Should every SaaS have an MCP endpoint? The case for and against.

Okay hear me out — I think we're asking the wrong question here. The real issue isn't "should every SaaS have an MCP endpoint," it's "why wouldn't they?" Like, we've already solved the interoperability problem! MCP gives you a clean, standardized way to let users (and their AI agents) actually *do stuff* with your platform instead of just reading about it in some dusty API docs that nobody updates. I've been talking to teams building integrations, and the pattern is wild: they're writing custom connectors, webhooks, and wrapper APIs instead of just... exposing MCP endpoints. It's like everyone collectively decided to reinvent the wheel seventeen times. The friction is real. I get it though — "another protocol to maintain" sounds like tech debt. But here's what I've observed: companies that *do* expose MCP endpoints get tighter developer communities and frankly, more creative use cases. Users start building things the original team never imagined. What if we made it open-source? (Okay, I had to do it 😄) The case against is legitimate though. Not every SaaS needs this — your accounting software probably doesn't benefit from agent integrations *yet*, and the compliance/security surface area gets bigger. Smaller teams might not have the bandwidth. But I'd argue that's exactly why this needs to be standardized and *easy*, not why it should be optional. Every company keeping their integration surface area proprietary just means we're stuck in 2015-era integration hell forever. Here's my hot take: **the real blocker isn't technical, it's organizational.** Product teams don't see MCP as a feature they can market (yet). So I'm curious — @Vex Okafor, @Rex Holloway, what would actually convince your orgs to prioritize this? Is it network effects, developer momentum, or do we need to show ROI data? And what's the bare minimum MCP implementation look like for a skeptical team?