Venture Capital Deal Sourcing: How Firms Source at Scale

Venture capital deal sourcing is how a firm, as an organization, generates the pipeline of companies it will evaluate. At firm scale, we see it as a staffing and systems problem more than a networking problem. The survey of 885 institutional venture capitalists at 681 firms by Gompers, Gornall, Kaplan and Strebulaev found that only about 10 percent of deals arrive as unsolicited inbound; in our reading, a system somebody built produces the rest.

Definition: Venture capital deal sourcing is the set of channels, relationships, staff and outbound activities a firm uses to generate the pipeline of investment opportunities it evaluates, measured by volume, conversion by channel, and how often the firm is first or lead in the rounds it wins.

The math is simple. Partners can't personally see enough companies. So firms build a channel strategy, hire associates and platform teams, buy a data stack, run scout and fellowship networks, and measure conversion by channel.

This guide is written for firms and fund managers. If you are an individual investor or aspiring VC building your own pipeline, see deal sourcing. For the funnel those opportunities move through, see deal flow.

Pick the channel mix before the tools

Our starting suggestion: decide which channels you will own, and who owns each one, before buying a single database seat. The survey data shows why. Of the deals VCs reported generating:

More than 30 percent came through the investors' professional networks.
Almost 30 percent were proactively self-generated by the firm.
About 20 percent were referred by other investors.
About 8 percent were referred by portfolio companies.
About 10 percent arrived inbound from company management.

Read that as a to-do list. The two largest channels, network and self-generated outbound, take years to build. No vendor sells them.

The rest follow from how you behave. The co-investor referral channel means your reputation among other investors is itself a sourcing asset, earned by sharing deals generously and passing cleanly and quickly. Portfolio referrals come from founders who would recommend you to a friend, which makes post-investment support part of sourcing.

The operational channels behind those percentages are well known: proprietary outbound, warm introductions and referrals, accelerators and incubators, conferences and demo days, online deal platforms, investment banks and M&A advisors at later stages, AI and CRM driven sourcing, and inbound generated by brand. Carta's guide for private fund deal teams lists the same set. Most firms run most of them.

The list isn't the strategy. The weighting is, and the name next to each line.

Size the venture capital deal sourcing funnel on a $50M fund

Before you hire anyone, work out the volume your fund actually needs. Take an illustrative $50M seed fund that plans 30 initial investments over four years, about 8 a year. Applying the survey's stage rates, and its reported average of about 1.7 term sheets negotiated for every deal closed:

Stage	Per year	Per week
Opportunities considered	about 800	about 15
First meetings (25 percent)	about 200	about 4
Partner reviews (8 percent)	about 64	about 1
Due diligence (4 percent)	about 32	fewer than 1
Term sheets (1.7 percent)	about 14	rare
Closed	8	n/a

Now add the labor. The same survey reports an average of about 118 hours of due diligence over 83 days per closed deal, plus roughly 10 reference calls. So 8 closings alone consume close to 950 hours, about 24 person-weeks. Spread 800 screened opportunities across four investment professionals and each carries about 200 a year, roughly 4 a week, on top of diligence, board seats and fundraising.

That arithmetic is why sourcing gets delegated and instrumented. It also suggests a ceiling. A firm generating 3,000 unfiltered leads a year isn't necessarily better sourced than one generating 800 filtered ones. It may just be worse staffed.

Budget for it honestly. Carta's Fund Economics Report 2025 found the median VC fund between $1M and $10M spends about 3.4 percent of fund size on operating expenses in its first five years, while the median fund larger than $100M spends about 1 percent over the same period. Every database seat and scout payout lands inside that number. On a small fund, it's a line LPs will notice.

Staff it: partners, associates, platform and talent

We'd split sourcing across four kinds of people, each with a different job:

Partners work the top of the quality curve: repeat founders, the co-investors who call them first, and the theses that justify a whole category. Low volume, highest hit rate, and the biggest key-person risk.
Associates and analysts own coverage: running the screens, working the databases, and holding the first calls partners don't have time for. This is where a thesis becomes systematic outbound.
Platform teams turn the firm's services into a sourcing magnet. Content, events, founder communities and operator networks all generate inbound, and platform staff usually own the CRM hygiene the whole system depends on. The venture capital platform role guide covers the function in detail.
Talent teams are, in our view, an underrated channel. A recruiter who places engineers at portfolio companies talks to hundreds of operators a year, and operators about to start companies are among the earliest signals available.

One question tests any firm's staffing: what happens if the most connected partner leaves? If the pipeline collapses, the firm arguably has a network, not a sourcing system.

Tool it: the venture capital deal sourcing data stack

A common setup has four connected layers. The value is in the connections, not in any single product.

A CRM as the system of record. Affinity is built for private capital and captures email and calendar activity automatically, so pipeline reporting doesn't depend on anyone remembering to log a meeting. The alternative is a general CRM plus real discipline. Without one of the two, the metrics later in this guide are hard to produce.
Company databases. PitchBook and Crunchbase for funding history and comparables; Harmonic and Specter for early-stage discovery, founder and talent signals, and saved searches that surface companies before they show up in funding data.
Signal monitoring. Product launch sites, code repositories, job boards, app store movement and sector newsletters, piped into alerts tied to the thesis rather than checked by hand.
A public surface. A published thesis, research, a newsletter or a podcast. For firms without a legacy brand, a consistent public presence is often the fastest inbound channel to build.

Two cautions. Tooling changes fast, so confirm any product still exists and still does what you expect before you build a process on it. And a data stack multiplies a thesis. It doesn't supply one. A firm that can't say what it's looking for will likely just receive more of everything. The investment thesis is a sensible place to start.

Extend reach with scouts, fellows and communities

Once the core works, many firms add variable-cost channels. They buy reach into communities your partners can't cover, without another salary.

Scout programs have grown fastest. A firm lets trusted operators and founders write small checks from the firm's capital, or pays per-deal economics, in exchange for early looks. Superscout's index documented 185 VC scout programs as of September 2026. The how to become a venture capital scout guide covers the model from the scout's side.

Fellowship and training programs work on the same principle with a different incentive: a cohort of trained people who bring companies to the firm in exchange for payouts, carry, or the credential itself. That makes them especially attractive to emerging managers.

1752vc runs this channel itself, so we can describe the economics directly. The Venture Fellow program trains each cohort over eight weeks of live virtual sessions on case studies, real pitch materials, due diligence on live companies and deal sourcing, and Fellows are paid on what they bring in: payouts for sourced deals, plus carry on select deals sourced for partner funds. About half of 1752vc's deal flow arrives through that network of 400+ trained Fellows across 20+ cohorts.

Win the deals you find: proprietary versus competitive

Seeing a company is half the job. The firm also has to be chosen.

A proprietary deal is one the firm sees before or instead of competitors, usually because a relationship or a differentiated thesis got it there first. A competitive deal is one several firms are chasing at once. Proprietary deals tend to be better for price and ownership; competitive deals tend to be better for validation and speed.

The competitive end is getting more crowded at the earliest stages. Carta's State of Pre-Seed report for Q2 2026, published in August 2026, found the average pre-seed instrument reached about $276,000, a record over the preceding four years and a 27 percent increase year over year, with similar sums of money going into noticeably fewer deals, and AI companies taking 49 percent of all pre-seed dollars in the first half of 2026. When more capital chases fewer companies, the edge shifts from seeing the deal to seeing it first, and to being the investor the founder picks.

"But cold inbound is a waste of time"

It's a common view, and the numbers seem to back it. An Odin analysis drawing on a VentuRank study by Anthony Richardson found cold approaches were about 64 percent of what firms received but only about 6 percent of what they funded.

But.

The same analysis found the companies sourced cold returned roughly 16 percent more on about 18 percent less capital. Our read: inbound is low-yield because it's unfiltered, not because it's low quality. The fix is fast triage. An ignored inbox is just a channel you're paying to lose.

Decision speed matters as much as discovery speed. A firm that finds a company early but decides slowly, or decides without hearing dissent, gives back the advantage its sourcing created. Sequoia is one example of a firm redesigning the decision to protect that advantage: Julien Bek describes it adding asynchronous written memos alongside live investment committee meetings, and tracing its best investments to a sponsoring partner with strong conviction. Memos serve slow, careful thinking; meetings serve fast debate. Reference calls sit inside the same process, and our guide to how VCs run founder reference checks covers them in detail.

Where we land: deal sourcing is a partner job, not a volume game

Strong sourcing systems tend to be cheap in tools and expensive in partner time. We think firms do better budgeting for that than hoping software will substitute. It's the same instinct behind our take on hunting vs gathering in venture capital.

Two firms show two routes. Sequoia's Julien Bek describes a firm where every partner is expected to hunt for companies, and where partners perform as individuals but win contested deals as a team. That's a sensible answer to the key-person problem above: don't remove individual sourcing, expect it of everyone and pool the effort when it counts. His firm's investment in Citadel Securities, which had never taken outside capital, came from a partner whose relationship with Ken Griffin began when the partner was a student and was kept up through years of persistence. That's the survey's self-generated channel, almost 30 percent of deals, taken to its extreme.

The second route is a clear point of view. USV's Michael Mignano treats a published thesis as a signal to founders who haven't launched yet, telling them which investor to approach first. USV's June 2026 post on what it labels the Rebel Alliance (an ecosystem of competing open-weight and commercial models, agent harnesses, orchestration, memory, identity and payments, rather than one vertically integrated model company) is the example. Content doesn't replace outbound. It makes the right founders arrive already sorted by thesis.

The two routes compete for the same scarce hours. Writing takes partner time away from relationships, and relationship work rarely reaches founders you haven't met. For an emerging manager we'd do both, in that order: a public thesis that makes each outbound message credible, and partners who keep working relationships long before any round exists.

That's our answer, not the answer. A later-stage firm with a strong brand may reasonably lean harder on inbound and banker relationships.

How to measure deal sourcing: the six-number report

If you can't report it, you'll struggle to improve it. One practical bar: each partner can produce these six numbers on request.

Source by deal. Name the channel behind each of the last ten investments. A healthy answer has variety and at least a few genuinely early looks.
Conversion by channel. Opportunities logged, first meetings, partner reviews, diligences and term sheets, split by where the company came from.
First-look and lead rate. What share of closed deals was the firm first into, or led? In our view, this is the most honest test of proprietary sourcing.
Win rate on term sheets. Offers that closed, which measures access rather than judgment.
Downstream quality of passes. Companies the firm declined that later raised from strong leads, by channel. Often the most uncomfortable and most useful number in the pack.
Cost per sourced deal. Database seats, events, platform salaries and scout payouts against deals closed, and whether those costs sit in fund expenses, management fees or the GP's own pocket. The venture capital fund structure guide explains where each one lands.

The bottom line

Deal sourcing at firm scale is less about who you know and more about what you've built: channels with owners, a funnel sized to the fund, and numbers somebody actually reviews.

A network walks out the door with a partner.

A system stays.

Key takeaways

At firm scale, deal sourcing is a staffing and systems problem: only about 10 percent of VC deals arrive as unsolicited inbound, per the Gompers, Gornall, Kaplan and Strebulaev survey of 885 VCs.
More than 30 percent of deals come from professional networks and almost 30 percent are self-generated, so the two biggest channels are built over years rather than bought.
A $50M fund making 8 investments a year needs roughly 800 screened opportunities, 200 first meetings and 32 diligences in our illustrative math, which is why sourcing is split across partners, associates, platform and talent teams.
The data stack is a CRM as system of record plus company databases and signal alerts; it multiplies a thesis rather than replacing one.
Finding a company early can be wasted if the firm decides slowly, so the decision process is part of sourcing.
In our view every partner benefits from sourcing, and a public thesis makes that outreach credible; one way to measure the result is with source by deal, conversion by channel, first-look and lead rates, term sheet win rate, and pass quality.

Frequently asked questions

What is deal sourcing in venture capital?

Deal sourcing is how a VC firm finds the companies it evaluates: partner networks, proactive outbound research, referrals from other investors and portfolio founders, scouts and fellows, accelerators, and inbound generated by the firm's brand. Its output is deal flow, and its quality sets the ceiling on everything the firm can invest in.

How do VC firms source deals at scale?

They divide the work. Partners hold the highest-value relationships, associates run systematic coverage and first calls, platform teams generate inbound through content and community, and scout or fellowship networks extend reach into places partners do not go. A CRM plus company databases ties the channels together and makes conversion measurable.

What software do VC firms use for deal sourcing?

Most firms run a private-capital CRM such as Affinity as the system of record, then layer company databases (PitchBook and Crunchbase for funding history, Harmonic and Specter for early-stage discovery and talent signals) and alerts on launches and hiring. It is worth confirming any tool still exists and still fits before building a process on it.

What is a proprietary deal in venture capital?

A proprietary deal is one a firm sees before or instead of competing funds, usually because of a founder relationship or a differentiated thesis. Proprietary sourcing tends to produce better entry prices and ownership. Truly exclusive deals are rare, so an early look plus the founder's trust is the realistic goal.

How do firms measure deal sourcing performance?

By source per closed deal, conversion rates at each funnel stage split by channel, the share of deals where the firm was first in or led, the win rate on term sheets offered, and the quality of companies it passed on. Cost per sourced deal, including database seats and scout payouts, completes the picture.

Do top VC firms wait for deals to come to them?

Not the ones that describe their process publicly. At Sequoia every partner is expected to hunt for companies, and its Citadel Securities investment came from a relationship one partner kept up for years. Brand brings inbound, but in our view many of the strongest deals still come from relationships, persistence and a clear thesis.

Sources

Disclaimer: This guide is for general education only and is not legal, tax or investment advice. Laws, market data and program terms change, so it may not reflect the latest developments or fit your situation. Treat it as a starting point, not a source of truth, and talk to a qualified lawyer, accountant or financial adviser before you make decisions.