Can AI-native sales stacks disrupt Salesforce's $200B? Owning the sale, not the CRM

CCgong · x · 2026-09-18

A long analysis posted while stuck in Dreamforce traffic argues Salesforce ($200B market cap, 27 years old) is vulnerable to a wave of AI CRMs with different flavors: rip-and-replace (Attio), full-stack sales OS (MonacoGTM), and agents layered on existing CRMs (rox.ai).

The core thesis: the opportunity isn't a better CRM, it's owning the sale. Deals increasingly live across texts, in-person events, email, calls, Slack and contracts—whoever captures the traces of decisions leading to a closed deal builds a learning loop Claude can't reconstruct from Salesforce's fields. Sales is a great learning loop: high-frequency work where judgment matters, so a good AE can grade an agent's output in seconds.

Salesforce's schema is a tax for relationship-driven jobs (recruiting, VC, real estate, agencies), and seed-to-Series-B companies can't be served profitably with a Salesforce implementation—which is why forward-deployed seller + software models like Monaco's are landing. Its 150k customers won't churn tomorrow, but the next 10k AI-native companies running on 5 humans and 50 agents will never install it. As for companies building their own CRMs: they end up building a schema, which was never the hard part—the hard part is integrations that break weekly, licensed external data, deliverability, and a learning loop sharpened by ten thousand teams correcting the same agent.

Original post →

More from Companies & People

Companies & People channel →