1) make an MVP. minimum viable PRODUCT (not prototype)
2) Pick an idea that you would pay to use (product champion). If you are not the target market, you need to find a product champion who will join your team prior to the MVP creation.
3) Do things that don't scale. don't future proof your MVP, just make it so you can validate that your product has a market beyond your product champion.
4) Create a website for your MVP and make sure you can run it at low/now cost for at least 6 months before you decide to abandon. Marketing is hard and product discovery might be the biggest challenge you face. As long as you have more customers every month you are doing ok.
5) Even if you give your product away for free during MVP/beta, figure out some way that customers who want to pay you can. This is very valuable for determining product fit. if nobody wants to pay, figure out what would make them decide to, and use that for determining how to pivot.
source: myself, I made/make phantomjscloud.com