How VC scouts source startups: a working method before you have a famous network

How VC scouts source startups: a working method before you have a famous network

The scout loop is short and it repeats: get a clear brief from the fund, pick two or three places where the founders you want already gather, become useful there, reach out to specific founders for specific reasons, qualify what you find against the brief, write a short memo, and ask the fund what you got wrong. Do that for a quarter and you have a sourcing practice. Skip the brief or the feedback and you have a hobby.

Nothing here requires an existing reputation. It requires you to be specific about what you are looking for and honest about what you are. The templates, checklists, cadence and plan below are editorial guidance from us, not rules any fund published.

Start with the fund's sourcing brief

Before you look at a single company, write one page you can show the fund and ask them to correct. Editorial guidance for what belongs on it: the thesis in a sentence, the stage and round size the fund actually writes, the geographies it can invest in, sectors and business models that are out of bounds, what the fund considers too early, whether it will look at companies already raising, and how fast it can respond.

Then write the part that protects you: exactly what you may say about your relationship with the fund. Do not describe yourself as investing on the fund's behalf, or as able to commit capital, unless the fund has told you in writing that you can.

Ben Casnocha of Village Global, writing in October 2019, described scouts as people who invest money in startups on behalf of a fund, sometimes with full decision-making autonomy, which tells you the arrangement varies by program and is a question you must ask rather than assume. Anything touching compensation, referral fees, conflicts or investment authority needs review by counsel in your jurisdiction, not a template from the internet.

Map the places where relevant founders already gather

Coverage is the job. Hunter Walk of Homebrew argued in April 2017 that seeing the best opportunities is a stronger indicator of future fund success than winning the ones already in front of you, and that much of the most interesting seed activity is dark to most investors, either competitively privileged to a few firms or so far outside the usual pattern that the path is random. You cannot fix that with volume. You fix it by being present where a specific kind of founder already is.

Build the channel map as a short list, not a spreadsheet of everyone. Editorial guidance on where to look: the professional community you already belong to, your university and its research groups, operators at companies in your sector, founders you have worked with, sector-specific groups and newsletters, public launch channels, and recurring events.

Structured events are one channel with published rules. Y Combinator's Demo Day is invitation-only and, per YC's Demo Day FAQ, invitations are software generated based on recent investment history in YC startups, are not transferable to someone else, and investors who are not invited may apply directly through the official route.

YC also notes that companies launch throughout the batch on its Startup Directory and Launch YC, so most of a batch is visible without an invitation. Access rules like these change, so check the current page before you plan around it.

Do not buy lists and do not run mass scraping. Both produce contacts you cannot explain to the fund, and neither survives the first question a founder asks you.

Build signals before you build a famous network

If nobody knows what you look for, nobody sends you anything. Walk's own account of missed deals names this: one of the two main causes was founders not knowing the firm cared about that area, and his answer was to be more open and learn in public about areas of active interest, even ones without a current investment. The same move works at your scale, and it costs nothing but consistency.

The clearest sourced example is a newsletter writer who announced in February 2025 that they were joining a16z as a scout. By their own account, the role came from covering more than 100 startups in a consumer newsletter and from a relationship with an a16z partner that started in 2020 with sharing promising startups. That is one person's path, not a job spec, and it took years rather than weeks.

Allied Venture Partners, writing about how to become a scout, points the same direction in general terms: build the profile through networking, creating industry content, and angel investing where you can. Pick the one of those you can sustain weekly and let the other two wait.

Run respectful outbound

Editorial guidance for the message. Say who you are and what your actual relationship with the fund is, in one line, with no implied authority. Say the specific reason you are writing to this founder, which should be about their work rather than about your process. Ask for one thing, usually a short call. Make declining easy and cost-free, and mean it. Then log the outcome whether or not they reply.

Two things to avoid. Do not present yourself as the fund. Do not send the same paragraph to fifty companies; if the reason you are writing would fit any of them, it is not a reason.

Qualify before you send

Sending everything is how a scout gets ignored. Run each company through a checklist before it reaches the fund. Our editorial checklist: thesis fit against the brief, stage and round timing, what evidence exists about the founders and their history with the problem, what evidence exists about demand, whether the company is already raising and from whom, how you came across it and how warm the context is, any conflict on your side, and the two or three things you genuinely do not know yet.

That last item matters more than it looks. A submission that names its open questions is more useful than one that hides them, and it is how a fund learns to trust your filter.

Write a useful scout submission

Keep the memo to one screen. Editorial structure: what the company does in one sentence a non-specialist can follow; why now, meaning what changed that makes this possible or urgent; why this fund, tied to the brief; how you found it and your relationship to the founder; the evidence you have; the risks you can already see; your open questions; and the specific next action you are asking for, usually an introduction or a pass.

Do not write an investment recommendation you cannot support. Your job is to make the fund's decision faster and better informed, not to make the decision for it.

Close the feedback loop every week

This cadence is editorial guidance, not an industry benchmark. Once a week, look at four numbers you keep yourself: founders you spoke with, companies you qualified, submissions you sent, and responses you got back. Then read the passes. Ask the fund which submissions were off-thesis, which were right but too early, and which they had already seen, because those three failures have three different fixes.

Track your own misses the way Walk describes tracking them at Homebrew, as rigorously as the rest of the funnel: companies in your area that raised without you seeing them, sorted by whether you never saw them or saw them and passed. Coverage problems and judgment problems look identical until you separate them.

Keep the relationships warm between cycles. Affinity's 2025 benchmark report, built on aggregated data from almost 3,000 venture firms across 68 countries, reports that the top-performing firms in its 2024 benchmark worked on more than twice as many deals as the wider set. That is a platform observation about firms rather than scouts, and it is not a target for you, but the direction is a fair prompt: sustained activity in a defined area beats bursts.

A 30-day practice plan

Editorial guidance, sized for someone doing this alongside a job.

Week one: write the sourcing brief and get the fund to correct it. Write down the exact sentence you will use to describe your role.

Week two: pick two or three channels, no more, and show up in each. Start the log that will hold every company, source, date and outcome.

Week three: run ten pieces of specific outreach against the brief, and publish one useful thing about the area you are exploring.

Week four: qualify what you found, send two or three memos, and ask for feedback on each. Write down what you learned about the fund's filter, then start again with a sharper brief.

Start your scout profile to set up a profile, a deal log and a memo habit in one place. If you are still deciding what the role involves, read the scout role guide, compare structures in the VC scout programs guide, and browse the current program index.

Sources: Hunter Walk, "Seed investing: dark dealflow," April 24, 2017; Affinity, venture capital benchmark report, 2025 edition; Y Combinator Demo Day and Demo Day FAQ, accessed 2026; Ben Casnocha, "Venture capital scout programs: FAQs," October 2019; Consumer Startups, "I am joining a16z as a venture scout," February 25, 2025; Allied Venture Partners, "How to become a venture capital scout." Program and access rules are as of those dates and can change.