Ask the magic question - 'Will you pay $X for this per month?' That differentiates ppl who find it 'nice' vs people who are desperate for it. This will also give you baseline pricing. Pitch it high and come low. If you think it's worth $100/m, start asking at $250/m.
What's an interesting revenue number for you from this? Say it's $10000. Say your price is $50/m [Are your customers ready to pay this much?]. Then you need 200 customers. Is the market big enough to support that? Can you reach this many customers?
Meanwhile, do the pipeline research. What do people look for when they want this product? What search terms do they use? Use Adwords Explorer to find out how many people in which region are searching for this - https://adwords.google.com/o/Targeting/Explorer . See how many competitors exist in this space, are they easy to find, are they targeting the right keywords [semrush.com]
If everything is positive, create a landing page, start doing some basic SEO or even SEM [if the price is low enough], to get signups, and if you're up to it, pre-orders.
Then build the product :)
P.S. I've done bits and pieces of this before - not done all the way. If you want to chat abt this, I'm prasanna 79 at google's email service.
P.P.S. I'm surprised this is not on the front page. I'd have thought this is clearly the most pertinent of HN material.