Hariom Palkar

Thrilled to share this milestone collaboration between Fynd and BeyondSeed! 🚀 At Fynd , we always believed that innovation truly flourishes when the right ecosystem comes together. With this partnership, we’re inviting startup founders who are ready to take their ventures to the next level to join us. Why this is special: AI-Native Future: Get hands-on with building AI-first solutions that can define the next wave of products. Tech + Backend by Fynd: Leverage our robust platform and infrastructure to scale faster without worrying about the heavy lifting. Funding & Acceleration via BeyondSeed: Access capital, mentorship, and a structured program that helps you learn, pitch, and execute quickly. If you’re a founder looking to supercharge your journey, this is your chance to immerse in a program that doesn’t just teach—it accelerates your vision into reality. Apply here: https://lnkd.in/duStjgDe Let’s build the future together. 💡 #BeyondSeed #AI #StartupGrowth #Partnerships #Innovation #Founders Fynd Boltic BeyondSeed Farooq Adam Sreeraman Mohan Girija Ragini Varma Ronak Modi Jigar Dafda Kushan Shah Kuldeep Mirani Mili Shah

14

1 Comment

IRCTC is one of the highest-traffic transactional platforms in India. Millions of people use it to book trains, manage journeys, and handle payments tied to personal ID data. They recently launched a beta redesign, and it's worth looking at closely (seriously). Because it shows something that happens often when large platforms redesign: The interface gets "cleaner", but functionality greatly disappears. The beta navigation is less cluttered. And it does have real gains: → The date selector now shows a full week with train counts per day, instead of making users click "Next Day" over and over. → Quota and concession fields are labeled and explicit, where before they were easy to miss. But the beta also removes the multi-date availability grid that lets users compare waitlist status across several days for the same train and class in one view. It moves filters from a persistent sidebar to a pop-up modal, which means an extra click every time you want to narrow results on a high-frequency interaction page. And the language toggle is gone. The meal booking entry point is gone. The AI chatbot assistant is gone. And the hero banner uses AI-generated imagery on a platform where visual trust directly affects whether users feel safe entering their Aadhaar number and making a payment. This is one of the most common mistakes in large-scale redesigns. Designers focus on making the interface look modern and end up stripping out the things that made it functional. They forget that a redesign should make the product easier to use, not just easier to look at. We did a detailed walkthrough comparing the current site and the beta across navigation, homepage, and search results (swipe through below). The goal was to document what improved, what regressed, and what should be reconsidered before the beta moves past testing. Check it out here:

8

We trashed 1000s of lines of code in our 2nd week of Y Combinator, to restart completely from zero Necessary. But absolutely terrifying 💀💀 Were we about to go through the entire YC batch not actually knowing what to build? It because obvious to us that no one wanted really what we first built (VMs for developers to build AI computer use agents) So we were about to pivot, until we got trajectory defining advice from our YC partner Jon Xu ✨✨ He told use to stop coding and go talk to 100 devs in tech/vertical SaaS for 🏥 healthcare, 💰 finance, 🚛 supply chain, etc. who were trying to automate things He’s a wizard. He knew we were missing something out there 🧙🪄 We already talked to a ton of devs interested in computer-use AI, but we were talking to the wrong people So we went all in: cold emails, LinkedIn messages, Reddit threads. We even paid devs on Upwork to talk to us. And it clicked 💥💥 Devs wanted reliable, desktop automation tools that could autonomously handle edge cases and dynamic tasks. That was their only way to integrate with a large subset of the digital world: API-less web/desktop apps. They wanted it fast, affordable, plug and play, and ✨✨ just worked ✨✨ Computer use AI was the first breakthrough that made that possible, so we got cooking to leverage it to make that dream a reality By the end of the batch, more than a dozen vertical SaaS companies were using Cyberdesk (YC S25) We wrote tons of code before talking to enough people the first time, and nobody wanted it. We talked to a ton of people before writing a singe line of code for the second time, and people actually wanted it Talking with the right users can really be more powerful than you can imagine. 

92

12 Comments

Most PMs know the metric names. Few know which number to trust right now. Every product has a different north star depending on its stage, model, and user behavior. Tracking all 15 at once is noise. Knowing which 3 matter today is the skill. 𝟏. 𝐀𝐑𝐏𝐔: Revenue per user per period. Tells you monetisation health. 𝟐. 𝐂𝐀𝐂: Cost to acquire one customer. Must be justified by CLTV. 𝟑. 𝐂𝐋𝐓𝐕: Lifetime value of a customer. The real unit economics number. 𝟒. 𝐂𝐡𝐮𝐫𝐧 𝐑𝐚𝐭𝐞: Users lost over time. 20% monthly = broken retention. 𝟓. 𝐂𝐨𝐧𝐯𝐞𝐫𝐬𝐢𝐨𝐧 𝐑𝐚𝐭𝐞: Step-to-step funnel ratio. Diagnoses drop-off points. 𝟔. 𝐌𝐑𝐑: Monthly recurring revenue. The SaaS growth pulse. 𝟕. 𝐃𝐀𝐔/𝐌𝐀𝐔: Daily and monthly active users. Defined by core action only. 𝟖. 𝐃𝐀𝐔/𝐌𝐀𝐔 𝐑𝐚𝐭𝐢𝐨: Stickiness score. High ratio = habitual product. 𝟗. 𝐒𝐞𝐬𝐬𝐢𝐨𝐧 𝐃𝐮𝐫𝐚𝐭𝐢𝐨𝐧: Time spent per visit. Matters for engagement products. 𝟏𝟎. 𝐁𝐨𝐮𝐧𝐜𝐞 𝐑𝐚𝐭𝐞: Users leaving without acting. High = friction or wrong fit. 𝟏𝟏. 𝐑𝐞𝐭𝐞𝐧𝐭𝐢𝐨𝐧 𝐑𝐚𝐭𝐞: Users who come back. The most honest product health signal. 𝟏𝟐. 𝐀𝐜𝐭𝐢𝐨𝐧𝐬 𝐏𝐞𝐫 𝐒𝐞𝐬𝐬𝐢𝐨𝐧: Engagement depth per visit. Likes, plays, clicks. 𝟏𝟑. 𝐍𝐏𝐒: Net promoter score. Customer loyalty on a 0-10 scale. 𝟏𝟒. 𝐀𝐎𝐕: Average order value. Drives upsell and basket-size decisions. 𝟏𝟓. 𝐂𝐓𝐑: Click-through rate. Ad and CTA effectiveness in one number. A number without a decision it changes is just noise. Save this. Which of these 15 do you track most closely in your product right now? Follow Bhavishya Bharadwaj if you're figuring out product or AI without a coding background. That's exactly what I post about.

70

15 Comments

If I were a Product Manager at Swiggy , I wouldn't start by asking: “What feature should we build next?” I'd ask: “Where is the biggest friction between user intent and a successful order?” One hypothesis I'd explore: choice overload. A user opens Swiggy hungry. Hundreds of restaurants. Thousands of dishes. Different prices, ratings, offers and ETAs. Sometimes the problem isn't finding food. It's deciding what to eat. My idea: Smart Meal Decision Instead of endless scrolling, ask: “What are you in the mood for today?” Then use signals like: • Time of day • Budget • Cuisine and dietary preferences • Past orders • Delivery ETA • Restaurant reliability And surface 3–5 high-confidence options. The MVP would be simple: Intent → Recommendations → Key Info → Order → Feedback No massive redesign. No complicated AI assistant. First, validate whether helping users decide faster actually creates a better ordering experience. What would I measure? My North Star: Successful, satisfying orders per active food user. Supporting metrics: Conversion Time-to-order Repeat orders Cancellations Refund/contact rate Restaurant conversion Contribution margin Then A/B test the experience and watch the guardrails. Because personalization also creates risks: Recommendation bias. Reduced restaurant discovery. Filter bubbles. Impact on partner economics. The goal isn't: More recommendations. It's: Better decisions. My PM approach would be: Problem → Evidence → Hypothesis → MVP → Experiment → Outcome For Swiggy, I'd want to move the experience from: Endless choice → Confident choice. Because sometimes the best product experience isn't giving users more options. It's helping them find the right option faster. This is a hypothetical product case study, not a description of an announced Swiggy feature. If you were the PM at Swiggy, what problem would you solve first? #ProductManagement #ProductManager #ProductThinking #ProductStrategy #Swiggy #ProductCaseStudy #ProductDiscovery #UserResearch #Prioritization #Experimentation #ProductLeadership #PMInterview

1