Read time: 7 min

Every few months I talk to a founder who just fired their accountant. Not because the books were wrong. Because the accountant had no idea what ARR was, had never heard of ASC 606, and sent a PDF invoice by email with net-30 terms.

If you're building an AI SaaS company, your accounting needs are genuinely different from a restaurant or a law firm — and most accounting firms aren't set up for you. Here's what you should actually be evaluating when you choose one.

They Need to Speak SaaS Metrics Natively

Your accountant should know what MRR, ARR, NRR, CAC, and LTV mean without you explaining them. Not because they're impressive buzzwords, but because these numbers live at the intersection of your financials and your business reality. If your accountant can't reconcile your recognized revenue to your MRR motion, your board package is wrong.

This matters most at the point where your accounting meets your investor narrative. Series A investors will ask why your revenue recognition doesn't match your ARR. If your accountant set up your books on a cash basis and you're presenting ARR to investors, you have a gap that will surface in diligence — and it won't be a small conversation.

ASC 606 Has to Be Set Up Right From Day One

This is where most early-stage SaaS companies carry silent technical debt.

ASC 606 requires you to recognize revenue when you deliver your service — not when you invoice, not when you collect cash, and not when you sign the contract. For a simple monthly SaaS subscription, that's straightforward. The moment you add annual prepays, usage-based components, onboarding fees, or professional services to a contract, the accounting splits into multiple performance obligations that each recognize differently.

The mistake founders make: they grow to 50 customers on a cash-basis or simplified setup, then hit Series A diligence and discover their entire revenue history needs to be reconstructed. That's a six-figure cleanup project at the worst possible moment.

A good accounting firm for AI SaaS sets this up correctly before you have a single paying customer, documents your revenue recognition policy in writing, and makes sure your billing system, CRM, and general ledger are talking to each other. If you're selling usage-based AI APIs, per-seat subscriptions, and annual enterprise contracts simultaneously, you need someone who's done this before — not someone who will Google ASC 606 after you ask.

They Should Know the R&D Credit Landscape Cold

AI companies spend disproportionately on engineering. That spend is almost certainly partially qualifying research under Section 41 — and if no one has built a documentation infrastructure for it, you're leaving money on the table every year.

A good firm will proactively identify your qualifying research expenses, build the documentation process alongside your engineering workflow, and make sure you're actually capturing the credit instead of discovering two years later that you could have been. After the OBBBA changes in 2025, there's also a retroactive relief window for companies under $31M in gross receipts that closes July 6, 2026. If your firm hasn't brought this up, ask them why.

For AI companies specifically: your model training costs, inference infrastructure, and proprietary algorithm development often qualify. The four-part test isn't as narrow as founders assume.

409A Timing Has to Be Proactive

If you're issuing options, you need a current 409A valuation. If you grant options with a stale 409A, you've created a tax problem for your employees — not your company. Most founders find this out after the fact.

An accounting firm that works with startups will either coordinate your 409A directly or have a tight relationship with a firm that does it, and they'll flag when you need a refresh: after a funding round, after a material change in the business, and at least annually. They should be reminding you, not waiting for you to ask.

Fundraise-Ready Financials Are Not Optional

When a VC or strategic buyer asks for your data room, you have roughly 24 hours to make a good first impression. Disorganized financials, cash-basis books, or missing board-approved financial statements are immediate yellow flags that slow diligence and give buyers leverage.

Your accounting firm should be producing monthly GAAP financials, a board-level financial package, and clean deferred revenue schedules — not just a P&L and bank reconciliation. If you've never been asked "what's your deferred revenue balance and how does it reconcile to your contracts?" you haven't been working with a startup-focused firm.

Multi-State and International Exposure Starts Earlier Than You Think

The moment you hire your first remote employee outside California, you may have payroll tax nexus in another state. The moment you have a customer in Europe, GDPR and VAT questions start. The moment you do more than $100K in sales in Texas, you have sales tax exposure.

Most early-stage founders assume this complexity starts later. It starts with your first hire in Austin and your first enterprise customer in Germany.

A firm that works with SaaS and AI companies will have this mapped out ahead of time — or will tell you clearly that you need a specialist — rather than discovering the liability when they're reconciling year-end.

The Things That Actually Separate Good Firms From Generic Ones

When you're evaluating accounting firms, the questions that tell you the most:

Ask them to walk you through how they set up revenue recognition for a SaaS company with annual and monthly tiers plus a usage component. If they can't answer this specifically, they've never done it.

Ask how they handle the federal vs. California Section 174 treatment difference. California didn't adopt the OBBBA rules. If your firm doesn't know there's a difference, your state tax filings are probably wrong.

Ask what they do when you're 60 days from a Series A close and diligence requests hit all at once. The answer tells you whether they've been through it before.

Ask what tools they use. A firm still doing everything in spreadsheets is not a firm built for a company growing 3x a year. You want QBO or Xero as the GL, a proper document management system, and ideally integrations with your billing stack.

Ask who your day-to-day contact will be. You should never be working with someone who doesn't know your business.

What to Look For, Summarized

You need a firm that:

• Understands SaaS metrics and can connect your financials to your business model

• Has set up ASC 606 for subscription and usage-based revenue before, specifically

• Proactively manages your R&D credit documentation throughout the year

• Coordinates 409A timing with your equity grants

• Produces investor-grade monthly financials without being asked

• Has seen multi-state payroll and sales tax nexus issues before they become your problem

What you don't need: a generalist firm that will learn on your dime, a solo CPA who handles your taxes once a year, or anyone who bills by the hour with no predictability.

The right accounting partner pays for itself. The wrong one costs you a funding round.

Anelya Grant is the founder of AG Accounting Inc., an accounting firm serving tech startups and healthcare organizations. She is also co-founder of JustPaid.ai, an AI-powered billing and contract-to-cash platform for growing companies.