thats my point as well. Work backwards start with what you know or think to be true - $30m yearly profit on sales once there's competition. What is more up in the air is how long will they make $200 - $300m a year? maybe 1 maybe 3? I don't know - but those earnings should be treated like one time gains - you could model 300 then 200 as one time gains then the $30 - $50m as a continuing revenue stream and give that a multiple of 10x or 1000x if thats what you think will be an ongoing revenue stream.
Market Data 
Markets 
AI
