You install an embedded chat widget and watch support costs climb immediately. Messages at ₹5 to ₹50 each, depending on the platform. Your team gets faster at closing, but your per-conversation cost stays stubbornly high until you hit a magic volume number. We mapped that number—and why it matters. The cost structure: why embedded chat looks expensive Three platforms dominate embedded chat pricing, and they diverge sharply: Intercom: ₹50 per message (roughly $0.60 USD); unlimited team members, $39/month minimum per seat. Drift: ₹30 per message (roughly $0.36 USD); similar per-seat model. Orin: ₹5 per message, flat-rate pricing—no per-seat surcharge. At first glance, Intercom's model seems criminal. A single conversation with 20 back-and-forths costs ₹1,000. Scale to 500 conversations per month and you're spending ₹10,000 just on message throughput, before seats. But message count is not the real metric. What matters is resolved conversations —and what each one generates. Revenue per resolved conversation: the number that justifies cost A resolved chat conversation typically does one of four things: Books a meeting. A prospect asks a question, chat agent answers, prospect schedules a demo. Revenue impact: depends on your close rate and deal size, but a booked meeting is worth ₹5,000–₹50,000 in pipeline. Closes a subscription upgrade. Existing customer sees your chat, upgrades mid-session. Immediate revenue: ₹2,000–₹20,000. Prevents a churn conversation. Customer support agent resolves billing confusion, contract question, or feature request in real-time. Prevented churn value: ₹5,000–₹100,000 (depending on your customer lifetime value). Qualifies a lead faster. Sales rep moves a prospect from interest to evaluation in one chat instead of three email threads. Acceleration value: ₹10,000–₹100,000 in NPV. The math only works if your team is actually trained to use chat as a revenue tool. If chat is treated as a cost center—"answer tickets, move on"—then you're paying message fees for zero return. Chat ROI flips when your team stops treating chat as a support channel and starts treating it as a sales and retention channel. The breakeven point: 500 conversations per month Here's the model. Assume: Average conversation length: 15 messages (roughly ₹75–₹750 depending on platform). Resolved conversations per month: 500. Revenue per resolved conversation: ₹5,000 (conservative; assumes 1 in 3 chats books a meeting worth ₹15,000). Monthly cost at Intercom: 500 × 15 messages × ₹50 = ₹375,000, plus 2-seat minimum at ₹39/month each = ₹78/month base. Total monthly: ~₹375,100. Monthly revenue: 500 conversations × ₹5,000 per resolved = ₹2,500,000. ROI: (₹2,500,000 − ₹375,100) / ₹375,100 = 566%. Your embedded chat pays for itself 5.6 times over. But drop to 100 conversations per month—a realistic starting volume—and the equation breaks: Cost: 100 × 15 × ₹50 = ₹75,000 + base seats = ~₹75,100. Revenue: 100 × ₹5,000 = ₹500,000. ROI: 565%. Still positive, but your cost per conversation (₹751) is high relative to ramp-up risk. The true inflection point sits around 300–500 conversations per month. Below that, you're paying message fees that make sense only if your average revenue per conversation exceeds ₹3,000. Above that, you're paying sub-₹2 per message at Orin or ₹50 per message at Intercom—and the volume makes either math work. Load time destroys ROI faster than message cost Here's what almost every embedded chat ROI calculator misses: how much the widget itself costs you. A badly optimized embedded chat widget adds 200–500ms to page load time. Google's 2021 research showed that every 100ms of delay costs you 1% of conversions. At a ₹5,000 average order value and 10,000 monthly visitors, 500ms of delay costs you: 5% conversion loss × 10,000 visitors = 500 fewer conversions. 500 × ₹5,000 = ₹2,500,000 in lost revenue per month. Your chat widget costs you ₹375,000 in message fees, but the load time tax is ₹2,500,000. The message fees are noise. Platforms vary wildly on optimization: Drift and Intercom load their full JS bundle synchronously on many sites, often adding 300ms+ to page load. Orin's chat widget uses lazy-load and async initialization, typically adding 50–100ms. Before you pick a platform based on message cost, audit its page load impact. Use Chrome DevTools or WebPageTest to measure widget injection before and after. A 300ms widget that costs ₹50 per message can destroy ₹2M in annual revenue. A 50ms widget that costs ₹5 per message pays for itself in optimization alone. When to embed vs. when to link to WhatsApp instead Embedded chat makes sense for high-volume, quick-resolution conversations: product questions, billing clarifications, feature requests. It does not make sense if your typical chat leads to a 30-minute phone call or a contract negotiation. For complex sales conversations, a unified WhatsApp inbox often wins on economics: users already expect WhatsApp, no widget load time, and your team can move convers