Basically it boils down to your prediction if the stock is going to be worth more than your strike price when it becomes liquid, and if you can tolerate the loss if you predict so and are wrong. Some scenarios this is a very good bet: you were early, your strike price is low, the company has had several valuations well above the strike price, the financials are healthy, and you can absorb the loss if you are wrong. The magnitude of the upside seems like it shouldn’t be a factor, if these are true, if you think the risk adjusted return can beat the market. In most cases where it is sane to exercise your options you’re expecting a multiple return, not a few points, so it basically should be irrelevant how much upside there is in absolute terms, once you’ve determined a full loss is tolerable.