SaaSDefend: Proprietary Logic Lock for AI-Vulnerable B2B SaaS

B2B SaaS: customers churning for in-house solutions?

second hand information here, effectively what I've compiled in my head from a meetup I attended recently where this was a big topic. Proprietary algorithms or data that's difficult to acquire seem to be good moats to defend a product, as do network efects (which you mentioned in the post). All those take time and domain expertise to replicate, which can't be done quickly using an AI agent. One of the people presenting at the meetup had a HR SaaS with some prop scoring algorithms, where several customers cancelled the subscription, built their own solution and a few months later came back because Claude could replicate the UI/UX easily but not the scoring. The curious thing was that customers effectively wanted to keep their UI but have access to scoring, so the SaaS opened it up as an API and now this part of the business grows rapidly, even their competitors are using the API because the scoring algorithm is so good. Another place where a SaaS can defend itself is data that's not easy to collect or aggregate. Very similar to prop algorithms, the API to access the data might actually be the valuable part of the product in the future.

The data/onPrem sales pitch is probably the best there is, but it still isn't the silver bullet. The defense is still, and will always be, cost. What you're describing is probably more expensive than just paying for it. Take a PTO management system, they're about $1 per use per month. The rules and ui is relatively straightforward, and your ai of choice could generate one in a day. A hundred-person company could easily negotiate a $3000 for 3 years deal, maybe a little extra for some support, and for that they no longer need to think about it again. Updates will come, support queries get answered and that's it. To do it in-house, you need infra to run it, so that's \~$30 a month (one-third of your allowance). Then you need someone to manage it, handle any queries, fix any bugs, and handle general support queries. But in most places, even a fraction of a headcount will run into the 10s thousands. So it's not financially viable. Obviously, that's a very cheap SaaS, if you found something like a contract management platform where people are paying $10k+ a seat, you'll be celebrated as a hero. But they're probably more complicated than a CRUD app :(