core channels: Pinterest on Facebook, Airbnb on Craigslist then search, Squarespace on podcasts. |
| Mechanics | Contract a specialist for each channel test rather than a generalist, so a failed test is not confused with weak execution. Keep a documented record of failed tests so they are not repeated. |
| Tools | Low-code and no-code tooling for landing pages and copy tests, owned by the marketer. A customer database with lead tagging, scoring jointly owned with sales, and real-time notification. |
| Metrics in priority order | Sales-accepted leads delivered by marketing first. Then pipeline from those leads, as a weighted dollar value and as a share of total pipeline. Win rate and net dollar retention as the lagging proof. Clicks, downloads and subscriber counts are directional only. |
| Money | Once the core channel matures, 70% on the core and 30% on testing. Hire a marketing analyst before $1M of run-rate marketing spend. |
| What not to do | Avoid brand agencies, where $10K a month is on the lowest side and you will still be deprioritised. Do media only if the founder is genuinely good at it. Do not hire a product marketer before there are real user cohorts. |
| Sequence | Buy brand design once, early, then stop. |
| Who to hire | Not published |
| Channels | Not published |
| Mechanics | The entire deliverable is a logo, a marketing website and a style guide you can build the product against. Get it, implement it, put it behind you, and move on. |
| Tools | Not published |
| Metrics in priority order | Not published |
| Money | $0 to $20K maximum, once, with someone competent. The failure pattern is paying $80K to $100K to an agency over three months. |
| What not to do | Do not treat a visual rebrand as a growth lever. It does not move the needle for 99% of companies, and the 1% know they are the exception. |
| Does a self-serve motion work for this category at all | Sequoia's operating partner says enterprise AI demands executive-level engagement and traditional enterprise sales. a16z says in almost every case they have been wrong and there has been a bottoms-up motion. Two Conviction portfolio companies rejected self-serve outright. |
| What settles it, and two funds publish the same rule | Index says stay product-led until average contract value passes $100K and the buyer stops being the user. Greylock gives three zones: under $25K bottoms-up self-serve, $25K to $100K high-velocity inside sales, $100K and above field enterprise. The named failure is sitting between zones with one motion. |
| Where that puts Hyrax | $30 per user a month against 50 to 250 engineers is $18K to $90K a year. The typical account sits below $100K, so by the only published decision rule the product-led motion is correct for this stage. |
| Whether to invest in search content | One fund says B2B software should never have done search marketing and the money belongs in press and community. Another says do it early when creating a category, to own keywords before they go mainstream. A third says the surface is being replaced by assistant recommendation. All three agree on paid search and on comparison pages. |
| Whether to buy visibility software | Accel names Profound for measuring visibility inside generated answers. Greylock says do not buy that category at all, because if your brand is not winning at the prompt you are not winning at brand. |