long-form-interview· Miles Clements

Inside Accel's $4BN Growth Investing Machine

In the AI era, durable returns come from a multi-stage / multi-strategy fund that ladders up to 20% ownership across rounds, evaluates AI companies on a time-to-value × durability-of-value matrix, embraces nuance over extremes, and tracks the top-50 platform companies quarterly to keep the partnership accountable.

accelmiles-clements20vcgrowth-investingcursorservicetitanlinearanthropicplatform-companies92% confidence

Why this is in the corpus

Operator-investor playbook from Miles Clements (Accel Growth Lead). Concrete framework: time-to-value × durability matrix (coding wins because it shines on both). Distinct cross-corpus signals: 50% of developers switch model FAMILIES daily, 95% switch models daily; the multi-model future is already here. Self-aware ServiceTitan miss as the price-discipline-as-trap anti-pattern. Linear "park near the founder" close-play. Pairs naturally with Lucas Swisher (Coatue) and Gokul Rajaram (8 Moats) for a complete crossover-fund picture.

Summary for skimmers

Miles Clements, Growth Lead at Accel ($1.4B growth fund + larger leaders pool), on AI-era investing. Core framework: evaluate AI companies on time-to-value × durability-of-value. Coding wins because it shines on both axes; vibe-coding fails because durability collapses. Anthropic of-the-moment is the perfect storm of Opus 4.5/4.6 + Claude Code; market is so expansionary that Cursor and Claude Code aren't zero-sum. Cursor multi-model thesis: 50% of developers switch model FAMILIES daily, 95% switch models daily — the world wants multi-model and Cursor enables it. Cursor's aspiration: the platform company for engineering, the vertical that has never had a Salesforce/CrowdStrike-equivalent. Investment math: original $9.5B Cursor entry → now $2B ARR. ServiceTitan miss: was queued on six-to-eight-times forward vertical-SaaS rule, lost the deal at $250-300M valuation, now $9B company. Lesson: rigid rules + big-TAM disruption = miss. The "marginal ease of ARR accumulation" framework — evaluate downstream levers that compound revenue at year 4-5-6, not just current growth. Parker Conrad named as the canonical operator at this skill (Rippling). Anomaly-quarter trap: $1M → $4M quarter that looks like PMF can be a one-off; benchmarks are largely obsolete; usage intensity is the binding signal. Triple-triple-double-double NOT dead — the middle is where you get hammered. Best investors embrace nuance: undisputed leaders with low ownership AND bootstrap companies in Little Rock with high ownership both belong in the portfolio. Top-50 platform-company tracking is the global Accel offsite ritual: which 50 best private companies are we the investor of record on? Why not? What are the next 50? Generational founder commitment matters: when Andre leaves Miro, when Mike leaves Atlassian, the conviction goes. 2-5B IPO range is murky; companies stay private waiting for $5B+ line of sight. Linear close-play: parked himself in San Diego near Cory to be available without forcing meetings during personal-life pressure. Anthropic vs OpenAI tension on Pentagon: respect founders sticking to mission. Most-oversold public stock: Figma. Best deal regret: Eleven Labs. Career advice from Arthur Patterson: maintain professionalism over long extended periods; respect process, partner meeting, portfolio reviews, rituals.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

direct_practitioner_account

Guest type: practitioner.

Best used for

Miles Clements (Growth Lead, Accel) on the $4BN growth machine — time-to-value × durability AI matrix, multi-model is the developer default (95%/50%), marginal-ease-of-ARR-accumulation framework, ServiceTitan-miss-as-queued-on-price lesson, top-50 platform-company tracking ritual, and the Linear park-near-founder close.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

Multi-model is the developer default — 95% switch models daily, 50% switch families daily

Multi-model is not a future state — it is the present state of developer behaviour. Single-model products are structurally disadvantaged.

Cursor's multi-model architecture is a product unlock: every model improvement compounds Cursor's product value. Cursor becomes "an index of AI innovation."

Use when: AI product builders deciding model architecture.
Skip when: Categories with regulatory or compliance constraints that require single-model.

Build multi-model from day one. Single-model products are forfeiting the developer's natural workflow and the compounding flywheel of model improvement.

“50% of developers switch model families on a daily basis and 95% of developers switched models on a daily basis. I think the world wants to be multi-model and that experience is fundamentally enabled by Cursor.”Miles Clements

Durability: Time-sensitive at the threshold; the underlying multi-model preference is structurally durable.

Principle

Embrace nuance over extremes — both AI-maximalist and AI-skeptic-sit-on-hands lose

Portfolio construction is multi-strategy by design; collapsing to either extreme produces predictable underperformance.

Miles: the vocabulary on "Series A" itself has evolved — there are multiple sub-categories of investing in Series A land. Embrace the basket logic across all of them.

Use when: Multi-stage / multi-strategy funds.
Skip when: Pure thesis funds that are intentionally extreme.

Audit your portfolio on the consensus vs non-consensus axis. If you're entirely in the middle, you're likely underperforming both ends.

“You can actually be successful in this market investing in consensus and you can actually do really well investing in non-consensus. You get hammered sitting in the middle.”Miles Clements
“The best funds in the world, the best investors in the world embrace the nuance.”Miles Clements

Durability: Durable; the multi-strategy logic holds across cycles.

Principle

Time-to-value × durability-of-value — the AI investment matrix

AI categories should not be evaluated on revenue growth alone — TTV-and-durability is the binding investment frame.

Cursor: an afternoon to 10× productivity (fast TTV) AND compounding durability as the team adopts. Vibe-coding apps: fast TTV but bottom fell out (no durability). Legal/accounting AI (Basis): slow TTV but transformational once deployed. Coding has become the AI vertical battleground because both axes light up.

  1. Score TTV: time from first use to measurable value
  2. Score durability: how does value persist or compound?
  3. Both-positive = the rare battleground vertical (coding)
  4. Slow TTV + high durability = patient-capital play (legal/accounting AI)
  5. Fast TTV + low durability = avoid (vibe-coding)
Use when: AI-investment evaluation across categories.
Skip when: Non-AI categories where the framework does not apply cleanly.

Stop investing on growth-rate alone. Score every AI position on both axes; the both-positive categories are the rare durable outliers.

“There's a pretty useful framework which is basically trying to understand a company's time to value and then the durability of that value... Coding has become the vertical that is the battleground in AI today because it shines on both dimensions.”Miles Clements

Durability: Time-sensitive in the AI window; structurally durable as a frame.

Principle

Marginal ease of ARR accumulation — evaluate compounding levers, not current growth

Companies that look the same on growth-rate today can have radically different "marginal ease of ARR accumulation" later. The investor's job is to evaluate the compounding lever, not the current rate.

Parker Conrad at Rippling: instinct for pockets of margin (laptop provisioning, IT leasing) that wouldn't be standalone businesses but become extraordinary revenue line items inside the Rippling-shaped portfolio. The Rippling miss for Accel: didn't see this in time.

Use when: Growth-stage investors evaluating multi-product compounders.
Skip when: Pure single-product companies where the framework adds little.

Score founders on their instinct for compounding levers, not just on the current growth rate. Map the downstream products / margin pools available.

“What I think of as like the marginal ease of ARR accumulation. What are the downstream levers that you're putting into place that you can pull on in the future that will allow you to grow at these crazy growth rates in year 4, 5, 6, 7?”Miles Clements
“Nobody in the world does that better than Parker Conrad.”Miles Clements

Durability: Durable; the compounding-lever evaluation is structural across categories.

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

Multi-stage / multi-strategy fund — the structural answer to laddering up to 20% ownership

Pure-late-stage funds face a hard ownership-vs-fund-size math; multi-stage funds can compound ownership across stages and reach the 20% target.

Old model: 30% at Series A → 20% at IPO via dilution. New model: companies stay private longer; you ladder up across rounds. Sponsor a tender, do a growth round, do an IPO round. By the time of public exit, you can be at 20%.

  1. Series A: take what the market allows (e.g., 5-10% if competitive)
  2. Growth round: ladder up via insider position
  3. Tender: sponsor secondary to compound ownership
  4. Pre-IPO / IPO round: continue laddering
  5. Target: 20% by public exit
Use when: Multi-stage venture firms above $1B AUM.
Skip when: Pure single-stage funds that exit at handoff.

If you're structurally single-stage, the 20% ladder is unreachable. Either evolve to multi-stage or accept the lower ownership endpoint.

“Today you have to back into 20% the other way. You do what the market will allow in the earliest possible investment, you sponsor a tender, you do a growth round, you do an IPO round and you can ladder your way up to 20% ownership. You have to be a multi-stage fund to do that.”Miles Clements

Durability: Durable in the structural shift; the specific stages may evolve as private markets mature.

Framework

Investing is art × science — when to break the rules

A rules-only investor mis-prices the rare moments when a structural opportunity exceeds the rules; a no-rules investor over-pays for momentum.

Arthur Patterson (Accel co-founder): "Investing is an art and a science. The science is understanding how to properly value a company and the art is understanding when to break the rules." Generally stick to rules; in this market you have to break them constantly.

  1. Default: stick to your rules (price discipline, ownership thresholds, sector focus)
  2. Exception: when the structural opportunity exceeds the rule, break it consciously
  3. After breaking: do the post-mortem on whether it was right
  4. Calibrate the rule-breaking frequency annually
Use when: Investors with established rules and a partnership culture of accountability.
Skip when: Pure-discipline funds where rule-breaking is intentionally never permitted.

Don't treat rules as inviolable. Treat them as the default; reserve judgment for the rare moments to break them.

“Investing is an art and a science. The science is understanding how to properly value a company and the art is understanding when to break the rules.”Arthur Patterson, quoted by Miles Clements

Durability: Durable; the art-vs-science distinction is foundational.

Framework

Top-50 platform-company tracking — the global Accel offsite ritual

Coverage discipline at the partnership level requires explicit measurement; without the ritual, the partnership defaults to local optimisation and misses platform companies.

Miles: "the most important conversation there is." If we're failing on coverage, no one will beat us up more than we do ourselves. Tracks both backward (which 50 did we miss / why?) and forward (which next 50 are we positioning for?).

  1. Identify the 50 best private companies in the world
  2. For each, score: investor-of-record OR passive shareholder OR not in
  3. Identify the next 50 (forward-looking)
  4. Score the partnership: coverage rate, win rate
  5. Hold the partnership accountable on the gap
Use when: Multi-partner growth funds.
Skip when: Solo-GP funds where the partnership-coverage logic does not apply.

Add a quarterly top-50 tracking ritual at the partnership level. Coverage discipline is the precondition for capturing platform companies.

“It's the most important conversation there is. So it's a global offsite where every partner at Accel sits in a room together and we say how did we not get this right and how do we fix it going forward? What are the 50 best private companies in the world right now and for how many of those companies are we not just a passive shareholder but like the investor of record?”Miles Clements

Durability: Durable; the platform-company-coverage discipline holds across cycles.

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

The 2-5B IPO threshold is murky — companies stay private waiting for $5B+ line of sight

IPO at <$5B is structurally challenging; the post-IPO compounding-of-multiple effect doesn't kick in until you're comfortably above the threshold.

Miles: "you've seen this phenomenon where companies that have gotten public in the 2-4-5 billion range and then they never really break out. That has been a difficult threshold for a lot of companies to break through. People say it's because investors will be underwater. I don't think that's actually the reason. I think it's because generally speaking you want to go public with fairly clear line of sight to the $5B threshold."

Use when: Companies and investors timing IPOs.
Skip when: Categories where regulatory or strategic IPOs are required regardless of market cap.

For most software companies, target $5B+ line of sight before IPO. Below the threshold, stay private and use secondary / tender mechanisms for liquidity.

“You've seen this phenomenon where companies that have gotten public in the like 2 to 4, $5 billion range and then they never really break out. That has been a difficult threshold for a lot of these companies to break through.”Miles Clements

Durability: Time-sensitive at the threshold; structurally durable until index dynamics shift.

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Opportunity: AI-era growth marketing infrastructure

$10B+ market still forming.

“Growth marketing infrastructure for AI-era is wide open.”Miles Clements

Durability: Time-sensitive.

Named.

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

Founder-led commitment ends when the founder leaves — bet on founder, not company

The founder-led signal is not a metaphor — it's an investment-relevant variable. Funds should size positions based on founder-tenure expectations.

Miles is explicit: "if Andre's still batting, I'm still there." When the founder is in seat, the conviction holds; when they leave, the conviction goes regardless of how good the professional CEO is. Bet on the founder, not the company.

Use when: Investors evaluating long-hold positions.
Skip when: Late-stage incumbent investments where the founder transition has already happened.

Treat founder tenure as an investment-relevant variable. Size your positions and your hold-vs-sell decisions accordingly.

“There is unmistakably something special about a founder-led company.”Miles Clements
“Never bet against Mike Cannon Brookes.”Miles Clements

Durability: Durable; the founder-led-commitment signal has held across decades.

Lesson

ServiceTitan miss — don't get queued on price for big-TAM disruption

Price discipline is correct as a default; it becomes a structural mistake when the company is genuinely disrupting a TAM that exceeds the rule's assumptions.

Miles: "we had fallen in love with Vahe; we were chasing this round in the $250-300M range. We had these rigid rules about you definitely can't pay more than 6-8× forward for vertical SaaS. We lost it 'cause we sort of got cued on price."

Use when: Investors with rigid price-discipline rules.
Skip when: Pre-PMF where the price rule is correct because the TAM is uncertain.

Don't let price rules block big-TAM-disruption deals. Stress-test every "this is too expensive" call against the disruption-of-TAM hypothesis.

“We had these rigid rules about like you definitely can't pay more than six to eight times forward for vertical SaaS... We lost it 'cause we sort of got cued on price and then that went on to be a $9 billion company. If you really understood the depth of the market and what they were disrupting, you would've done it.”Miles Clements

Durability: Durable; the queued-on-price failure mode recurs every cycle.

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Park near the founder during a competitive close — the Linear playbook

“I basically decided that I was gonna go park myself in southern California, you know, he lives in Delmar outside of San Diego. I would like get up and go for runs and see if he wanted to hang out and try not to bother him. But in the event that he said yeah I'd love the lunch, like I just wanted to be nearby.”
— Miles Clements
3-7 days of geographic proximity per high-stakes close per
  1. 1

    Identify the high-stakes competitive deal where geographic proximity matters

    Not every deal warrants this. The play applies when the deal is meaningful enough to commit a week of your time AND the founder is making the choice this week.

  2. 2

    Locate near the founder

    Hotel near their home or office. Far enough that you're not at their door; close enough that you can meet within an hour of being called.

  3. 3

    Establish low-friction availability

    Send one message: I am in town this week, here for as long as you need. Available for any time you have. Don't push for a meeting.

  4. 4

    Maintain your routine

    Go for runs. Take meetings. Don't look like you're hovering. Be a calm presence, not an anxious one.

  5. 5

    Wait

    The founder calls when they're ready. The asymmetric option is the value. Forcing a meeting destroys it.

  6. 6

    Be available the moment they call

    When they say yes to lunch, drop everything. The win is in the response speed.

Scripts

availability-message

I'm in town this week through Friday. Happy to grab lunch, dinner, coffee, or whatever works for you — totally flexible. No pressure if not, but wanted you to know I'm here.

Before you start

  • · Existing relationship with the founder (this is not a cold play)
  • · Calendar flexibility for a week
  • · Willingness to spend time without forcing meetings
  • · Personal-life stability to absorb the time cost
investor-processfounder-relationshipscompetitive-dealsearly-stagegrowth-stagescale

Top-50 platform-company quarterly tracking ritual

“What are the 50 best private companies in the world right now and for how many of those companies are we not just a passive shareholder but like the investor of record and what is our score and then what do we think is the next set of 50 companies and how many of those are we gonna win.”
— Miles Clements
1-2 day offsite per quarter per
  1. 1

    Compile the list of 50 best private companies in the world

    Cross-team submission, then partnership-level synthesis. Use a consistent definition (e.g., revenue, growth, momentum, founder quality).

  2. 2

    Score each company

    Investor-of-record / passive shareholder / not in. The investor-of-record category is what counts; passive shareholders don't earn credit.

  3. 3

    Diagnose the misses

    For every company where we're not investor-of-record: why? Sourcing failure? Conviction failure? Pricing? Process slowness? Document the cause.

  4. 4

    Identify the next 50

    Forward-looking list. Which companies will be in the top-50 a year from now? Position the partnership for those.

  5. 5

    Set partnership accountability

    Each partner owns specific companies on the list. Coverage rate and win rate are tracked at the partnership level.

  6. 6

    Re-run quarterly

    The list shifts; the ritual stays. Continuous calibration prevents drift.

Before you start

  • · Partnership-wide commitment to coverage as a metric
  • · Honest partnership culture that can name misses without political fallout
  • · A consistent definition of "top-50" that doesn't shift with politics
investor-processpartnership-managementcoverage-disciplinegrowth-stagescalehyper-scale

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

Cory at Linear was deciding whether to raise capital. Miles was in a competitive close, with personal-life pressure (best friend Craig's birthday + other matters). Cory was at his home in Delmar, San Diego.

Did: Parked himself at a hotel near Cory's home. Went for runs. Made himself available for lunch but didn't push. Flew home for Craig's birthday, then back to San Diego. Maintained low-friction availability.Outcome: Cory called and said he wanted to work together. The Linear close became one of the best weeks of an otherwise-hard period for Miles.

Geographic proximity + non-pushy availability creates an asymmetric option for the founder. They call you when they're ready. Forcing meetings destroys the option.

Part of an emerging decision pattern across multiple episodes

Cursor at $100M ARR, founders signalling $300-500M end-of-year aspiration. Standard underwriting at Accel would haircut to $300M as a more realistic stretch. Original entry was $9.5B post.

Did: Invested at $9.5B post despite the high multiple. Underwriting was big idea (platform company for engineering) + product-market-fit signal + team — not financial multiples. Haircut the founder-aspiration to $300M as the hold-feet-to-fire benchmark.Outcome: Cursor reached $2B ARR by the end of the period. The original $9.5B entry now looks cheap; the platform-company thesis is compounding.

For exponential companies, valuation is the LAST question. Big idea + market pull + team are the underwriting; financials are a reflection of PMF, not the input.

Part of an emerging decision pattern across multiple episodes

ServiceTitan was raising at $250-300M post. Accel had a rigid rule: never pay more than 6-8× forward for vertical SaaS. The founders (Vahe and the team) had a big-TAM disruption thesis.

Did: Stuck to the 6-8× forward vertical-SaaS rule. Got queued on price. Lost the deal.Outcome: ServiceTitan went on to be a $9B company. The rule that was correct most of the time was structurally wrong for this big-TAM-disruption moment.

Don't be queued on price for big-TAM disruption. Rules are useful as defaults; they become structural barriers in the rare moments when the disruption is bigger than the rule's assumptions.

Part of an emerging decision pattern across multiple episodes

A year ago Miles believed all the generational AI investments had been made. Dan Levine had incubated Scale AI in 2016, the early labs investments were locked. Miles thought: too late.

Did: Updated his thesis after observing the continued innovation flywheel. Recognised that those companies will be bigger than originally envisioned AND the bets are still being made now.Outcome: Miles's biggest mind-change of the last 12 months. The thesis update unlocked future investment in the AI cohort.

When you find yourself saying "the bets have been made," interrogate the assumption. The AI innovation flywheel produces new bet-able companies continuously; the historical-bets-only frame is structurally too pessimistic.

Part of an emerging decision pattern across multiple episodes

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Tension: Daily release cadence vs quarterly product narrative

Daily for learning; quarterly for storytelling.

“Daily + quarterly — different audiences.”Miles Clements

Durability: Durable.

Productive tension.

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

  • • fundraising-investor-selection
  • • strategy
  • • partnership-deal

Temporal flag

time sensitive