Your startup needs an organizational structure. But which one? This piece walks through the decision a founder faces in 2026: which structure, which state, and which service to use. The structure and state answers have barely changed in a decade. The services market is the one piece quietly in flux. And the biggest news hides inside the structure question: the tax law rewrote its math last year, and most founders missed it. I’ll state my views and label them as views.
The structure question is usually over before it starts.
If you plan to raise outside capital, you will almost certainly end up a Delaware C corporation, because that is what the money still assumes. SAFEs, preferred stock financings, option plans, acquirer diligence checklists: the entire paperwork stack of the venture industry is written for one entity in one state. You can fight that default. You will lose slowly and expensively. I have watched lots of startups over the years regret the LLC choice foisted on them by lawyers because it was “more flexible,” which, to my cynical lawyer self, I hear as “more billable.”
What changed in 2025 is how much that default choice pays off. The One Big Beautiful Bill Act rewrote Section 1202, the qualified small business stock rules, for stock issued after July 4, 2025. The per-shareholder exclusion rose from $10 million to $15 million of capital gains, tax-free. The old five-year cliff became a ladder: 50% of the exclusion at three years, 75% at four, 100% at five. The asset ceiling for qualifying companies rose from $50 million to $75 million. Only C corporation stock qualifies. Not LLC interests, not S corporation shares.
Run the arithmetic on a modest outcome. A founder who sells $15 million of qualifying stock after five years keeps $3 million or more that would otherwise have gone to federal capital gains tax (more, once you count the 3.8% net investment income tax that QSBS gain also escapes). That single provision dwarfs every state-fee comparison ever written. The clock starts when the C corporation issues your stock, which is the strongest argument against the “start as an LLC, convert later” path for anyone on the venture track. Converting is routine and usually tax-free. The years you spent as an LLC still don’t count.
My base case: for a venture-path company, the LLC detour saves hundreds of dollars and costs years of QSBS clock. Take C corp for the win.
If you are not on the venture path, the answer inverts. A profitable agency or services business distributing its earnings wants a home-state LLC, possibly with an S election once profit per owner clears roughly $60,000 to $80,000 a year (a CPA call, not mine). The C corporation’s 21% flat rate plus a second tax on dividends is a bad deal for a business that pays out its profits every year.
Delaware, after the exodus headlines
You have probably read that companies are fleeing Delaware. The headlines are real, but the relevance to startups is low. Glass Lewis counted 18 reincorporation proposals from Delaware companies in the 2025 proxy season, and Delaware’s share of large IPO charters slipped from its 80%-plus norm of 2022 to 2024 down to roughly 77% in 2025, with Nevada the main beneficiary. Delaware responded with SB 21 in March 2025, adding safe harbors for interested-party transactions and narrowing books-and-records demands.
Look at who actually left: the completed departures were overwhelmingly controlled companies, and Glass Lewis found 55% of the season’s reincorporation proposals came from companies with a significant or controlling shareholder. Founders with super-voting control and litigation scars have reasons to prefer Nevada’s management-friendly statute. A seed-stage startup asking strangers for money has the opposite incentive: investors price unfamiliarity, and Delaware Chancery’s case law is the devil everyone knows. Private companies, two-thirds of Delaware’s franchise base, are staying put.
The tail wags the dog here. What would change my mind is standard venture documents shipping in Nevada or Texas variants from the major accelerators and law firms. Watch for that; it hasn’t happened.
One practical Delaware note that saves real money: the state’s default franchise tax bill uses the authorized-shares method and routinely shows a five-figure number to a startup with 10 million shares. Recalculate under the assumed-par-value method, and the same company usually owes around $400 to $800. Every year, some founder panics at that letter. I did. Twice. Don’t.
And if you aren’t raising at all, ignore Delaware entirely. Your home state is cheaper and sufficient. In Ohio, where we operate, an LLC costs $99 to form, files no annual report, and owes no commercial activity tax until it has $6 million of gross receipts. Wyoming, at $100 down and about $60 a year with real privacy protections, remains the right answer for holding companies, not operating startups.
The services, graded
The formation-service market sorts cleanly once you know what each one optimizes. Stripe Atlas ($500, Delaware only) optimizes speed: days from idea to formed entity, 83(b) election handled, then it steps back and leaves compliance to you. Clerky (about $819 for a lifetime package) optimizes document hygiene, with templates that startup law firms actually accept, which is why it remains the quiet recommendation for anyone heading into a priced round. Firstbase (about $399 plus subscriptions) optimizes breadth, bundling compliance reminders, mail, and banking. Doola (from $297 plus plans) optimizes for non-US founders, where its ITIN support is sometimes the only realistic path. The budget players are fine for a home-state LLC and wrong for venture paperwork. Three-year all-in costs cluster between $2,000 and $6,500 regardless of sticker price, and the expensive mistake is bad early documents, not the formation fee.
Don’t forget to get an EIN, the corporate version of the SSN. You can get one separately, but I always did it as part of the incorporation process. Not cheaper, but easier. You will also need a registered agent (Ohio calls it a statutory agent), for which you should use a service.
One more way to slice it: incorporation is a one-time decision for most founders, so the services compete hardest on the thing that matters least, the formation fee. What you are actually buying is the quality of the documents that follow you for years and the compliance model that keeps you out of trouble in year two. Pick on those.
Final thoughts
Founders overthink geography and underthink timing. The state question has a boring answer (Delaware if raising, home state if not), and the structure question has a default with a $15 million tax benefit attached. The decision that actually goes wrong is waiting: potential personal liability, IP accumulating in personal names, a co-founder handshake hardening into an accidental partnership, an 83(b) window closing 30 days after stock issues, with no cure available. Incorporation is cheap insurance priced in hundreds of dollars and regret priced in equity. Form the entity when the work gets real, paper it properly, and get back to the product.
Regards,
Charles Stack
Founder, Coworkers.Global
Research by Lexi, a managed paralegal agent at Coworkers.Global.
This article discusses general concepts, not your situation. For that, talk to a lawyer.
Sources. QSBS changes: Perkins Coie. Delaware reincorporation data: Glass Lewis, State of US Reincorporation in 2025; Analysis Group, DExit Trends.