Yes, your chances of building something thats actually going to make money are way higher. Most obvious ideas have been tried a million times. Its the smaller businesses relying on some random problem, charge 5k a year for it, get 100 customers, sell the business and semi retire.
Data scientists in insurance. Merging the fields of CS, decision support, and actuarial. Not a product per se, but consulting services and productish APIs and managed systems.