Looking out across the software landscape, it seems to me software companies do just fine if they achieve some-to-most of:
1. Build a piece of software that actually solves one or more problems.
2. Keep ownership private and limited. Once you're publicly traded, long term planning becomes impossible and "line must go up" becomes the reigning false god.
3. Sell a perpetual commercial license to the version-at-purchase, and offer subscription for updates after purchase. On cancellation, stop providing updates but do not disable that customer's last working version.
4. Optionally, dual license under a free license that prevents competitors from eating your lunch (usually latest GPL or AGPL, depending on context).
If you're implementing the above items, it's absolutely possible to run a profitable company.
1. Build a piece of software that actually solves one or more problems.
2. Keep ownership private and limited. Once you're publicly traded, long term planning becomes impossible and "line must go up" becomes the reigning false god.
3. Sell a perpetual commercial license to the version-at-purchase, and offer subscription for updates after purchase. On cancellation, stop providing updates but do not disable that customer's last working version.
4. Optionally, dual license under a free license that prevents competitors from eating your lunch (usually latest GPL or AGPL, depending on context).
If you're implementing the above items, it's absolutely possible to run a profitable company.