CPA Services for Startups and SaaS

The entity choice is the first decision a founder makes that a startup CPA will weigh in on hard, because it shapes everything that follows: how you are taxed, whether you can grant options, whether investors will write a check, and how much tax you pay when you finally sell. For a company that intends to raise venture capital, the answer is almost always a Delaware C corporation, and the reasons are practical rather than sentimental. Institutional investors are set up to buy preferred stock in a Delaware C-corp. Their fund documents, their board seats, and their liquidation preferences all assume that structure, and asking a venture firm to invest in an LLC is a good way to end a conversation early. Delaware itself is chosen not for tax reasons, since it taxes little of a startup’s early activity, but because its corporate law is the most developed and predictable in the country, which investors value.

An LLC is not wrong for every founder. It is cheaper to form, simpler to maintain, and by default it is a pass-through, so early losses flow to your personal return and can offset other income. For a bootstrapped SaaS company that may never raise outside money and expects to distribute profit to a small number of owners, an LLC can be the better home. The problem is timing. If you start as an LLC and later need to convert to a C-corp to take investment, the conversion can be a taxable event and, worse, it resets the clock on qualified small business stock, the very break that can eliminate tax on an exit. You lose years of holding period at the moment the company finally has value worth protecting.

The tax mechanics are straightforward once the structure is set. A C corporation pays a flat 21% federal rate on taxable income. Most early startups do not owe much because they are spending more than they earn, but the rate matters once the company turns profitable. The offsetting concern people raise is double taxation, the idea that C-corp profits are taxed once at the company and again when distributed as dividends. In practice this rarely bites a growth-stage startup, because such companies reinvest everything and pay no dividends, so there is no second layer to worry about until an exit, and at exit the qualified small business stock rules can do the heavy lifting instead.

Consider a concrete case. Two founders start a SaaS company and expect to raise a seed round within a year. If they form a Delaware C-corp now and issue founder stock immediately, their Section 1202 holding period starts today, and their cap table is ready for investors. If instead they form an LLC to save a few hundred dollars in setup and convert eighteen months later when the seed investor demands a C-corp, they may recognize tax on the conversion and they restart the 1202 clock from zero, potentially costing each founder a seven-figure exclusion on a later sale. The few hundred dollars saved up front is dwarfed by what the delay can cost. We run this decision against your real fundraising plan and handle the formation, founder stock, and 83(b) elections through our entity work at entity formation and structuring, and the IRS entity classification guidance together with the Form 1120 instructions lay out how each structure is taxed.

Qualified small business stock, usually shortened to QSBS, is one of the most valuable breaks in the tax code for a founder, and a startup CPA earns their fee many times over by making sure a company qualifies for it from the beginning. Section 1202 lets an eligible shareholder exclude a large portion, in many cases all, of the gain on a sale of qualifying stock from federal income tax, up to a generous per-issuer cap. For a founder who builds a company from nothing and sells it for millions, the difference between qualifying and not can be the single largest tax outcome of their life, easily larger than every other planning decision combined.

The catch is that the rules are technical and unforgiving, and most of them have to be satisfied at the moment the stock is issued, not at the moment you sell. The company must be a domestic C corporation. Its gross assets must sit below a statutory ceiling when the stock is issued and immediately after. The stock must be acquired at original issuance, meaning you got it directly from the company rather than buying it from someone else. The company has to run an active qualified business, which excludes certain service and investment fields but comfortably includes most software and SaaS companies. And the shareholder generally has to hold the stock for the required multi-year period measured from issuance to get the full exclusion. Because so much of this is fixed at issuance, the value of a startup CPA is front-loaded: we confirm the company qualifies, we document the asset level at the date of issuance, and we start tracking the holding period the day founder stock is granted, so that years later there is a clean record proving the stock qualifies.

Where founders get hurt is by treating QSBS as something to think about near an exit. By then the facts are frozen. If the company was an LLC for its first two years, that time does not count. If it raised so much that its assets blew past the ceiling before founder stock was properly issued, the window may have closed. If nobody documented the asset value at issuance, proving qualification during acquisition diligence becomes a painful scramble. None of these problems are fixable in hindsight, which is exactly why the planning belongs at formation.

Here is the math that makes people pay attention. Suppose a founder holds QSBS with a near-zero basis and sells their stake for $8 million after satisfying every requirement including the holding period. Under Section 1202, that entire gain can be excluded from federal income tax, saving roughly $1.6 million in federal capital gains tax at a 20% rate, before even counting the additional net investment income tax that is also avoided. Now suppose that same founder had formed an LLC first and converted late, restarting the clock and selling six months short of the required holding period. The exclusion is lost, and the full gain is taxable. The cost of that one structural misstep is over a million and a half dollars. We build the QSBS analysis into the entity setup, document qualification at issuance, and monitor the holding period as part of ongoing planning through tax strategy consulting. The statute itself lives at 26 U.S.C. Section 1202, and the broader small business tax rules sit in the IRS starting a business center. Getting it right early is nearly free, and getting it wrong is one of the most expensive mistakes a founder can make.

The research credit under Section 41 is the tax provision most often left unclaimed by early companies, and a startup CPA who knows how to use it can hand a pre-revenue business real cash rather than a future tax break. The credit is a dollar-for-dollar reduction in tax for qualified research spending, and the qualifying costs map almost perfectly onto what a software company already does. Wages paid to engineers writing and testing new code, a portion of contractor costs for development work, and supplies consumed in the process can all count, provided the work meets a four-part test that centers on developing or improving a product through a process of technical experimentation. Building new SaaS functionality, resolving genuine technical uncertainty about whether or how something can be done, generally qualifies.

The obvious objection is that a startup losing money owes no income tax, so a credit against income tax seems worthless. This is where the provision built for startups changes the picture. A qualified small business, broadly one under a gross-receipts ceiling and within its first years of having receipts, can elect to apply a capped amount of its research credit against the employer portion of payroll taxes instead of income tax. That converts the credit from a paper asset into a reduction of a bill the company actually pays every pay period. For a startup burning cash, offsetting payroll taxes is close to receiving money, and it arrives quarter by quarter rather than someday when the company turns profitable.

The work required to claim it is real, which is why it belongs with a professional. You have to identify which employees and projects qualify, allocate wages to qualified activities with documentation that would survive an examination, calculate the credit under the chosen method, and file the payroll-tax election correctly and on time. Sloppy claims draw scrutiny, and the substantiation matters as much as the arithmetic. We handle the study, the allocation, and the filings, and we tie the payroll offset to the work our payroll compliance team already does so the credit actually lands against the right liability.

A worked example shows the scale. Suppose a seed-stage SaaS company spends $500,000 on engineering wages for employees doing qualifying development. Depending on the method and the mix of costs, the research credit might come to roughly $50,000 for the year. A profitable company would use that to cut its income tax, but our pre-revenue startup owes none, so instead it elects to apply the credit against employer payroll taxes up to the allowed cap. Over the following quarters, that startup simply pays tens of thousands of dollars less in payroll tax than it otherwise would, cash that extends the runway by weeks at exactly the moment runway is most precious. Layered on top, current law restored the immediate deduction of domestic research costs under Section 174, undoing the earlier requirement to spread engineering salaries over five years, which had bizarrely created taxable income at companies that were losing money. We calculate the credit, make the election, and coordinate the Section 174 treatment through tax strategy consulting, and the IRS research credit guidance defines what qualifies. Leaving this on the table is leaving startup fuel unburned.

Revenue recognition is where SaaS accounting diverges sharply from the way founders instinctively think about money, and a startup CPA who understands ASC 606 keeps a company out of trouble that a generic bookkeeper would create. The instinct is simple: cash arrived, so we earned revenue. Under the ASC 606 standard that governs how subscription businesses report, that instinct is wrong. You recognize revenue as you satisfy your obligation to the customer, which for a SaaS company means as you provide the software over the subscription term, not when the payment clears. Cash timing and revenue timing are two different things, and conflating them produces financial statements that mislead you, your board, and eventually an acquirer.

The mechanism that reconciles the two is deferred revenue, a liability on the balance sheet. When a customer pays upfront for an annual plan, you have the cash but you have not yet delivered eleven of the twelve months of service, so most of that payment sits as deferred revenue and is recognized into income month by month as you earn it. This matters for real decisions. A startup that books an entire annual prepayment as revenue on day one looks far more profitable and faster-growing than it is, and when a diligence team recalculates revenue correctly during a fundraise or acquisition, the correction can be embarrassing at best and deal-threatening at worst. Investors and acquirers expect ASC 606-compliant numbers, and metrics they care about, such as recognized revenue and the reconciliation to bookings, depend on the deferred-revenue schedule being maintained correctly from the start.

Keeping it clean is ongoing work rather than a year-end fix. Every new subscription, every renewal, every upgrade or downgrade, and every cancellation changes the deferred-revenue balance and the amount recognized each month. Multi-year deals, discounts, and usage-based components add wrinkles that have to be handled consistently. We maintain the deferred-revenue schedule as part of the monthly close so that recognized revenue is always defensible and the balance sheet always reflects what the company genuinely owes in future service. That discipline pays off the moment someone looks closely, because the books already tell a clean and consistent story rather than requiring a frantic restatement.

Here is the concrete picture. A SaaS company sells a $12,000 annual subscription and collects the full amount in January. Under ASC 606, it recognizes $1,000 of revenue in January and records $11,000 as deferred revenue. Each following month it recognizes another $1,000 and the deferred balance falls by $1,000, reaching zero at year end when the service has been fully delivered. If that same company instead booked all $12,000 as January revenue, its first-quarter revenue would be overstated by $11,000, its growth rate would look inflated, and a careful investor would catch the error and question every other number in the deck. We tie this schedule to the reporting our team produces through monthly financial reporting so the revenue you show is the revenue you can defend. The framework sits within the recordkeeping expectations of the IRS starting a business center, and clean recognition is what lets a growth company raise on its numbers instead of apologizing for them.

Equity is how startups pay people and raise early money, and it is also where founders make the most expensive avoidable tax mistakes, which is why a startup CPA gets involved the moment stock or options change hands. Three instruments dominate the early cap table: SAFEs and convertible notes on the fundraising side, and stock options and restricted stock on the compensation side. Each carries tax and accounting treatment that is easy to get wrong and painful to unwind.

Start with the fundraising instruments. A SAFE, a simple agreement for future equity, and a convertible note are ways to take money now and convert it to stock later at a priced round. Receiving cash through a SAFE is generally not taxable income to the company, because it is a financing event rather than revenue, but it changes the balance sheet and it dilutes the cap table when it converts. A convertible note may carry interest and has debt characteristics that a SAFE does not. The conversion terms, discounts, and valuation caps all have consequences that surface at the next round, and if they are recorded loosely, a priced financing opens with an accounting cleanup that slows the deal. We record these instruments correctly when they are issued so the cap table and the books agree when it matters.

The compensation side is where the individual tax stakes are highest. Options come in two flavors. Incentive stock options can receive favorable capital-gains treatment if a set of holding requirements is met, though they can trigger alternative minimum tax on exercise. Nonqualified options are taxed as ordinary income on the spread between the exercise price and the value at exercise. Which type an employee holds, and when they exercise, drives very different tax outcomes, and founders should understand the split before they hand out a grant. The single most important item, though, is the 83(b) election. When a founder or early employee receives restricted stock that vests over time, the default rule taxes the value as it vests, which for a company that is appreciating means a growing tax bill on paper gains with no cash to pay it. Filing an 83(b) election within thirty days of the grant flips this: you elect to be taxed on the value now, when the stock is worth almost nothing, and all future appreciation is taxed later as capital gain on sale. Miss the thirty-day deadline and there is no fix, the election is simply gone.

The numbers show why the deadline is sacred. Suppose a founder receives 1,000,000 shares of restricted stock at formation, worth a fraction of a cent each, so the total value is essentially nil. File an 83(b) within thirty days and the founder recognizes almost no income now and starts the capital-gains and QSBS clocks immediately. Skip it, and suppose the stock is worth $2 per share when it vests two years later. The founder would then recognize $2,000,000 of ordinary income at vesting, with a tax bill in the hundreds of thousands of dollars, on stock they cannot yet sell to pay it. The election that would have prevented this takes one page and a stamp, but only inside a thirty-day window that never reopens. We flag the 83(b) at every grant, prepare the filing, and coordinate the withholding on option exercises with our payroll compliance team, since equity events flow through payroll. The IRS stock options guidance covers the ISO and NSO treatment, and getting a founder to file that one small form on time is among the highest-value things a startup CPA does all year.