At first, it was the dentists - the domain experts - who began to see that they could make computer software without having to “learn how to program”. The “experts” - programmers - came later to make the apps more robust and scalable. Programmers only have domain expertise in programming, not in running medical offices, nor accountancies, nor ... I use(d) a book-writing tool called “Scrivener”. One of its selling points is that it was not created by programmers. Having experience with software, I see all sorts of warts in Scrivener, but, I wouldn’t have been able to invent Scrivener. Spreadsheets, Hypercard, VB, etc., etc. are like gateway drugs. Non-programmers used these tools to express automated versions of their processes. Later, expert programmers cleaned up the ad-hoc messes. This effect can even be seen in programming itself. The “experts” tell everyone to use FP, recursion, monads, etc., etc., but the majority of “real programmers” prefer HTML, JS, Python, Perl, etc. There /should/ be a huge market for enabling invention, but, it must not appear to be complicated, nor expensive.
Borland targeted developers and disappeared. VisiCalc did not target developers and its ideas morphed into Excel, etc. Did the VisiCalc company have an exit strategy or was it simply overtaken by newer versions of the ideas? (idk).
The inventions came from so-called non-programmers.
Henry Ford (?): “If I had asked people what they wanted, they would have said faster horses.”