Harshit Tiwari

Harshit TiwariOct 5, 2026

3,000 followers. Not a viral moment. No giveaway. No engagement pods. Just showing up, sharing what I'm actually learning while building Postessia. Here's what I've figured out so far: 1. Outreach fails before the first message is read. Not because the offer is weak, but because the message feels like everyone else's. 2. Pain is not a pitch. When someone admits a problem, listen first. The moment you pitch, the conversation dies. 3. Manual beats automated, at the start. Researching one person properly beats messaging 100 people badly. 4. Honesty compounds. No fake screenshots, no made-up numbers. People can tell, and they remember. I'm still early. Still learning. Still building in public. But 3,000 of you decided this was worth following, and that means a lot. If you're a founder or agency owner working on outreach, what's the one thing that's stuck for you right now? Tell me in the comments. I'll reply to every one. Thank you for being part of the journey. 🙏

Harshit TiwariOct 7, 2026

The problem isn't your copy. It's your deliverability. I learned this the hard way. Spent eighteen months rewriting subject lines, testing send times, obsessing over word choice. Open rates were down to 27.7%. I kept assuming the message was broken. Then we checked inbox placement. Turns out 40% of our emails weren't arriving at all. They were in spam. We could've had the perfect subject line, the perfect offer—it didn't matter. Nobody saw it. Once we fixed deliverability, the same copy that "wasn't working" pulled a 5.3% reply rate. The founders who've actually nailed this all say the same thing: above 90% placement changes everything. Below 70%, you're looking at 0.8% replies on identical messaging. Here's what actually happens: you rewrite. You test. You optimize for open rates—which are broken anyway, thanks to Apple privacy filtering everything. Meanwhile, your emails are bouncing off inboxes that don't want them there. The real work isn't in the copy. It's in the decision to stop blaming the message when the problem is whether it arrives at all. Start tracking placement and reply rate. Not opens. Those numbers lie to you. Placement tells you if you're actually in the conversation.

Harshit TiwariOct 6, 2026

The problem isn't that AI is bad. It's that we keep trying to remove the human entirely. Artisan AI spent $2M on billboards saying "STOP HIRING HUMANS." Then they hired a human BDR and started building a dialer. A dialer. That's not a pivot—that's an admission. Per School of SDR's breakdown, AI SDRs got fewer meetings and a higher spam rate than humans. Not my data. Theirs. Here's what actually happened: they didn't fail because the technology was broken. They failed because they bet everything on a premise that turned out to be wrong. One problem. One solution. Remove the human. Done. Same thing is happening on LinkedIn right now. Open your feed. Half the posts sound identical. The cadence is identical. The structure is identical. And nobody's buying from identical. They're buying from the person who actually has skin in the game. I'm building an AI tool. I know this is uncomfortable to say. But here's the truth: Postessia can draft your post in your voice. It can save you 40 minutes. That's real. But the opinion—that has to stay yours. The replies have to be yours . The moment someone admits a pain point in your DMs, that's not for a tool. AI can draft. You have to mean it. The real work isn't finding a tool that sounds human enough. It's deciding which parts of you stay irreplaceable. The moment you hand all of it over—the thinking, the conviction, the actual skin in the game—you're not ahead. You're just faster at being replaceable. What part have you handed to AI that you probably shouldn't have?