Two main things to keep in mind here:
Have an MVP mindset — try and continually restrict your ideas to the minimum set of features you need to create something useful and testable for real users. Don’t worry, you can always add more later if your core product takes off.
Test your ideas early and often — more on this below, but the bottom line is there’s no use building something that costs a hundred grand if you’ve not even shared the idea with someone who’ll actually use it.
Taking a lean approach to both the refinement and scope of your idea, as well as how you actually plan, design and test your product will save you massive chunks of time and heartache.
Don’t assume, sprint.
A relatively new approach to discussing big ideas, building prototypes and testing your assumptions is called a design sprint. The design sprint was devised by Jake Knapp from Google Ventures and it’s been tested and proven 100s of times to be a super effective method for the rapid assessment of an idea. What’s really attractive about this process is that you get feedback on your concept within a 5 day period rather than months of design and development, working in a customer feedback void.
We recently ran a design sprint for a long term client of ours and the results were very interesting. While I’d love to report that the idea was well received by the testers — it wasn’t and our client had to radically rethink his approach but the net result was that we tested his assumptions, got the answers we needed and importantly saved him hundreds of thousands in product building costs.
What often kills ideas early, especially in risk-averse large tier companies, is the “shit that’s a lot of investment” factor in building software. The big advantage of a design sprint is that it significantly reduces the risks by getting rapid feedback from people who might actually use the product.