The strangest flex in venture capital is giving the money back. The business is built to accumulate commitments, convert them into ownership and let a handful of improbable companies carry the portfolio. More capital usually means more fees, more influence and a larger claim on the future. In October 2024, CRV went the other way. The firm released $275 million of uncalled commitments from a $500 million late-stage vehicle after deciding that many follow-on rounds could make its returns worse.
Then came the second half of the maneuver. In August 2025, the 55-year-old partnership raised $750 million for Fund XX, its twentieth flagship fund. It was smaller than the $1 billion early-stage vehicle raised three years earlier, contained no new late-stage sidecar and closed in four weeks. Reported demand was roughly twice the amount accepted. In an industry where fund size often stands in for momentum, CRV made narrowness sound like a product feature.
The product is an early decision
CRV sells capital, but that description is as incomplete as calling a newsroom a supplier of ink. Its real product is a decision made when evidence is thin. The firm leads seed and Series A rounds, generally taking a minority stake and a durable role around the company. Founders use the money to hire, finish a product, find customers and discover whether an unusual insight can become a repeatable business. Limited partners supply the fund capital; CRV earns its upside when portfolio companies are acquired or go public.
The customer, therefore, comes in pairs. A founder wants speed, conviction, useful introductions and a board member who will still answer when a launch goes sideways. A limited partner wants the investment team to resist mediocre opportunities, reserve enough capital for genuine breakouts and return more than the public markets could. Those interests usually rhyme. At overheated prices, they can clash. CRV's giveback made that tension visible.
“My sweet spot is post-product, pre-traction.”Saar Gur, CRV general partner
Before the graph learns to brag
CRV's best-known cases are comically early. The firm led DoorDash's first financing in 2013 when the company was nine weeks old. Gur has said he signed Mercury's Series A term sheet when the banking product was two weeks old. CRV partnered with Vercel in 2015, the year it was founded, later joining rounds from B through E. None of these bets arrived with the sort of revenue history that makes an investment committee comfortable.
What was available instead? A founder's peculiar command of a problem. A product that made a small group unusually enthusiastic. A market shift that could be described before it could be measured. In developer tools, that has meant backing the plumbing of modern software: Kong for APIs, Postman for API collaboration, Vercel for web development, Cribl for telemetry and CodeRabbit for AI-assisted review. In consumer and fintech, it has meant looking for new habits: local delivery through DoorDash, startup banking through Mercury and flexible work software through Airtable.
Lead seed and Series A rounds, with reserves for the companies that keep earning conviction.
Pressure-test the founder's insight before conventional traction can settle the argument.
Help recruit executives, reach customers and navigate later financing from a 750-company network.
Stay involved across product revisions, board decisions and the long route toward an exit.
The anti-bloat fund
A venture fund can become too large for its own strategy. If a partnership raises more than it can sensibly place into early rounds, it must write bigger checks, own more companies or drift later. Each solution changes the original product. CRV has confronted that arithmetic twice in public. After the dot-com crash, it cut a $1.2 billion fund to $450 million. In 2024, it released the unused portion of Select Fund II after concluding that expensive later rounds demanded exit values too large to justify.
Selected fund commitments
Fund XX is 25% smaller than Fund XIX and arrives without a new late-stage Select vehicle.
That choice separates CRV from multi-stage platforms built to finance companies from inception through pre-IPO rounds. It also distinguishes the firm from accelerators that trade a small standardized check for a cohort program, and from seed specialists whose smaller funds cannot always support later rounds. CRV sits between those models: large enough to lead and reserve, but now explicitly unwilling to turn scale into a mandate for late-stage deployment.
The trade-off is real. Passing on later rounds can mean surrendering ownership in winners. A focused early fund still needs outlier exits, and a $750 million pool is hardly tiny. The strategy works only if the team can find enough companies early, win competitive rounds and maintain ownership without paying away the return. CRV's advantage is not immunity from venture math. It is five decades of seeing where that math breaks.
A river runs west
The firm began in Boston in 1970 as Charles River Ventures, named for the water between Boston and Cambridge. Co-founder Richard “Rick” Burnes Jr. helped build it during an era when venture investing often meant commercializing technical work from the region's labs. Early portfolios leaned toward infrastructure. The Silicon Valley office opened in 2004; by 2014, the shorter CRV name advertised a center of gravity that had shifted toward the Bay Area and consumer software.
Today the firm lists offices in San Francisco and Palo Alto and invests far beyond either address. The current featured portfolio ranges from security agents and AI code review to satellite platforms, robotics, procurement and healthcare. The apparent sprawl makes more sense when viewed through product behavior. CRV likes tools that remove a technical bottleneck, products that give an individual more agency and founders whose lived knowledge makes a market look different.
Born beside the Charles
Charles River Ventures starts in Boston with a technology-investing mandate.
The westward office
A Silicon Valley presence begins reshaping the portfolio toward software and consumer companies.
DoorDash, Vercel, Mercury
Three early leads become shorthand for CRV's product-first, pre-traction appetite.
Give back, then narrow
CRV releases late-stage capital and raises a smaller flagship for seed and Series A.
No playbook, but a point of view
CRV's public culture is allergic to deal-shop language. The team page says it forms lasting partnerships rather than doing “deals.” Partner biographies read like field notes, full of founders met over dim sum, products observed through portfolio companies and investments that took years to ripen. The pose is deliberately personal: listen closely, say what you think and avoid forcing every company through the same operating manual.
There is a useful contradiction here. “There is no playbook” is itself a kind of playbook. CRV has a repeatable operating shape: lead early, concentrate attention, recruit through networks, help founders reach customers, take board responsibility and reserve capital where conviction deepens. What changes is the company. That distinction matters to founders deciding among Sequoia, Accel, Benchmark, Greylock, Lightspeed, General Catalyst, Bessemer, First Round and dozens of specialist funds. Check size is only the cover charge. The lasting choice is whose judgment enters the room.
The firm's sharpest recent investment was the one it declined to make: more of its limited partners' money at prices it no longer believed.YesPress
What founders can steal
CRV's reset offers a clean exercise for any company. Name the stage where you are unusually useful. Identify the adjacent work that looks like growth but weakens the core. Put numbers on the point where expansion becomes dilution. Then make the constraint legible to customers. CRV did not merely announce a smaller fund; it connected the size to seed ownership, Series A leadership and the decision to stop underwriting late-stage valuations.
For a founder approaching the firm, the practical pitch is equally specific. Show the non-obvious insight, the product response and why this team learned the problem first. Massive revenue is not required at CRV's preferred moment. Evidence of craft is. So is a credible route from a few intense users to a market large enough to return a venture fund. The firm can provide money, pattern memory and access. It cannot manufacture the founder's conviction.
That is where CRV fits in the market: not the earliest check in every case, not a passive index of startups and no longer an aspiring cradle-to-IPO capital stack. It is a long-running lead investor trying to arrive after a product becomes imaginable but before success becomes obvious. Fifty-five years in, the interesting part is not that CRV has survived. It is that the partnership still believes its best work happens when almost nothing is certain - including whether the sensible move is to invest at all.