One of the first SaaS products my co-founder Neil Patel and I launched was a podcasting toolset and advertising network called Fruitcast in 2005. We didn't know anything about podcasts, but we were ambitious and thought we might have the next AdWords on our hands.
We stepped in too quickly without taking the time to properly research the space. A year later, we were out over $45,000 with nothing to show for it.
That's why I believe that in today's world, you don't need to start with a SaaS product. You can start by building a much smaller product. It can be as simple as a template in Google Sheets for calculating churn, or something more complex like Moz's Open Site Explorer. Starting small forces you to learn about your customer before creating a lot of software.