The Twenty Minute VCElevenLabs: Building an AI Sales Machine & Why We Set a 20x Sales Quota
CHAPTERS
AI-first CRO mindset: distribution, speed, and “revenues of tomorrow”
Carles frames why the CRO role has changed: it’s no longer just closing deals, but architecting distribution and building a system that compounds. He emphasizes embedding AI across go-to-market so teams can do more with less while staying execution-oriented.
- •Modern GTM is about distribution strategy, not just deal-making
- •AI should be embedded end-to-end in revenue workflows to increase leverage
- •A good CRO prioritizes future revenue engines, not only current quarter performance
- •High-velocity markets demand faster iteration and decision-making
Why most AI outbound tools fail—and what ElevenLabs built instead
Carles argues most “AI SDR” tools treat outreach as a transaction, creating spam and collapsing response rates. He explains ElevenLabs’ internal approach: custom AI agents that support humans with context-rich drafts and workflows across inbound, proposals, and customer success.
- •Outbound fails when AI automates generic messaging without channel/recipient preference
- •Response rates and tolerance for automated outreach have collapsed
- •ElevenLabs built internal AI agents: inbound SDR, RFP/RFI scanning, CS email drafting
- •Closed-loop learning: store drafts/sends/responses to improve and personalize per account
- •Human-in-the-loop, customer-specific tone/language is what makes AI effective
Will AI shrink sales teams? The 50% productivity target and talent concentration
Rather than replacing sales entirely, Carles expects AI to drive major productivity gains—his target is ~50%. That enables smaller, more elite teams, with commissions paid even on AI-assisted or AI-originated upsells to keep incentives aligned.
- •Goal: 50% productivity improvement across the revenue org
- •Smaller teams, higher compensation, and higher talent density
- •AI-closed expansions can still earn human commissions to maintain incentives
- •Impact: fewer hires than traditional scaling, but more concentrated top performers
The 20x quota model: commissions, accelerators, and why big checks are good
Carles breaks down ElevenLabs’ unusually high quota philosophy and how comp is structured to drive ambition. He explains baseline commission, accelerators for overperformance, and why paying large commissions is rational given the valuation impact of incremental ARR.
- •20x quota is meant to be challenging but ‘fair’; adjust if consistently unrealistic
- •Baseline commission (e.g., ~5%) plus tiered accelerators (1.1x, 1.2x, 1.3x, 1.5x…)
- •Extra incentives (spiffs) can target priority products but can also distort behavior
- •Large commission checks are celebrated because incremental revenue massively increases valuation
- •Commission credit can apply across roles (SDR/AE/CSM/engineer contribution)
Avoiding comp traps: pilots, retention/expansion rules, and strategic account windows
He warns that too many incentives create perverse selling. ElevenLabs avoids paying commissions on pilots and instead rewards annual/multi-year value, with special rules for retention and strategic accounts to keep teams focused on durable revenue.
- •Too many spiffs/accelerators can encourage ‘selling for the incentive’
- •No commissions on pilots (don’t meaningfully add valuation; avoid gaming)
- •Commissions paid on annual contracts; expansion in first 12 months is comp-eligible
- •Strategic accounts can unlock longer commission windows (e.g., up to 2 years)
- •Comp design should align seller behavior with durable revenue and renewals
Hunters, farmers, and customer success as a revenue function
Carles values hunters but warns unmanaged hunters can damage long-term account health through inconsistent pricing and messaging. He argues customer success must be oriented toward expansion and retention (revenue), especially in AI where switching costs can be low.
- •Hunters matter, but unchecked hunting creates pricing/value inconsistencies
- •CS should not be ‘happiness’ only; it should drive expansion, cross-sell, retention
- •AI markets increase competitive pressure—customers can spin up alternatives quickly
- •Services can help, but over-transactionalizing services can hurt community/trust
- •Balanced coordination between hunters and CS reduces renewal risk
“One market at a time” is obsolete: parallel bets and hiring experienced closers
Carles challenges the classic VC playbook of sequential geographic expansion. He argues companies must parallelize GTM bets to outrun fast-follow competitors, and that hiring seasoned sellers (even 20+ years experience) can compress cycles via established relationships.
- •Sequential market expansion is too slow in AI-era competitive dynamics
- •Parallelize across markets and channels with a clear thesis per bet
- •Controversial take: experienced enterprise sellers can be a superpower, not a culture mismatch
- •Veteran reps can go straight to C-level buyers and shorten sales cycles
- •Culture fit matters, but hunger + alignment can bridge age/experience gaps
What didn’t work: early media/entertainment push, agentic pivot, and the India reset
Carles shares failed and successful experiments: selling directly into major studios didn’t convert as expected, but shifting to media creation platforms did. He also describes mistakes in India—verticalizing too early—followed by a “back to zero” reset using stronger pipeline construction.
- •Direct entertainment studio selling was slower than expected; industry readiness matters
- •Pivot to media creation platforms improved traction
- •Early bet on agentic systems paid off as AI agents became a major growth area
- •India mistake: vertical segmentation too early with too small a team hurt a quarter
- •Reset: return to horizontal selling + named accounts + rebuild confidence via liquidity
Pipeline construction like portfolio construction: liquidity, whales, and hard vertical bets
Carles explains his “pipeline construction” ritual: design a pipeline mix like a VC portfolio—some fast-closing deals to maintain momentum and some large strategic whales. He uses government as an example of a hard, sticky bet driven by mission and long-term payoff, despite slower ramp.
- •Weekly pipeline construction: design mix by segment/market, not just quotas
- •Balance ‘liquidity’ deals (quick wins) with ‘whales’ (large enterprise)
- •Too few quick closes erodes rep confidence; too few whales caps growth
- •Government is a deliberate hard bet: sticky, societal impact, long-term defensibility
- •Experimentation mindset makes traditional forecasting less precise
Brand and enterprise cycles: “no one gets fired for buying IBM”
Carles strongly asserts brand reduces enterprise sales cycles and procurement risk. He points to a small set of AI ‘blue chip’ vendors and argues ElevenLabs must replicate that trust position to accelerate adoption and reduce perceived buyer career risk.
- •Brand materially shortens enterprise cycles and reduces buyer risk
- •Procurement favors ‘safe choices’; trust is a competitive advantage
- •Examples of perceived AI blue chips: OpenAI, Anthropic, Cursor (as discussed)
- •Once enterprise motion starts, stickiness compounds despite hard initial entry
- •Brand-building is strategic GTM work, not just marketing vanity
Scaling the sales org without dilution: transparency, leaderboards, and hiring for obsession
Carles discusses doubling the revenue team while protecting culture and standards. He describes radical transparency (per-rep quota attainment), how to interpret performance beyond a leaderboard, and his fast pattern recognition for “obsessed” talent in short interviews.
- •Primary scaling risk is cultural dilution if expectations aren’t explicit
- •Public internal stats/leaderboards create accountability and competition
- •Evaluate reps on pipeline-building actions, not only near-term revenue
- •Hiring signal: energy, sharpness, depth, preparation—often clear in ~20 minutes
- •Set ‘this is hard’ expectations upfront to reduce later churn
Partner ecosystems done right: CVC strategy partners, incentives, and time horizons
Carles outlines a partnership model anchored in strategic corporate investors (CVCs) who can unlock distribution and industry insight. He emphasizes partner motions take a long time, require dedicated resourcing, and need clear incentive structures (including performance commitments).
- •CVCs can be powerful ‘strategy partners’ who navigate big enterprises internally
- •Align incentives: investment allocation tied to revenue/pipe commitments
- •Enforcement mechanisms can include buyouts/penalties if commitments aren’t met
- •Partnerships require dedicated people, tracking SQOs/SQLs before revenue is reliable
- •Best ecosystems are bidirectional (partners send and receive deals, like Salesforce)
Operators as investors: when it helps, when it distracts, and how to add value
Carles argues operators should invest because it reinforces learning and lets them help founders with real execution. Harry pushes back on attention conflicts; Carles counters that it works if the operator remains the hardest-working person and focuses on being useful beyond capital.
- •Debate: institutional investing vs commitment to the day job
- •Carles’ view: investing and operating can reinforce each other if performance stays elite
- •Best operator value-add: GTM, hiring, customer calls, intros, pitch feedback
- •Small checks matter less than time, insight, and hands-on help
- •Key diligence mistake: backing founders who won’t iterate on go-to-market
Quick-fire and closing: goals, constraints, and what’s next in foundation models
In rapid Q&A, Carles shares personal and professional targets, what he finds hardest (time trade-offs), and how he recharges. He also looks ahead to a new wave of foundation model companies and discusses why focus beats spreading too thin—both for AI leaders and for his own work habits.
- •2026 happiness metric: ElevenLabs at $1B+ revenue and multiple fund unicorns
- •Hardest challenge: time allocation and personal trade-offs
- •Personal escape: gardening/plant care as stress relief
- •Excitement: next wave of foundation model startups and M&A by incumbents
- •Preference for focus: doing fewer things better vs overextending product roadmaps
