Benchmark

Snapshot

• Founded: 1995
• HQ: SF Bay Area
• Stage: Early-stage (seed → Series A); concentrated portfolio (~10-12 deals per fund)
• Model: Equal-partner partnership
• Cap-table conflict status: Verified clean across all 8 Init Intelligence competitors as of 2026-05-05.

Lead deal partners

Roster turnover is significant — the entity stub previously listed partners no longer active.

Partner	Status	Why fit
Chetan Puttagunta (GP)	✓ Primary target	Most-aligned thesis voice at Benchmark. Cites “$1 trillion annual enterprise IT and IT services TAM” publicly + Sierra case study (his portfolio) on outcome-pricing-vs-seat. Same TAM framing as BCV/Echelon. Open-source/dev-tools/enterprise SaaS legacy (MongoDB, Elastic, Modern Treasury)
Eric Vishria (GP)	✓ Secondary	AI infrastructure GP since 2014. Led Fireworks AI ($25M Series A). Logan Bartlett show ep. 90. Most likely to think infrastructure-rigor about agent reliability/SLAs
Ev Randle (GP, joined 2025 from Kleiner)	✓ Pipeline urgency	At Kleiner backed Glean + Harvey — both services-as-software prototypes. Newest GP → most pipeline urgency. 20VC episode comments on “why margins matter less in AI”
Jack Altman (GP, joined Feb 2026)	Backup	Lattice founder; workforce SaaS lens. Brought Alt Capital roster (Bala Chanrasekaran, Vivek Katara) along
Peter Fenton (GP)	Senior, AI-cautious	Long-tenured

Removed from prior stub: Sarah Tavel (limited VP only since April 2025), Miles Grimshaw (left → Thrive 2024), Victor Lazarte (left → own fund July 2025).

Notes

• Benchmark’s concentrated model (10-12 deals/fund) means a single partner must develop deep conviction.
• Chetan’s “$1T enterprise IT” TAM framing and Sierra precedent (customer support) are a close thesis match. His TAM framing aligns 1:1 with BCV/Echelon’s framing — Benchmark and BCV are the only two firms publicly using this number. BCV has already deployed against the thesis (Echelon); Benchmark has not.

Related

• cap-table-patterns-across-startup-competitors
• Init Intelligence

---

Fundraise intel (2026-05-25)

Appended per the investor-intelligence build (Pass 1). Content above this line is unchanged. Sources: benchmark.com, Wikipedia, TechCrunch (Marina Temkin / Julie Bort), 20VC, KP “Welcoming back” post, The Information, partner LinkedIn/personal sites — all double-verified (Claude WebSearch/WebFetch + Exa CLI).

Firm Profile

• What they back: Concentrated early-stage (seed → Series A) across consumer + enterprise; pioneers of the equal-partner model (no hierarchy, no CEO, equal comp). Famous for network-effect / category-defining bets (eBay, Uber, Snap, Instagram, Twitter; recently Cerebras IPO).
• Fund discipline: **Deliberately keeps each main fund under ~$450M\ast \ast andrunsaconcentratedportfolio(10–12coredeals/fund)—asinglepartnermustdevelopdeepconviction.Foroversizedopportunitiestheyraise\ast \ast special-purposevehicles\ast \ast (e.g.two\ast \ast "BenchmarkInfrastructure"\ast \ast vehiclestotaling\ge$225M to double down on Cerebras in Feb 2026).
• Performance: 2020 fund reportedly >10x; 2024 fund ~3x (per Wikipedia/press). Cerebras alone: Benchmark owned 9.5% (≈$3.3B+ at IPO).
• HQ: 140 New Montgomery, San Francisco (some records list Woodside, CA).
• Stage focus for us: seed → Series A; they write the early lead check and stay concentrated → one GP carries the deal.

Partners (current GP line, 2026)

• Ev (Everett) Randle — GP (joined Oct 2025 from Kleiner Perkins). Our primary POC. Growth-stage enterprise/fintech pedigree (Vista → Founders Fund → KP); first Benchmark deal = Gumloop $50M Series B (Mar 2026).
• Chetan Puttagunta — GP (since Jul 2018; ex-NEA GP). Our secondary POC. OSS/dev-tools/enterprise; “$1T enterprise IT TAM” + Sierra precedent.
• Eric Vishria — GP (since 2014); AI-infra (Cerebras board since 2016, co-led its 2016 Series A; led Fireworks AI).
• Peter Fenton — senior GP, long-tenured.
• Jack Altman — GP (joined Feb 2026; Lattice founder; brought Alt Capital roster).
• Sarah Tavel — now Venture Partner (limited role since ~2025).
• Departed: Miles Grimshaw (→ Thrive, 2024), Victor Lazarte (→ own fund, 2025).

Recent activity (Pass-1 sketch — full data → seed-market-scan/firms/benchmark.csv)

• Gumloop — $50M Series B, Mar 12 2026, Randle led (his first Benchmark deal). “Every employee an AI agent builder.” ✅
• Cerebras — ≥$225Mintolateroundvia\ast \ast BenchmarkInfrastructure\ast \ast SPVs,Feb2026(TigerGloballedtheround);Vishriaboardsincethe2016$27M Series A. ✅
• LangChain — $10M seed, Apr 4 2023, Benchmark led (partner inferred = Chetan, OSS-first-lead pattern; not named in announcement). ⚠️
• Mercor / Sierra / Fireworks / Legora — the 2020-fund AI winners (Mercor $100MARRin11mo;Sierra$5–10B val; Fireworks ~$130M ARR; Legora). Round-level detail pending Pass 2. ❓/⚠️
• Note: Benchmark is multistage-concentrated, not a pre-seed shop — expect ~0 pure pre-seed; their early checks are seed/Series-A leads.

Best data channels for this firm

• TechCrunch — Marina Temkin is the most reliable beat reporter on Benchmark (broke Gumloop, Cerebras SPVs, the Randle/Annie-Case KP turnover). Julie Bort covered the Cerebras IPO.
• The Information + Eric Newcomer (Newcomer.co) — best for partner transitions / firm strategy.
• 20VC (Harry Stebbings) — has unusually deep Benchmark access; often breaks/amplifies Benchmark partner news first; both Ev and Chetan have appeared.
• Partner X/Twitter + personal sites (evrandle.com; @chetanp) — primary voice + thesis.
• benchmark.com/news — thin and infrequent but authoritative when present.
• Codex spine covers them reasonably for headline rounds; partner-per-deal attribution is weak (use press bodies / the LangChain-style “first lead in OSS” pattern to infer).

Fundraise Relationship (init.inc)

• Status: Benchmark is a clean lane (no ITSM/AI-MSP conflict — see cross-check). Their concentrated model means a single GP’s conviction carries the firm. Live outreach leads with Ev Randle — warm intro landed, meeting happened; his theses are Gumloop “every-worker-an-agent” and “margins matter less / gross dollars per customer.” Chetan is the thesis-fit secondary (Sierra-for-IT / $1T TAM).
• Conversation log (newest first):
  • 2026-05-26 — Raise-open follow-up email sent to Ev Randle (no reply since 5/17 post-meeting).
  • 2026-05-17 — Last contact with Ev; silent since.
  • 2026-05-15 (Fri 11am) — Met Ev Randle; went pretty well; he said he’ll find time again.
  • 2026-05-12 — Alex landed the warm intro to Ev (Quinn working a parallel path).
  • Chetan: 2026-05-25 set to Secondary / on hold; warm path via Isaiah if triggered.
• Mutual connections / warm path: Alex (Ev intro, primary) + Quinn (parallel) → Ev. Isaiah → Chetan (reserve).
• Personal & rapport notes (firm-level): Equal-partner culture → no “champion + IC theater”; the deal hinges on a single GP’s conviction. Ev is on record that margin is the wrong metric for AI (prefers gross $/customer, NRR over classic SaaS gross margin). They value non-consensus, well-argued theses (Ev’s “Playing Different Games”).
• Live stage: tracked in the CRM (Google Sheet) — Ev = Follow Up Phase; Chetan = Secondary.

Cross-check vs existing wiki

• ✅ Clean-lane verdict holds and is extended. The page above declares Benchmark “Verified clean across all 8 Init Intelligence competitors as of 2026-05-05.” A recency sweep (Benchmark investments since 2026-05-01; enterprise-AI-agent seed rounds in May 2026) surfaced no new Benchmark ITSM/AI-MSP overlap. The newest Benchmark deal closest to init.inc’s space — Gumloop (horizontal “agent builder,” not ITSM) — is adjacent but non-conflicting. Lane remains clean as of 2026-05-25.
• ⚠️ Sequencing note. Earlier framing (and cap-table-patterns-across-startup-competitors, line ~117) put Chetan first on thesis-fit (“Sierra for IT”), with Ev under “pipeline urgency.” Live outreach as of 2026-05-25 leads with Ev Randle because the warm intro landed and the meeting happened first — a connector-availability / live-pipeline fact, not a thesis contradiction. The cap-table-patterns line is stale relative to current sequencing (left unedited — out of this build’s scope).
• ✅ Existing roster notes (“Tavel limited; Grimshaw/Lazarte gone; Altman/Randle joined”) all corroborated. Refinement: Ev came from Kleiner Perkins but his full arc is Vista → KP/Bond → Founders Fund → KP → Benchmark (the existing “from Kleiner” is accurate-but-partial).

Related (this section)

• Ev Randle · Chetan Puttagunta · sierra · cap-table-patterns-across-startup-competitors

---

Pass 2 deal pattern (2026-05-25)

Exhaustive ~18–24mo pre-seed/SEED pull (Nov 2024 → May 2026), triangulated across Exa CLI, press (TechCrunch/Fortune/Newcomer/CNBC/Business Wire), and firm/founder posts. Full row-level data + verification labels → /Users/sazzad14/seed-market-scan/firms/benchmark.csv (20 rows; 15 net-new vs Pass-1). Content above this line is unchanged.

⚠️ Scope honesty (read first). The strict pre-seed/seed cohort in window is N = 1: Eigen ($15Mseed,Apr2026).\ast \ast Benchmarkessentiallydoesnotdopre-seedandrarelydoesalabeled"seed"—their\ast \ast effectivefirst-institutionalcheckisa$25–75M Series A lead (Greptile, Manus, Starcloud). To make a meaningful deal-pattern view, the CSV therefore also carries (a) Benchmark’s in-window Series A/B leads and (b) participate-only mid/late rounds in window, each tagged accurately by Stage and Firm_Role — plus a handful of pre-window anchors (Sierra Feb ‘24, Fireworks Mar ‘24, Mercor A Sep ‘24, Legora seed May ‘24, LangChain Apr ‘23) kept like Pass-1 kept LangChain. Stage labels are not inflated (a Series A is never relabeled “seed”).

Lead rate, check, stage, cadence

• Lead/co-led rate is very high when they engage: of the in-window rounds where Benchmark led or co-led, they were the (co-)lead in 6 of 6 (Eigen, Greptile, Manus, Starcloud co-lead, Gumloop, Monaco). Their “participate-only” rows (Mercor B/C, Sierra E, Legora B/C, Cerebras, Decart) are all follow-ons on companies they already led/seeded earlier — Benchmark almost never enters a new logo as a passive follower.
• Stage-equivalent: seed-labeled = 1 (Eigen $15M).Series-Aleads=Greptile$25M, Manus $75M,Starcloud$170M (co-led, outlier). Series-B leads = Gumloop $50M,Monaco$50M. **Typical Benchmark “first/lead check” in this window clusters $15–75M\ast \ast (the$170M Starcloud is an orbital-infra outlier).
• Cadence is genuinely thin (concentrated by design, fund discipline ~$500M/4 GPs per Chetan, Mar ‘26): 2025 produced only 2 new leads (Manus Apr, Greptile Sep); 2026 ramped to 4 in-window leads in ~5 months (Gumloop Mar, Starcloud Mar co-led, Eigen Apr, Monaco May) — i.e. ~2 leads/yr in 2025 accelerating to a ~6–8/yr run-rate in early 2026 as the partnership rebuilt. Effective new-lead cadence ≈ 3–6 deals/yr.
• Sector mix (in-window leads): dev-tools/AI-infra (Greptile, Starcloud), AI agents (Manus, Gumloop, Monaco), consumer-social AI (Eigen). Heavy AI-application + AI-infra; no ITSM/AI-MSP → clean lane confirmed (see cross-check above; Gumloop/Monaco are horizontal agent/sales tooling, adjacent not conflicting).

Notable seeds / first-checks

• Eigen — $15M seed, Apr ‘26 (the only true seed; consumer-social “mutual friend” AI; angels Silbermann/Söderström).
• Greptile — $25M Series A lead, Sep ‘25, Eric Vishria (AI code review; seed was Initialized’s, Benchmark took the A).
• Manus — $75Mlead,Ap{r}^{\prime }25,$500M val → acquired by Meta ~$2.5B Dec ‘25 (fast win; Chetan’s deal; drew US CFIUS review for China exposure).
• Starcloud — $170MSeriesA\ast \ast co-ledwithEQT\ast \ast ,Ma{r}^{\prime }26,$1.1B val (YC’s fastest unicorn; do not inflate to sole-led).

Partner-by-deal attribution

Partner	In-/near-window deals (role)
Eric Vishria	Fireworks A (led, Mar ‘24, board) · Greptile A (led, Sep ‘25, board) · Cerebras (board since ‘16; co-led ‘16 Series A; SPV double-down Feb ‘26) — the AI-infra GP
Chetan Puttagunta	Legora seed (led, May ‘24, board) · Sierra (co-led founding round Feb ‘24, board) · Manus (led, Apr ‘25) · LangChain (‘23 seed, inferred) — OSS/dev-tools/consumer-AI; the “Sierra for IT” anchor
Ev Randle	Gumloop B (led, Mar ‘26 — his first Benchmark deal, board) · Starcloud A (co-led, Mar ‘26 — inferred ^[ambiguous], partner not named)
Jack Altman	Monaco B (led, May ‘26 — his first Benchmark deal since joining Feb ‘26, board)
Victor Lazarte + Bill Gurley	Mercor A (led, Sep ‘24, Lazarte board) — both now departed/inactive (Lazarte left Jul ‘25); their deals are legacy, not a live channel
Eigen partner	Unconfirmed — not named in coverage; inferred Chetan or Fenton ^[ambiguous]

Partnership-compression timeline (explains the cadence): 6 GPs → 3 by Aug ‘25 (Grimshaw → Thrive ‘24; Tavel → Venture Partner Apr ‘25; Lazarte → own fund Jul ‘25), then rebuilt to 5 (Randle Oct ‘25, Altman Feb ‘26). The 2025 lull and the 2026 ramp track this directly — new GPs = new pipeline.

Role corrections vs Pass-1 / existing page (⚠️)

• ⚠️ Sierra was CO-LED (Sequoia + Benchmark), not Benchmark-led — the $110M founding round, Feb ‘24. Pass-1 had it ❓.
• ⚠️ Mercor: Benchmark LED only the Series A ($30M,Se{p}^{\prime }24,Lazarte+Gurley);theBandCwere\ast \ast ledbyFelicis\ast \ast withBenchmark\ast \ast participating\ast \ast .Pass-1hadMercorasageneric"Led❓."Also⚠\stackrel{◯}{R}\ast \ast amountdiscrepancy\ast \ast :PRNewswire$30M vs later TechCrunch “$32M"—using$30M (primary).
• ⚠️ Legora: Benchmark LED the SEED (Chetan, May ‘24) then participated only in the A (Redpoint led), B (ICONIQ/GC led), C. The existing page’s framing of Legora as a Benchmark “winner” is right but the role is seed-lead-then-follow, not lead-throughout.
• ⚠️ Monaco’s Series A (Feb ‘26) was Founders Fund’s, not Benchmark’s — Benchmark entered at the B (May ‘26, Altman). Do not mis-attribute the A.
• ✅ Gumloop, Cerebras re-verified unchanged from Pass-1.

Refined best-data-channels note (Benchmark-specific: truth vs noise)

• TRUTH: Eric Newcomer (Newcomer.co) is the #1 channel for partnership dynamics (the “Benchmark Has Lost Its Way” Aug ‘25 piece + the Altman-hire scoop nailed the 6→3→5 GP arc). Marina Temkin (TechCrunch) = best for deal scoops (Greptile-in-talks, Cerebras SPVs, Gumloop). Business Wire / GlobeNewswire / company blogs (Legora, Monaco, Starcloud) = authoritative for amount/lead confirmation. Founder/portfolio posts + Artificial Lawyer-type trade press recovered the seed-lead role databases dropped (Legora seed). The Peel / Turner Novak podcast = Chetan’s own words on strategy + Manus.
• NOISE / weak for Benchmark: benchmark.com is empty (just an address card — confirmed NOT a portfolio source). vcsheet/aggregator partner rosters are stale (still list Grimshaw/Lazarte). Partner-per-deal attribution is the weakest field — press names the GP only ~half the time (Eigen, Starcloud partners unnamed); inference via thesis-fit is required but must be flagged ^[ambiguous]. The Codex spine (per playbook) understates participate-only follow-ons and is partner-attribution-weak for Benchmark specifically.
• Single most-reliable channel for Benchmark: TechCrunch (Marina Temkin) for deals + Newcomer.co for partner/firm dynamics — together they covered every in-window lead and the roster churn that drives the cadence.

Pre-seed verdict + COULDN’T-GET

• Pre-seed = 0 (confirmed) across two independent sweeps — Benchmark runs no pre-seed vehicle; their early checks are seed→Series A leads. True seed = 1 (Eigen).
• COULDN’T-GET: (1) Eigen + Starcloud partner attribution — not named in any coverage found; inferred only, flagged ^[ambiguous]. (2) Legora Series C lead-of-round — headline didn’t name it; Benchmark was follow-on regardless. (3) Mercor A exact amount ($30Mvs$32M) — left at primary-source $30M with the discrepancy flagged. No deals were silently skipped.