Achint Setia

OpenAI declared war on Google yesterday and 82% of Indian shoppers are already on their side without knowing it. ChatGPT now has "Instant Checkout." You can search for products, compare options, and complete purchases without leaving the chat window. 700 million weekly users can now shop directly through conversations. Within hours of this announcement: 📌 Etsy's stock jumped 16%.  📌 Shopify gained 6%. The market immediately understood what this means for traditional e-commerce. When people use ChatGPT for shopping, they convert at 4× higher rates than regular websites, and that is because users spend 14 minutes per session on Ai tools like ChatGPT as compared to Google's 5 minutes. India's position in this makes it even more interesting: → 87% of Indian shoppers believe AI tools help them shop faster, with AI-powered interactions showing higher engagement rates (Walmart Retail Rewired Report India, 2025) → 62% of Indians make purchases influenced by AI recommendations vs 30% globally (EY India, Aug 2024) Traditional shopping makes you open multiple tabs, visit different websites, and compare prices manually. ChatGPT collapses this entire journey into one conversation. For small businesses, this could level the playing field. A seller in Jaipur gets the same access to global users as Amazon does. For established platforms, the threat is real. Why would customers visit your website when they can complete everything through a conversation? The companies optimizing for AI discovery today will own tomorrow's market. Is your business ready for AI-powered shopping or still relying only on traditional search?

252

52 Comments

90% of startups don’t die from competition. They die from 4 silent traps. Trap 1: No Product–Market Fit Pets.com raised $82M in the 90s and collapsed in under two years. They sold what local stores already offered. India’s quick commerce feels eerily similar. Burning crores to deliver things your kirana already sends for free. Trap 2: Vanity Spending OurBeginnings.com spent $4M on a Super Bowl ad. Their revenue that year? Less than $4M. Today, it’s like a startup with ₹20 crore revenue dropping ₹25 crore on IPL ads. Hype doesn’t pay the bills. Trap 3: Regulation Napster had 80M users before lawsuits shut it down. Closer home, ONDC is rewriting the rules for food delivery. If Swiggy or Zomato lean too hard on data-led cloud kitchens, they could face the same fate. Trap 4: Too Early Z.com tried to be YouTube in 1999, before broadband. It died. YouTube launched in 2005, when the market was ready, and thrived. Being early doesn’t make you a visionary. It makes you broke. Most founders think they’ll die fighting competitors. In reality, the killers are invisible until it’s too late. Which Indian startups do you think are closest to falling into one of these traps?

158

60 Comments

I asked ChatGPT, What do I mention the most to you and what impression gave I given to you of them. I was quite surprised with my secrets. 1. HACA What you mention most: - Scaling the business, hitting revenue goals, and expanding into new markets (UAE, Kerala, South India). - Differentiating from competitors. - Building a strong culture (Culture Test, Culture Handbook, Founder’s Code). - Improving sales systems, marketing strategies, and content-led growth. - Enhancing student experience, placements, and product quality. Impression you’ve given: - You’re deeply committed to building HACA as a standout ecosystem, not just a typical edtech company. - You balance ambitious growth goals with a long-term people-first mindset. - You care a lot about structure, clarity, and high standards in operations, culture, and leadership. 2. Leadership & Team Building What you mention most: - Hiring key roles (Head of Sales, Head of Marketing, Head of Media). - Training and empowering leaders (Project Leads, Builders, mentors). - Driving accountability through processes (daily P&L, dashboards, weekly reviews). Impression you’ve given: - You’re very hands-on as a leader and have high expectations. - You value talent creation as much as talent acquisition (giving freshers opportunities). - You want leaders around you who are self-driven and can scale your vision. 3. Culture & Values What you mention most: - Freedom with ownership, empathy, transparency, responsibility, and client-first thinking. - Building the Culture Test and embedding cultural alignment into hiring and operations. - You dislike shortcuts, politics, and misalignment in teams. Impression you’ve given: - Culture is non-negotiable for you; you see it as the foundation of long-term success. - You’re inspired by best practices from companies like Netflix, Google, Zoho but want HACA to have its own unique playbook. 4. Content & Personal Branding What you mention most: - Your own content on LinkedIn, YouTube, and The Mallu Show. - Positioning yourself as a relatable mentor/big brother figure for career aspirants. - Scaling HACA’s organic marketing through storytelling and student/alumni success. Impression you’ve given: - You’re a strong communicator and educator, with a focus on depth and authenticity. - You want your personal brand and HACA’s brand to mutually reinforce each other. 5. Ambitions & Standards What you mention most: - Clear financial goals (₹8 Cr MRR, 10x growth in 2 years). - Building HACA to be “Global.” - Long-term product iterations (AI-first, experiential learning). Impression you’ve given: - You’re ambitious but methodical. - You hold yourself and your team to very high standards, which sometimes makes you frustrated when processes aren’t followed. - You’re very self-aware about your growth as a CEO and want to get better at delegation and strategic focus. PS: Photo was real candid. Not plandid. What impression do I give you. Let me know 🤓

186

35 Comments

𝐆𝐓𝐌 𝐢𝐧 𝐁𝐅𝐒𝐈 𝐢𝐬 𝐡𝐚𝐫𝐝 And if you’re building in India, you quickly realize: it’s less about speed and more about understanding how things actually move. 1. There’s no single buyer. Product might love you, but Ops is risk-averse. Compliance wants a legal opinion. Tech wants infra docs. Everyone has veto power, and timelines don’t move till all green lights align. 2. Bottom-up sounds democratic, but is painfully slow. You find a champion, you do a free POC, you show result, and still nothing moves. Because decision-making sits way higher. And often, the folks you start with aren’t even in the room when final calls are made. 3. Top-down works better - if you can get in. Senior leaders care about org-wide impact and risk savings. If your pitch solves a board-level problem, you skip 6 layers of internal selling. Otherwise, you’re just another tool in the pile. 4. Security, infra, and compliance are not “later” conversations. If you’re not enterprise-ready from Day 1, most BFSI players won’t even let you inside the building. 5. Everyone’s overloaded. The risk lead is doing 3 roles. The Ops lead is managing 10 vendors. If your solution isn’t dead simple to understand - and implement - they’ll thank you and never call back. This is a market that tests your GTM clarity more than your tech Get it right, and you build deep moats. Get it wrong, and you waste a year chasing the wrong door. #AI #Founderslife

54

2 Comments

everyone wants to feel “in the know” on ai. few actually touch the tools. but you don’t learn to swim by debating water temp - you jump in. into the deep end. last weekend with Code with AI, 🥃 Agrim Singh and i taught a room full of ppl who never touched code before. folks who usually build skyscrapers 🏢 or theme parks 🎢 suddenly building software. and here's the thing: being “in the know” about ai means accepting that it’s MESSY. you’re going to npm install the wrong thing. you’ll get stuck in infinite loops. you’ll feel like you've been dropped into an alien world 👽 with no language to describe your bug. but that discomfort is necessary if you want to speak from real experience instead of empty speculation. because only builders don't put ai on a pedestal - they know it intimately, flaws and all: - when you pollute your context window and the model hallucinates - when a single missing sentence in a prompt derails your entire chain - when you ask ai to do 10 things at once and it does all 10 badly - when the limits of ai mirror the limits of human cognition - rendered in code and only by building (and breaking, and rebuilding) do you truly live the real power of these tools - not just hear it echoed in panel debates. and yet our students shipped anyway: - an ios app to track stools (yes... 💩 ) - a singapore transit optimizer factoring weather, distance, and price - one person even built 12 apps in 24 hours (shoutout Jelili Odunayo Kazeem!!) this is what “exposure to ai” actually looks like: messy, hard, exhilarating. if you want to dive into the deep end too, waitlist for the next Code with AI class is in the comments.

124

13 Comments

Ever watched your startup go from calm to chaos in 48 hours? 48 hours ago, Kartik and I thought we had steady growth figured out. Nothing crazy, just consistent users. Boy, were we wrong. Today? Our servers are literally melting 🔥 Yesterday alone, Omnara processed 30,000+ messages. Our single cloud instance? OBLITERATED. We had to emergency scale to FIVE Render instances just to stay online. The catalyst? One Hacker News post hit #2 and triggered an avalanche. Lessons from our first viral moment: → Always have a scaling plan (we definitely didn't) → Server costs can 10x overnight (painful lesson) → Community-driven growth is incredibly powerful Easily the wildest 48 hours of our startup journey. Want to see what caused all the chaos? Check out our open-source repo: https://lnkd.in/gJjjGtVc

289

39 Comments