I have incredibly mixed feelings about this. Zero-negotiation policies are great tools for eliminating unfairness <i>if executed well</i>, the problem is that you still force the employee to end "trusting someone at their word", and <i>every company</i> claims they're going to make a fair, standard offer. Here's the thing:<p>1) Yes, being strict and almost formulaic will reduce inequality and increase probability that folks are compensated according to actual value.<p>2) But you need <i>meaningful transparency</i> around this. I guarantee you, 100% that if I had a live offer at Reddit, I could find some way to negotiate some additional crap that amounted to a meaningful compensation bump in the end.<p>3) <i>Every</i> company that does this ends up making exceptions for people the higher-up you go. Wealthfront, Stack Exchange, and now Reddit, will join the club of companies that negotiate with execs they hire, VCs, and bizdev partners, suppliers - basically, everybody except their employees. And even then only "most of the time"- there are always exceptions - just hold out for a higher comp band. Get a stronger inside referral. It's always possible.<p>Something very close to the Buffer model is the only real way to do things. You can tweak the variables, but you need strictness <i>and</i> transparency. <a href="https://open.bufferapp.com/buffer-open-equity-formula/" rel="nofollow">https://open.bufferapp.com/buffer-open-equity-formula/</a><p>Ellen, Alexis - If you're reading this, I'd love a chance to understand your challenges in crafting these policies, and see if I could offer any input from my perspective as well.<p>(Source I founded <a href="http://OfferLetter.io" rel="nofollow">http://OfferLetter.io</a> - we help engineers and other tech workers negotiate for what they're worth. I've personally had literally hundreds of conversations with folks about this.)