Starting a Business: The Legal Work That Actually Matters
Most founders overdo legal, complicating their business and sapping energy. Here is the minimum viable way: which entity, dividing ownership, and the one deadline you can't undo.
Written by

Ryan Wenger
Startups
Ryan is the CEO and cofounder of Inhouse. He exited his last startup and has nearly 20 years of experience as a corporate lawyer.

Key takeaways
Opt for the minimal viable legal structure to channel the maximum time into the business. Just like a product, you want to build a lean structure that you can scale later.
Whether you plan to raise institutional capital in the short term decides your entity. C corp if yes, LLC if no. Converting later is usually fine.
Separate economics from control. Who gets the money and who makes decisions do not have to match, and treating them as one question is what creates deadlock.
You do not need to know what each founder is worth to split equity.Â
File your 83(b) election within 30 calendar days of receiving vesting founder stock. Missing this can’t be fixed.
On this page
Founders usually overdo legal
Every day, founders ask our law firm to review or create their initial legal package. As a lawyer, I get puzzled looks when I suggest pairing it down. The advice contradicts their favorite dramas and probably every lawyer they have spoken with. Most lawyers only see companies at the beginning and the end of the journey.
I say it with confidence because I have been in their shoes. I built and exited a venture-funded startup before founding and running Inhouse. Before that, I practiced corporate litigation at a big firm so I also know how to lawyer up.
I recommend lean legal because a new business's cause of death is almost always nonlegal: no product-market fit, bad timing, a bad team. Legal defects are only the 17th of the top 20 causes of failure.Â
Further, defects are often curable, particularly once you have revenue. When I founded my last company, I gave half of it at the onset to a mentor and angel investor whose role diminished over time and became what investors call dead weight on the cap table. When we raised our Series A, he agreed to give up most of that equity at the behest of the VC firm. Money talks.
This means you don’t have to worry about everything that can go wrong.
But a few things matter enormously, and my job here is to outline them for you, so you can go back to not thinking about legal work.
What the minimum actually includes
Four need-to-haves and one nice-to-have:
An entity, with articles and bylaws or an operating agreement. Commodity documents. Describe your business to Inhouse and it will produce them.
Written ownership terms, if anyone else owns part of this. The highest-value document you will sign.
The 83(b) election, within 30 days, if your stock vests.
IP assignments from anyone who writes code or builds brand assets, signed before work starts. Without one, the contractor owns what they made.
Nice to have: a privacy policy and terms, if your site collects anything. CalOPPA applies from your first visitor with no revenue threshold. It is on the nice-to-have list because enforcement follows scale, not because the rule waits for you.
LLC or C corp, and where
One question settles the entity: are you raising institutional capital in the short term?
If yes, form a C corporation. Funds will not buy LLC membership interests, SAFEs convert into C corp preferred, and only C corp stock qualifies for the Section 1202 QSBS exclusion, which the 2025 One Big Beautiful Bill Act raised to $15 million of excluded gain per issuer with a tiered holding period starting at three years.
If no, form an LLC. It avoids double taxation, passes profits and losses to your personal return, and carries no board or meeting formalities. Once profitable, ask your accountant about an S corp election, which is a tax election rather than a third entity type and can cut self-employment tax. It also bars entity shareholders and multiple stock classes, so it is wrong for anything you intend to raise on.
Converting later is easier than you have been told. An LLC-to-C-corp conversion is generally tax-free under §351 for a company with no debt. The real cost of waiting is that your QSBS clock restarts when the C corp issues shares.
Incorporate where you operate, unless a priced round is close. Delaware adds franchise tax, an annual report, a registered agent, and a foreign qualification back home. Check your own state first: every California LLC owes an $800 minimum franchise tax from year one, while new California corporations get a first-year exemption.
Dividing ownership and control
This is where the expensive mistakes live, and the fix is the same whether you are two engineers or two sisters opening a salon.
Separate economics from control. Who receives the money and who makes decisions are different questions, and collapsing them is what produces paralysis. You can hold equal economics with one person holding a tiebreaking vote.
Single owner. Name a successor in your operating agreement, or the business goes through probate and your family negotiates with a court. If you are married in a community property state like California or Texas, your spouse may already hold an interest in what you think is entirely yours, so get a written spousal consent while everyone is happy.
Family and close partners. Write it down precisely because it is family, since there is no future financing to force a cleanup. Separate the two hats: a relative who works in the business gets a salary for the work and distributions for the ownership, and conflating them poisons both conversations.
Even and uneven splits. Fifty-fifty is fine on economics and dangerous on control, because two equal votes with no tiebreak freeze the business the first time you truly disagree. Give one owner the tiebreaking vote, or name a third party to break deadlocks. Where the split is uneven, the minority owner needs consent rights on the decisions that matter (a sale, significant debt, new equity, a change in the business) and nothing beyond that, since a minority veto over daily operations is just deadlock in another form.
Everyone needs a buy-sell provision, covering the four exits that actually happen: death, divorce, disability, and wanting out. Name who may buy, how the price is set, and the payment period. Owner-operated businesses often fund the death scenario with life insurance on each owner, so cash exists to buy the shares rather than leaving you in business with your late partner's spouse.
On splitting when nobody knows what they're worth yet: you do not need to know. Vesting resolves it. Four years with a one-year cliff means equity is earned over time rather than awarded on a prediction, and a cofounder who leaves at eighteen months keeps roughly 37.5% of their grant. Split on expected forward contribution rather than on who had the idea, and default to roughly equal unless someone is part-time or funding the business.
The 83(b) election
The one item here with no remedy. When founder stock vests over time, IRC §83(a) taxes you as ordinary income at each vesting date on the value then, not at grant. That means annual tax on paper gains you cannot sell to pay the bill.
An 83(b) election flips it: you elect to be taxed at grant, when the spread is zero and the tax is nothing, and it starts your capital gains and QSBS clocks. File within 30 calendar days of the transfer date. Not from signing. There is no extension and no cure. The IRS standardized it on Form 15620 in 2025 and now accepts it electronically.
If you're self-funding or regulated
A personal guarantee defeats your entity. On an SBA 7(a) loan, every owner of 20% or more must sign an unlimited personal guarantee. No lender can waive it, many ask smaller owners to sign too, and on larger loans they will lien your home. Commercial leases work the same way, so negotiate a cap of a few months' rent and a good-guy clause that releases you on proper notice.
If you need a license, structure comes first. A med spa, clinic, staffing agency, or cannabis business cannot test demand and paper it later, because operating is the activity that requires the paperwork. A non-physician cannot own a medical practice in a corporate practice of medicine state, and traction does not cure it.
What can wait
Intellectual property, mostly. Nobody is going to steal your idea, and your job is finding out whether anyone will pay. Patents are five figures over several years. Trademarks cost $350 per class and can wait until you go to market. The one item that cannot wait is the assignment, which costs nothing.
Employees. Use contractors and advisors while you can, since employees bring payroll, withholding, workers' comp, and strict state law. But you have to mean it: the IRS control test treats someone working full time under your direction as an employee whatever the paper says. Physical businesses are the exception and will have real employees from day one.
Customer agreements. Do not write one before you have customers. Terms of service are different and you need them at launch.
Each of these gets fuller treatment elsewhere in the Playbook.
