Most legal work can wait for the budget to exist. A few items cannot, because getting them wrong costs far more than the advice would have.
"We will get a lawyer when there is money" is a reasonable position for most of the legal terrain, and a dangerous one for a small part of it. The skill is telling those two apart early, while both are still cheap.
Three things are worth doing before revenue. Founder and equity terms in writing, with vesting: an undocumented split is the most common fatal failure at this stage, and it detonates exactly when the company becomes worth something or somebody leaves. Lawful handling of user data from the first record, because for a company built on trust a privacy failure does not merely create legal exposure, it removes the reason anyone would use you. And clear written ownership of what gets built, including by contractors, who otherwise may own what they made.
Everything else can usually wait. But there is a difference between deferring and forgetting. Deferral should be a decision with a trigger attached: first paid customer, first hire, first raise, first new use of data. Written down, where you will see it again.
The metric that matters is not legal spend. It is the number of issues you know exist that are sitting in a blind spot with nobody watching them. Zero is achievable long before you can afford a lawyer on retainer.
This is a description of how we think about it, not legal advice, and it is no substitute for an actual lawyer looking at your actual situation.
