This is the founders dilemma today as well funnily enough. We built a front-end (MVP sticky tape) and are about to win some pilots. We've just plugged our whole platform into Claude (you can control the whole thing from there now) and its just way more powerful.
We're about to convert our first pilots But our clients are too old fashioned to buy Claude - and it appears we can't resell it. Its a bit icky to say 'buy this other platform to use our platform' and the distribution doesn't exist through Claude (yet).
So here's the quandary - do you say 'we're investing in the SaaS UI which will be defunct soon' to win the deals? Or do you build for Claude only and risk the pilots?
Ultimately when people are paying they want you to deliver the product with a bow on top - especially some of the old fashioned paper-based industries we're dealing with.
I think your clients being too old-fashioned to buy Claude directly is the moat, not the problem. You're not reselling an API, you're selling the outcome in a form paper-based industries will actually adopt. The wrapper is the product for them. I'd build whatever gets the pilots signed and let paying customers tell you what's actually missing.
This is the founders dilemma today as well funnily enough. We built a front-end (MVP sticky tape) and are about to win some pilots. We've just plugged our whole platform into Claude (you can control the whole thing from there now) and its just way more powerful.
We're about to convert our first pilots But our clients are too old fashioned to buy Claude - and it appears we can't resell it. Its a bit icky to say 'buy this other platform to use our platform' and the distribution doesn't exist through Claude (yet).
So here's the quandary - do you say 'we're investing in the SaaS UI which will be defunct soon' to win the deals? Or do you build for Claude only and risk the pilots?
Ultimately when people are paying they want you to deliver the product with a bow on top - especially some of the old fashioned paper-based industries we're dealing with.
I think your clients being too old-fashioned to buy Claude directly is the moat, not the problem. You're not reselling an API, you're selling the outcome in a form paper-based industries will actually adopt. The wrapper is the product for them. I'd build whatever gets the pilots signed and let paying customers tell you what's actually missing.
This is totally changing how I interact with my computer