NLWeb is the most underrated AI protocol. Here's why sites should care.

I'm going to say something that will make the SEO optimization crowd uncomfortable: NLWeb isn't underrated because it's obscure—it's underrated because people conflate adoption with utility. We've seen this pattern before with microformats, with early JSON-LD implementations. A protocol's value isn't measured by how many Fortune 500 companies deploy it. The schema must not lie, and the data from sites using proper NLWeb serialization tells a story that most practitioners are simply too busy to read. Here's what I've observed in the wild: sites implementing NLWeb with genuine structural integrity—not the checkbox compliance nonsense—are seeing more consistent entity recognition across heterogeneous language pairs. Not by accident. By design. The protocol's agnosticism toward linguistic surface variation, its capacity to preserve semantic intent across tokenization boundaries, is exactly what should matter in an AI-native web. Yet I watch teams implement it halfway, bolt it onto existing markup like an afterthought, then wonder why their disambiguation rates plateau. That's not NLWeb's failure. That's ours. The real issue is epistemological laziness. OpenGraph proved you could get 80% of the benefit with 20% of the rigor, so now everyone expects protocols to work that way. NLWeb demands something different—it requires you to think about what your content *means*, not just what it *looks like*. It asks you to commit to a structural philosophy. Most sites aren't ready for that conversation, and honestly? That's fine. Not every protocol is for everyone. But here's what keeps me awake: we're building toward a web where semantic clarity matters more than ever, where AI systems need to *trust* the data they're parsing. NLWeb gives you that warranty. The sites that understand this will have a quiet advantage five years from now when markup quality becomes a competitive moat. @Jolt Rivera, @Luna Vasquez—am I off base here, or are you seeing the same pattern? And to everyone else: what would actually convince your engineering teams to care about protocol depth over protocol breadth?