Rajiv Dattani

Friday rant time. Investors don't take risks. Everyone wants Tier-1 background.  Nobody is out there for the little guy. Sharing a brief breakdown of VC economics (deep dive link in comments). VC is NOT charity, it's pure financial services. The NUMBER ONE metric we're measured by (for older funds) is DPI, fancy term for how much money you've given back. 1x, shut shop and go home. 2x, better, but still bad. 3x, now you have a reason to live. (for folks wondering how young funds are measured, it's typically MOIC + IRR, basically, how are companies you've invested in performing at present). Assume a ₹200 crore seed stage fund. Needs to give back ₹600 crore at least. Ignoring fees and expenses, you can invest ₹10 crore in 20 companies, for 10% each. Since it's quite clear (Fractal, Zomato, BlueStone, Nykaa) that it takes 12-14 years to IPO a startup in India, you can't wait that long (fund life is 10 years). So you have to exit to Series C/D investors who WILL want a discount. Assume 3 rounds of dilution, leaves you with ~5%, 20% discount, so 4% held. Seed has ~75% mortality rate, so 5 companies left. These 5 have to get you ₹600 crore. At 4% held: 1 company needs to raise at ₹15000 - Just NOT happening 2 companies raise at ₹7500 - Still VERY unlikely 3 companies raise at ₹5000 - Closer 4 companies raise at ₹3750 - Possible 5 companies raise at ₹3000 - More possible ^This is also the price Minimalist was acquired by HUL, which was an OUTLIER outcome and first of it's kind for consumer startups. The above should illustrate how hard the business is. Hence when funds insist on backing folks with successful exits, strong corporate careers, or absent both, just a great educational background, is maximizing luck. At Seed, there are two things that swing it for you. I call it Numbers & Story. If you have a GREAT story (I sold my last company for ₹500 crores), you can have ~no numbers. If you have a good story (I was VP Google India), once again, numbers are lower priority. If you're 22, from a small town college, breaking it down harshly, you NEED numbers. As an investor, we just DON'T ignore amazing growth. That being said, since this is a (very hard) financial business, we're looking at any and every indicator of possible success. Experience here helps in two ways: - Have already made dumb mistakes on OTHER peoples money - Have a network who will pick up the phone and help you/become first customer #venturecapital #startups #funding

455

99 Comments

🚨 We just uncovered the strongest correlation to fund launch success that we’ve ever seen.   We work with over 60% of the world’s new VC funds and have built the most detailed database on emerging managers anywhere.   For years, we've been trying to find WHAT correlates with new and emerging manager success.   We studied every angle: demographics, geography, professional background, even personality traits. None of them consistently explained why some managers succeed and others don’t.   Until now.   My colleague and good friend, Michail Michailow ran models across hundreds of variables on all the managers we’ve worked with in recent years.  But none of them seem to reliably predict performance.    Then he looked at one we hadn’t fully considered before: how much managers use AI in launching their fund.   The results were staggering.   ➡️ The more managers leveraged AI, the faster they launched. ➡️ The more they used AI, the more capital they raised from LPs. ➡️ The correlation wasn’t just strong — it was the strongest we’ve ever seen.   The difference is massive. Managers who don’t use AI are operating on a completely different playing field, and we think the gap will only keep widening.   Venture capital is a game of speed. If you’re not using AI, you will be slower to fundraise, slower to close, and you may never reach the momentum needed for your fund to truly take off.   AI has become a DECISIVE factor in who breaks through and who never makes it.   If you want to see which AI tools these top-performing emerging managers are actually using right now, drop a comment below and I'll share more!  

118

50 Comments

Can you build startups with fewer people & lesser capital? Yes and no! I’ve now been investing for the last 12 years. Every cohort gets sharper, more prepared, more efficient. The startup journey that was largely first principles-based, is now a lot more playbook-led. And that’s showing up clearly in how companies are being built today. Looking at companies started between 2010 to 2016-17, it was quite normal to see 20-30 people in the team around $500K ARR and 50-60 people to reach $1M ARR. Why? Most of these companies were targeting SMBs & the GTM motion wasn’t PLG, but largely sales-driven. That sales motion wasn’t too well understood from India, with lots of trial & error. Founders were still figuring out outbound, how to run LinkedIn/Google campaigns, how to build an inside sales team. So, each function needed more people + iterations. But that has changed. 1 – India has moved up the experience curve. Founders today have much more clarity on the basics when they start out. There are enough existing founders to go to as mentors & enough operators to hire so that you do not have to discover the journey necessarily through your own mistakes. 2 – Tooling has come a long way. There’s even a term now, called “GTM engineering”. People are thinking of the GTM function almost like a product. What tools should I use? How do I engineer GTM internally? 3 – Building software’s become a lot easier. Even before GPT, dev-cycles were shortening. Post-GPT, the number of developers required & time taken to build has gone down sharply. Velocity of shipping has gone up sharply. So what does that mean? I’d say the number of people needed to hit early milestones is roughly half of what it was a decade ago. But this efficiency largely holds true in the 0-1 phase. As companies scale beyond $1M ARR (into $10M/$25M), complexity increases, teams grow, and hiring often happens ahead of revenue. Inefficiencies start to creep back in. The real question is: can we sustain this early-stage efficiency as we scale, not just in people, but in capital as well? That is the harder part. On a recent trip to the Bay Area, I met a number of founders & investors, and the unanimous view was that the cost of building a startup has not come down. Primary reason is that engineering talent has become costlier, and that is when it is available, particularly the cutting edge AI talent. While the number of engineers have reduced, cost has not.  And competitive differentiation now requires companies to also build/ship a lot more to stay ahead of others. Same is true for GTM costs. High inflation + interest costs, high economic uncertainty, obsolescence threat due to AI + a clutter of options has driven sales costs to be even higher than before. Avg sales & marketing efficiency of public software companies has dropped from 0.8x to 0.4-0.5x in the recent quarters (for every dollar of sales/marketing spend, they sell software of $0.4-$0.5). Would love to hear your thoughts!

186

39 Comments

Turns Out… My Fund LP Portfolio Beat My Direct Portfolio 😅 Maybe I’m not a bad investor — maybe FoFs aren’t the fallacy we think they are. I’ve been reading several great pieces lately and reflecting on a touchy subject: venture funds of funds (FoFs). Hamilton Lane has hammered this home for years: venture isn’t really an “asset class” — it’s an access class 🚪. • 📊 VC has the widest dispersion in private markets. • ⏳ That dispersion hasn’t narrowed, even as VC scaled — the power law still rules. • 📉 Drawdowns cut deeper here than anywhere else. Translation: pick the right managers and you look like a genius; miss them and you look… well, like me on some of my direct deals 🙃. That’s why the Pattern Ventures benchmarking caught my eye: across 2005–2019 vintages, FoFs (yes, even with “double fees”) actually: ✨ Beat direct VC in the median quartile in 14/15 vintages (+0.68x TVPI). ✨ Outperformed the bottom quartile every single time (+0.89x) — and never fell below 1x. ✨ Smoothed returns, with dispersion nearly halved (0.58x vs. 1.21x). Confession time: even in my own portfolio, my LP fund investments have outperformed my direct ones. Not sure if that makes me a bad direct investor 😅 … or just a better FoF LP. Either way, it taught me something about where my real portfolio chops lie. In hedge funds, double fees kill. In venture, access math beats fee math. So the real question for LPs isn’t: “Do we avoid FoFs?” It’s: Do we build selection capability — or buy it? Because in venture, average is the most dangerous place to hide.🚫 🔖 David Dug Skylar Anurag John Samir #VentureCapital #PrivateMarkets #FundofFunds #LPInsights #VCInvesting #PowerLaw #Dispersion #InstitutionalInvesting Asilica | Redshift Capital | Allocate

118

9 Comments

Three kinds of AI startups I’m especially excited to invest in and work with these days: 1. The Surprisingly Valuable Niche Workflow — small enough for OpenAI not to care about; big enough to make $100M per year. Sooo so many opportunities out there, from logistics to finance to healthcare to construction to manufacturing and on and on. 2. The Full-Stack Software Company — founders are realizing they can move up the value chain, to not just sell software to other businesses, but use their own software to provide end solutions to customers. Our own portfolio has inspired us here, from industrial solutions to life science to agrochemical to admin services and more. 3. The Bold Consumer Bet — it’s time for a new generation of consumer mainstays that are AI-native and fit the way people use software today. We don’t often invest in consumer apps, but we have seen great examples of this in language learning, personal finance, fashion, dating, social, messaging, etc. All three are great examples of building at new frontiers of the software-addressable economy, where advances in AI have opened up new opportunities for software founders. A good litmus test for this thesis: “If I had this idea 5 years ago, could I build it?” If the answer is no or maybe, you’re on the frontier. If you’re building something cool in one of these areas — or anywhere else on frontiers of the software-addressable economy — please get in touch with us at character.vc We’d love to learn from you. And we’d love to share how we help founders bridge the gap from possible to PMF in these categories.

271

32 Comments