>you know only how much %% of a company is offered to you at the moment, not the future exit value.But this "future exit value" is the same unknown for co-founders, angel investors, VCs, etc. Employees share the same unknown as everyone else as to what the eventual IPO or corporate acquisition valuation could be.
E.g. if an angel investor buys 5% of the company, he only knows his % ownership at that moment and not the eventual IPO price in 7 years. The angel also doesn't know what his eventual dilution will be. It really doesn't matter.
The angels and VCs ultimately cares about shares_times_price and that it has grown multiples of their investment. They already know they will get diluted because every owner gets diluted in the normal course of events. Selling equity to help it become a more valuable company means getting diluted.
E.g. When Accel Partners invested $12 million for 15% equity of Facebook in 2005, their investment got diluted when Microsoft later invested $240 million to buy 1.6% equity in 2007. That's how it mathematically works and it's normal. Dilution is a downstream math calculation side-effect of selling equity. The Accel partners didn't complain that they got screwed and owned less than 15% because the Microsoft investment diluted them. Also, neither Accel nor Microsoft knew what the eventual IPO price would be in 2012.
>how much %% of a company is offered to you at the moment
To be more precise, employees are typically offered a number of shares or options to buy shares instead of an explicit % of ownership. The % of ownership is a derived calculation at a point in time that's based on the number of shares. From that moment on, it's the price of the shares that ultimately matters and not the dilution because everybody's % of ownership will decrease as a normal event in growing companies that raise additional funding. Presumably, the startup hired employees that can contribute value so everybody's share price goes up which more than offsets the dilutive effects of accepting extra investment money.