software, entrepreneurshipI have been working in the industry for several years and I find a lot of flaws in some of the structural analysis FEM software that I use daily. Yet, for some reason, we still use these engineering software because we are merely used to them and we “trust” them.
That got me to think: why isn’t there any better software out there? Why can’t I make one?
How can the new software be competitive against more well-established software? Should I start learning how to code or is it better to find someone who is more of an expert in CS/computer programming? I suppose learning how to program FEM is very difficult..
It sounds like you have significant barriers to entry - you cannot code & you do not know someone who can. I will make the assumption that you do not have lots of cash to get going?
This is not meant to sound too critical, but just for you to understand the challenge. I would suggest a completely different approach. As you do not have the above, build a specification for a new system. Go into exhaustive detail covering every part. Identify the limitations of existing systems and give real-life examples (best if they are other people’s/prospective buyers). Then work out the size of the market.
Once you have these two, approach software houses that may be interested in developing this and giving you a small share (make sure you get them to sign a non-disclosure agreement before you share your details/knowledge). Good luck!
I’ve been in this situation with engineering software I’ve used. You know that a decent programmer could write something that does that functionality better in a couple of weeks, right?
If you’ve never written a substantial piece of software before for others’ consumption, you should know that the core functionality is generally 10% of the coding work (homepage, data munging, all the front-end widgets, styling, interactivity, user account creation+management, access control, database architecture, deployment & provisioning, server maintenance…), and the coding work is generally like 50%±20% of the total work required to receive money in exchange (sales and marketing, legal, outsourcing).
So the two weeks actually required to create the software that solves the core problem is in fact only 5% of the work required to create the business system that transforms the software into money. I.e. you’re looking at like a year of FTE work to have a sustainable business.
Most people opt to reduce this time and risk by:
Some people opt to keep the security of their day job until the MVP is validated, thus spreading those 2 MVP months out over 6-12.
And, all this assumes that the competitive circumstances are such that the competition wouldn’t respond credibly/quickly, and your business would succeed! Hopefully that explains why it’s obvious that the core problem could be solved quickly, but few embark upon commercialising the solution.
The way to start a structural software engineering company is to start a structural software engineering blog.
Your validation question is, do people who use this software care about the problems? So MVP is whatever gets you talking to users. And a great way to get there is to write. Don’t write, “here’s what’s wrong,” but, “here’s what I do to work around problem X.”
If you can establish an audience, you can work step by step towards finding the most common problems, which will help you spec out a software MVP that will solve just the biggest hassle, beautifully.
Who will write the software? If you’ve validated demand, you’ll have no trouble finding a risk partner.
All content is licensed under CC BY-SA 3.0.