I work in SaaS. Here are the fears, debunked one by one:
1. People will vibecode their own SaaS
Wrong. Cheaper software has always existed alongside the SaaS incumbents. People pick the incumbents still for a reason. Any growing business, especially at scale, would do better to focus on growing their core business rather than to dedicate expensive resources (engineers) to go on sidequests to vibecode all their SaaS.
2. AI will replace humans, thus cannibalizing seats.
What we're seeing anyway is that people are actually spending ***more***. Because all the SaaS companies are exposing their workflows & data via API/CLI/MCP, all the humans are now accessing all of the functionality through their native chat clients like claude & chatGPT. what happens here is because it's so easy to interact with SaaS (you don't even have to login to ask "what's going on with this latest lead?") you end up asking and doing ALOT of work. AI credit spend is going way up, WITH seats.
There *is* a doomsday scenario here - if AI replaces all white-collar workers entirely, then yes, seats are dead. But if all white-collar workers are out of work, we'll have bigger problems as a society.
3. The AI labs are coming for SaaS.
Possibly for some. Claude Design is certainly coming after Figma. What's less likely to be replaced are the systems of record (e.g. CRMs like Salesforce and Hubspot, HCM tools like Workday, Oracle, SAP) because ripping these out is really painful, and any security lapse of any sort (especially from a vibecoded tool) is tantamount to organizational suicide (think banks, etc.) We have an investor deck that shows a survey of 500+ companies asking them what they're thinking of replacing in the next 5 years. The lowest category was systems of record.