Charge tuition for training and offer "student loans" with no payments or interest while you're working for the company which are forgiven if you stay for five years or whatever. Then if somebody wants to poach you right away, that's fine, they can just pay off your training loan.
You could also just forget about the loan forgiveness, charge "tuition" unconditionally and then pay more to make up for it.
Pay higher than others and employees won’t leave you.
Make delayed payments part of compensation, so employees get the money only if they stay long enough.
Don’t hire employees who change jobs too often.
There are many ways to keep employees from leaving without damaging their careers.
If you train me and I get better at my job, I deserve more money. If you don't give it to me, I will work for someone else who will pay me more.
Good companies train their employees and also do what it takes to retain them. It's a two-way street. Why should you get to extract more labor value from me without paying me more?
CFO: We loose money if we train an employee and he leaves. CEO: What if we don‘t train them and they all stay?
I think the standard argument isn't trade secrets. The standard argument is that you want to incentivize employers to train employees.
If I, as an employer, train you for 6 months, then you leave at month 7 for a higher salary (which you can demand cause you have more training/experience), I lose out. So I have no incentive to train you. So training or hiring of people without experience never takes place (or does so at a lower rate).
If you are someone without experience, and there was a way for you to credibly commit to not moving to a new job (e.g. by signing a non-compete), then you might want to make that bargain - it's worth it for you to sign away some freedom in order to get some advantage (which is kind of the basis for all contracts).
That's the theory, anyway. I have no idea how much it holds up in the real world.
> The tech industry in California does just fine without noncompetes.
Obviously, the counterargument is that it could be doing even better, we just don't know it.
Note that if the standard argument is correct, then there's a chance that the biggest group of people hurt by lack of noncompetes are people without experience. You might not necessarily be feeling their pain, depending on your social circle (e.g. if you're a working developer in an SV company, you're surrounded by the people who did make it, not by the possibly numerous people who faced closed doors because of this).