Fewer people are talking about what it actually takes to do it - the decisions that aren't clean, the ones you sit with for a year before you make them.
This is what ours looked like.
THE BRAND QUESTION
SaaStock ran for ten years. It meant something. It still does, to a lot of people. But the name was doing quiet damage. "SaaS" in the title positioned us inside a category that's in the middle of being rebuilt, and the founders we wanted to serve were already moving past it.
I spent about twelve months going back and forth. Add a .ai? Reposition the website? We tried the repositioning route. It helped. But it didn't feel like enough, and when you're doing something cosmetic when the situation calls for something structural, you usually know it.
Launching Shift AI as a new event, a new brand, a new company felt like the honest version of the decision. It cost more. It was harder to explain. It was right.
THE TEAM QUESTION
My hand was partly forced. Closing SaaStock meant redundancies - real people, real careers, a team I'd built over a decade. But starting Shift AI from scratch also gave me a choice about what to build with.
I wanted a small team. I wanted people with AI fluency, not people who knew the tools, but people who thought differently about how work gets done. I brought a couple of people from SaaStock across. The criteria had changed though. That mattered.
THE TECH STACK QUESTION
SaaStock had around 60 SaaS subscriptions. We trimmed the list every year and somehow always ended the year with more tools than we started with. The biggest line was HubSpot at £3,500 a month, which, through a combination of testing AI agents and not fully understanding what we were buying, we managed to push to £8,500 a month before we caught it. We weren't using most of the platform.
Closing the company and starting fresh meant we got to make this decision with clean hands. We moved to Attio for CRM and Brevo for marketing. The savings are significant. For sales we're starting to layer in Crono, Perlon AI, and Outcraft AI. A Claude Team account handles more of our day-to-day work than I expected, including design.
THE BUILDING QUESTION
With SaaStock, I ran an events company. I never built a SaaS product. I always observed that world from the outside.
Roy Mann, CEO of Monday.com, talked about going back to coding to really understand what AI can do - not as a demonstration, not for content, but to actually learn. That landed with me.
Last Sunday I spent three hours building with Claude Code and used 375,000 tokens. I didn't ship anything meaningful. I learned more than I expected. Sunday mornings might become build time; not because I'm going to become a developer, but because there's a kind of understanding you only get by doing it yourself.
That's the Shift we're going through. Messy, unfinished, occasionally expensive, and heading somewhere worth going.
ONE QUESTION:
How are you Shifting right now?
Hit reply. I'm building a broader dataset from these conversations. If enough of you answer the same way, it becomes next month's benchmark issue.
Alex
Founder, Shift AI
P.S. Shift Europe is the only conference where software founders who are actually navigating the AI platform shift come together - no panels of pundits, no filler sessions. Tickets go live this Thursday and waitlist gets first access. That matters more than it sounds, because we're hard-capping attendance at 1,000. Curated, not crowded. Founders who figure out the new model in the next 12 months will pull away from those who wait. Shift Europe exists to accelerate that process. Join the waitlist and I’ll send you the link when we go live.
The Shift · Published every Tuesday · shift.ai
You're receiving this because you subscribed. Unsubscribe any time.