The information reports USDA has let ONE of it's SalesForce contracts expire

>The U.S. Department of Agriculture, which spends about $3 billion a year on IT, uses Salesforce for “pretty much everything” including case management and maintaining its public facing portals and websites, according to chief information officer Sam Berry. But over the last year, the sprawling agency reduced its use of Salesforce for some tasks as it turned to other software suppliers such as enterprise AI firm C3.

So "The Information" (what an incredibly dumb name to be honest) is a costly paid "Deep Research" entity that does write interesting things (imo) and also publicised Anthropic's private (quarterly?) financials.

Screw their insanely expensive paid subscription. However, if you subscribe to their mailing lists, you do get exposed to some interesting bits of information.

Like the excerpt above. Here's that last newsletter in PDF format for you to read. yw:
[https://files.catbox.moe/81sc3h.pdf](https://files.catbox.moe/81sc3h.pdf)

I'm still a strong believer in regression to the mean, for sure for SaaS (I do only have a small position in SalesForce). However - and I share this quote with you from "The Information" 's latest newsletter that I just glanced over - this situation, if true, shows some proof that the latest pro-SaaSpocalypse proponents' argument has some measurable truth to it: "The SaaS erosion will happen slowly over time while cheaper and/or self-built solutions are being built". I still don't think this holds for all companies. I still think that the moat is much more than just software: data, network effects, analytics over multiple data-sets that one company just can't have by definition, (customer) service, offloading responsibility & focusing on a core mission. Just like everything else is outsourced or bought that is not central to any specific business.

Yet maybe that truth is somewhere in between NO, AI IS A DANGLING CARROT OVER A TREADMILL PROMISE and YES ALL SAAS WILL DIE.

The question then becomes, where is the border? Which companies will try to do this (and succeed) and which ones will not even try? Which SaaS is easiest to replicate and why?