Earlier quoted context omitted.
Or, just throwing this out there as an option, you could solve any problem for a physician's practice -- like, say, patients not complying with their treatment regimen by failing to make scheduled check ups. You might even be able to do that with a SaaS app. Granted, it won't cure cancer, but you'll save more lives at the margin than doing Zynga ads. (And, n.b., healthcare clients often have more than $500 available…
How? Making software for government is like a two year bidding process. The barriers to entry are incredibly high when it should just be a case of, "Okay, you want X Y and Z as your most critical features, let's keep the initial product small, we can add features as we go along. It'll be done in two months and then we'll iterate." Instead it's more like, "We're going to have bids for two years for the project and the…
The not-so-subtle subtext was that "I, erm, actually sell software which does pretty much exactly this." The how, in relevant fashion: create software which does appointment reminding phone calls for, without loss of generality, HVAC contractors. Read up on HIPAA requirements and talk to people who've done it before. Do some very boring paperwork and not-all-that-impressive technical work for ticking off the boxes. Start landing hospital clients, through a combination of slackadaiscal meat-and-potatoes SEO and being a lot more hungry for their business at the low end than the competition is.
I don't typically compete in or win two year bid cycles. I go after smaller projects in the $X,000 to $X0,000 a year range. The sales process often involves convincing a single nurse or office manager that I'm not some slick sales guy from the big city who is going to sell her $10,000 of software where she needs $2,000 and then never be there for her if it breaks.
Your mileage (kilometerage?) may vary in the United Kingdom.