The Twenty Minute VCAngelList CEO Avlok Kohli: The Funding Market Today; Last-Mile Delivery; Competing w/ Carta | E1005
CHAPTERS
- 0:00 – 0:31
AngelList’s speed-first operating philosophy (no process theater)
Avlok opens with a blunt philosophy: eliminate any process that slows product shipping. He frames a company as a group of people building for customers, not maintaining meeting-heavy bureaucracy. The focus is speed, iteration, and customer feedback loops.
- •Cancel meetings that don’t directly drive shipping and customer value
- •“Meetings about meetings” and excess process are organizational drag
- •The core job is shipping products customers love
- •Tight iteration cycles with customers beat internal ceremony
- 0:31 – 2:00
How Avlok became CEO: from semi-retirement to “start tomorrow”
Avlok explains his path to CEO, built on a long relationship with Naval (an investor in all three of Avlok’s prior startups). After stepping back post-acquisition to focus on investing, he was pulled into AngelList after months of conversations. The transition was abrupt and intense, reflecting AngelList’s urgency.
- •Naval invested in Avlok’s prior startups and later recruited him
- •Avlok initially hesitated after the marathon of startups
- •Six months of discussions led to accepting the role
- •Immediate start date set the tone for high-velocity execution
- 2:00 – 3:41
Last-mile delivery lessons from Fastbite: utility functions, not single-variable optimization
Harry digs into Avlok’s Fastbite experience (7-minute average delivery, marketed as 15). Avlok shares two core lessons: building for yourself improves product-market fit odds, and optimizing the right combination of variables matters more than pushing one metric to the extreme. He highlights diminishing returns—sometimes “faster” isn’t actually better.
- •Building for yourself can sharpen product intuition and PMF
- •Every product has a utility function (price, convenience, speed, experience)
- •Diminishing returns apply—ultra-speed may not increase satisfaction
- •Optimize across multiple variables, not just one
- 3:41 – 4:51
Why “first to market” is overrated: the idea maze and building the map for others
Avlok argues that being first is often less important than successfully navigating the ‘idea maze’—the complex set of industry tradeoffs founders must learn. First movers may educate later entrants unless they execute exceptionally well. He uses Google and the iPhone as examples of winners that were not first.
- •Idea maze: first movers often create the map others use
- •First-to-market doesn’t guarantee category leadership
- •Execution and learning speed matter more than timing alone
- •Examples: Google not first search engine; iPhone not first smartphone
- 4:51 – 5:45
Defensibility at pre-seed/seed: you don’t have a moat on day one
On defensibility, Avlok defines it as the cost to replicate a business over time. By that definition, early startups generally have no moat because they don’t yet have scale, distribution, or hard-to-copy assets. Network effects rarely appear meaningfully in the earliest days.
- •Defensibility = rising cost for others to replicate you over time
- •At day one/day 30, most startups have effectively zero moat
- •Early network effects are rarely durable unless scale is real
- •Early-stage evaluation should not over-index on ‘moats’
- 5:45 – 7:22
15-minute delivery boom/bust: logistics depends on capital markets
Avlok reacts to the wave of rapid-delivery startups by emphasizing how capital-intensive logistics is. When rates rise and capital dries up, the model’s J-curve becomes hard to finance, especially when expansion requires subsidized unit economics. He predicts widespread failure during downturns due to cost of capital and funding scarcity.
- •Logistics (online-to-offline) requires heavy capital to scale
- •Higher interest rates raise cost of capital and kill expansion plans
- •Many models lose money per order until scale; financing the gap is key
- •Downturns make it “near impossible” to raise what’s needed
- 7:22 – 10:26
The funding market now: AngelList’s data shows fastest, deepest venture contraction
Avlok describes AngelList’s unusually broad data view (funds, syndicates, documents, cap tables) and a quarterly “boom times vs bad times” chart. The market has moved from top-right (high activity + many up rounds) to bottom-left (low activity + fewer up rounds). He stresses both the magnitude and the speed of the contraction since early 2022.
- •AngelList tracks activity rate and positive activity (up rounds) quarterly
- •2021/early 2022 = boom; current = one of the worst ‘bad times’
- •Compression is both historically large and unusually fast
- •Fewer startups are raising; fewer rounds are positive
- 10:26 – 14:07
Two-speed market: seed holds up while Series A/B/C resets (and Series C “basically gone”)
Harry challenges the apparent resilience of seed pricing; Avlok explains the disconnect: late-stage has repriced dramatically while seed remains buoyed. Multi-stage investors are moving earlier, increasing capital supply at seed. Meanwhile Series B volumes and valuations are down sharply and late-stage deal activity has largely evaporated.
- •Seed valuations are more stable than Series A/B/C right now
- •Series B: deal volume and valuations down ~50% (per Avlok)
- •Series C/late-stage deal flow is very rare
- •Later-stage investors have shifted earlier, propping up seed
- 14:07 – 17:37
Round dynamics: party rounds steady share, but ‘seed++’ era disappears; time-to-raise increases
They discuss how rounds are being formed: the proportion of party rounds vs led rounds is similar, but total volume is down. The ‘successive seed plus plus’ pattern of repeatedly topping up with small party rounds has largely ended. Avlok also notes time-to-raise and time-to-markups are increasing, with fundraising activity ~33% below historical averages.
- •Party-round share roughly unchanged, but fewer total rounds happen
- •‘Seed, seed+, seed++’ incremental raises are largely gone
- •Larger seed rounds (~$5M) are harder without a lead
- •Time-to-raise up; activity ~33% below historical averages
- 17:37 – 23:39
Repricing and survival mechanics: pay-to-plays start ticking up; M&A slows instead of surges
Avlok says many startups still have cash (often extended via layoffs), delaying repricing, but pay-to-play rounds are beginning to rise and will likely accelerate as cash-out dates approach. On acquisitions, he’s contrarian: both public and private acquirers face constraints, and integration costs are real. AngelList data shows M&A down significantly and likely to remain muted until macro improves.
- •Repricing is delayed by runway extensions, but it’s starting
- •Pay-to-plays are increasing and may rise more in coming quarters
- •Acquisition sprees are unlikely; acquirers face pressure and uncertainty
- •Integration overhead makes ‘cheap’ acquisitions non-trivial; M&A is down
- 23:39 – 25:41
Secondaries and LP behavior: more sellers, unclear buyers; 20–50% discounts on fund positions
Harry asks about liquidity; Avlok notes increased appetite from LPs to sell fund positions, sometimes even in high-quality names, driven by allocation mismatches or liquidity crunches. The market clears only when buyers and prices align. Typical discounts observed are ~40–50% from last mark, sometimes closer to ~20%.
- •More LPs want to sell positions due to liquidity and allocation pressure
- •Key bottlenecks: who buys and at what clearing price
- •Secondary transactions still happen, including in large funds
- •Discounts commonly 40–50% from last mark (sometimes ~20%)
- 25:41 – 30:38
Micro-funds and fundraising: fewer new funds, longer to first close, smaller first-time fund sizes
Avlok explains capital flows by LP type and product: rolling funds are down modestly, traditional venture funds saw individuals down sharply while institutions recently rebounded. Net-new fund formation is down, and time to first close has expanded from ~3 months in boom times to 6–7 months. First-time fund sizes are also shrinking materially.
- •Rolling Funds: commitments down ~20% (per Avlok)
- •Traditional funds: individuals down ~60%; institutions have rebounded recently
- •Net-new funds are down; experienced managers raise more per fund
- •Time to first close increased to 6–7 months; first-time fund sizes smaller
- 30:38 – 33:37
How AngelList uses AI: replacing ‘human judgment’ workflows like deal-email routing and doc checks
Avlok details how LLMs are automating operational workflows across fund administration. With massive historical training data, AngelList can route high-stakes emails, classify and validate legal documents, and auto-request missing items (e.g., cap tables). The effect is meaningfully reducing manual effort in processes previously assumed to require human judgment.
- •AngelList manages end-to-end fund workflows across a 10-year lifecycle
- •LLMs enable replacing some judgment-heavy work with code
- •Example: automate routing/triage across ~15,000 weekly fund emails
- •Auto-identify document types and detect missing pieces (e.g., cap tables)
- 33:37 – 41:51
Margins, pricing, and platform economics: 80%+ holistic margins; SPVs are harder than they look
Harry presses on pricing vs cost reductions; Avlok says the pricing change largely removed a cap that no longer matched larger fund sizes AngelList now serves. Funds are strong-margin, and overall customer-level margins are north of 80% due to platform compounding across products. Smaller SPVs can be loss-leading; he cites Assure’s failure as evidence that SPV ‘tails’ and long-term obligations create hidden costs.
- •Pricing increase tied to moving upmarket and removing outdated caps
- •Funds are high-margin; overall platform margins north of 80% (per Avlok)
- •Platform model: products feed each other, expanding margins over time
- •Small SPVs can be loss-leading; long-tail surprises make them complex
- 41:51 – 45:56
Product strategy calls: regretted incorporation product, Europe shutdown, and the ‘win the US first’ thesis
Avlok names incorporation as a misstep because AngelList didn’t bring unique advantage; they later partnered with Stripe Atlas. On Europe, he explains AngelList pulled back due to fragmented regulatory structures and to avoid splitting resources before dominating the US. Winning the US is defined as market share across funds and venture-backed startups, with a goal of 50%+ in segments they enter.
- •Regret: building incorporation tooling; better solved via Stripe Atlas partnership
- •Expansion discipline: sunset initiatives that aren’t scaling fast enough
- •Europe: regulatory complexity + resource dilution; focus on US first
- •‘Winning the US’ = >50% market share across targeted segments
- 45:56 – 49:51
Competing with Carta via cap table: customer obsession over rivalry, and execution speed as the weapon
On the cap table product, Avlok frames competition as something to watch but not fixate on—startups die by ‘suicide’ (internal mistakes), not competitors. AngelList is winning customers from multiple sources, including incumbents, by building a better experience and moving quickly once direction is clear. Execution speed is tied to ambitious hiring, aggressive deadlines, and rapid blocker removal.
- •Competition lens: prioritize customer experience; don’t obsess over rivals
- •‘Startups die of suicide, not homicide’—internal execution matters most
- •Customer acquisition comes from net-new and switching from other providers
- •Speed drivers: ambitious talent, clear deadlines, tight loops, kill blockers
- 49:51 – 58:45
Resilience to macro + quick-fire: AI’s impact on knowledge work and AngelList’s long-term ‘fabric of venture’ vision
Harry challenges macro risk to AngelList’s transaction-dependent model; Avlok argues the platform is stronger than ever due to moving upmarket and gaining share, plus countercyclical product planning. In quick-fire, he predicts 2023 worse but 2024 better, expects AI to rapidly reshape white-collar work, and prefers startups to capture value. He closes with a vision of AngelList as the software infrastructure layer for venture ownership and operations, with IPO status an open question.
- •Macro isn’t existential if you’re gaining share and moving upmarket
- •AngelList has planned for bear markets with diversified, countercyclical products
- •AI trend: rapid amplification of knowledge work; significant reskilling needed
- •10-year vision: AngelList as the infrastructure/fabric for venture activity
