CPA Positions Now Take 73 Days to Fill—41% Longer Than Non-CPA Roles. Is the Market Telling Us the CPA Model Is Broken?

General
1

I’ve been running Thompson & Associates for 15 years, and I have never seen a hiring environment like this. I posted a senior accountant position requiring CPA certification three months ago. 73 days later, I finally made an offer—and the candidate had two competing offers already.

This isn’t just my experience. The data is stark:

The Numbers

  • Finance roles requiring CPA credentials now take an average of 73 days to fill—41% longer than comparable non-CPA positions
  • Each additional week of vacancy costs roughly $3,000–$5,000 in lost productivity and delayed reporting
  • 87% of finance and accounting leaders report talent shortages
  • CPA exam participation has declined over 30% since 2016
  • 75% of working CPAs are approaching retirement age
  • The BLS projects ~124,200 annual accounting openings, but new graduates are nowhere near that number

Something structural is broken when the profession’s primary credential is actively driving people away from the profession.

The 150-Hour Barrier Is Finally Cracking

The good news is that states are finally listening. As of January 2026, 25 states have either removed or provided alternatives to the 150-hour college credit requirement. Ohio passed HB 238 allowing 120 hours + 2 years experience. Texas follows suit in August 2026 with SB 262. The AICPA/NASBA Competency-Based Experience Pathway lets candidates substitute the extra 30 hours with verified professional experience.

This is overdue. The 150-hour requirement was designed for a different era. Today it adds $30K–$50K in tuition costs and delays career entry by 1–2 years—disproportionately hitting candidates from lower-income backgrounds who the profession desperately needs.

But Here’s My Real Question for This Community

When you can’t hire CPAs, what do you actually do? In my practice, I’ve been forced into some combination of:

(A) Raise prices — fewer staff means longer turnaround, so I charge premium for the constraint. But this has a ceiling; clients will leave.

(B) Turn away work — I’ve declined 6 potential clients this year because I literally don’t have capacity. That’s roughly $90K in annual revenue walking out the door.

(C) Hire non-CPAs and train them — This is where it gets interesting for the Beancount community. I hired a bookkeeper last year with strong Python skills but no CPA. She learned Beancount in two weeks, built importers that saved 15 hours/month, and now handles work that used to require a CPA-level reviewer—because the automation catches errors that human reviewers miss.

(D) Automate with AI and serve more with fewer humans — We’re using AI categorization + Beancount scripts to reduce per-client hours by ~40%. This lets my 3-person team handle work that would have needed 5 people three years ago.

The Uncomfortable Implication

Option C raises a provocative question: does this shortage create opportunity for tech-savvy bookkeepers who can deliver CPA-quality work without the credential? Companies that opened positions to “CPA-eligible” candidates cut their time-to-fill by 22% while maintaining comparable performance outcomes.

If a bookkeeper with Python skills + Beancount automation + AI validation tools can produce work that’s as accurate and audit-ready as a CPA working in QuickBooks… what exactly is the CPA credential protecting at that point?

I say this AS a CPA. The credential gave me professional judgment, ethical training, and regulatory knowledge that matters. But I’m watching the market answer a different question: is 73 days of vacancy worth the credential requirement, or should the work get done by whoever can do it competently?

What’s Your Experience?

  1. Practice owners: Are you struggling to hire? What’s your actual time-to-fill, and what compromises have you made?
  2. Non-CPA bookkeepers: Do you feel the shortage is creating more opportunity for you? Are clients less credential-focused than they used to be?
  3. Tech-savvy practitioners: Has Beancount/automation capability changed how you think about staffing and credentials?
  4. Anyone tracking the 150-hour reforms: Which states are getting it right, and will it actually move the needle fast enough?

The market is telling us something. I’m not sure we’re listening.

2

Alice, this hits close to home. I’m the non-CPA bookkeeper who’s benefiting from exactly the dynamic you’re describing—and honestly, it’s complicated.

The Opportunity Is Real

Three years ago, I couldn’t get a meeting with certain clients because they wanted “a CPA” on their books. Full stop. Didn’t matter that I had 10 years of bookkeeping experience and was already doing cleaner work than half the CPAs they’d used before.

Fast forward to 2026: I’ve onboarded 8 new clients this year alone. Four of them explicitly told me they couldn’t find a CPA who would take on a small business account. One said her previous CPA firm dropped her because they were “consolidating to focus on larger clients”—translation: they don’t have enough staff for small accounts.

But Let’s Be Honest About the Limits

I’m not doing attestation work. I’m not signing audit opinions. I’m not representing clients before the IRS (that’s @tax_tina’s territory as an EA). The CPA credential protects specific functions that matter.

What I AM doing:

  • Full-cycle bookkeeping with Beancount (reconciliation, categorization, monthly close)
  • Automated bank imports via Python scripts that flag anomalies before I even look at them
  • Monthly financial packages that clients tell me are “better than what their old CPA firm delivered”
  • Connecting clients with CPAs at tax time for the work that actually requires the credential

The Beancount Advantage for Non-CPAs

Here’s what I think the Beancount community underestimates: plain text accounting is a credential multiplier for non-CPAs.

When I show a prospect my workflow—Git version control on their books, automated reconciliation scripts, Fava dashboards they can access anytime—they stop asking about credentials and start asking “when can you start?” The technology signals competence in a way that a framed certificate on the wall used to.

My time-to-close for a new client engagement? 3 days. I hear CPA firms are quoting 2-3 weeks just to do the intake because they’re drowning.

The Real Question

Alice, you mentioned hiring a bookkeeper with Python skills. Was that a conscious strategy shift, or were you just desperate? Because I think there’s a huge difference between “we’re deliberately building hybrid teams” and “we can’t find CPAs so we’ll settle for bookkeepers.” The framing matters—both for the bookkeeper’s career trajectory and for client trust.

I’d love to hear how other non-CPA practitioners are positioning themselves in this market. Are you explicitly marketing the shortage as your opportunity? Or does that feel predatory?

3

This thread is touching on something I’ve been thinking about from a completely different angle—as someone who came from tech, not accounting.

The Credential Isn’t the Problem. The Pipeline Is.

I want to push back gently on the “CPA model is broken” framing. The credential itself serves important public interest functions (audit independence, ethical standards, regulatory competence). What’s broken is the pipeline to the credential—and these are different problems requiring different solutions.

The 150-hour requirement created a bottleneck. States are fixing that. But even if every state drops to 120 hours tomorrow, you still have a fundamental attraction problem: talented quantitative people look at accounting and see:

  • Starting salaries 20-40% below tech equivalents
  • Rigid hierarchy and slow career progression
  • Tools from 2005 (looking at you, legacy ERP systems)
  • A professional culture that sometimes punishes innovation

Meanwhile, the same person could learn Python, pick up data engineering, and start at $120K with stock options and remote flexibility.

Where Beancount Fits In (And This Is Actually Hopeful)

Here’s my optimistic take: Beancount might be the gateway drug that brings tech talent into accounting.

I’m a software engineer who started using Beancount for personal finance. Through this community and tracking my rental properties, I learned more about double-entry accounting, GAAP concepts, and financial reporting than I ever expected. If someone had told 25-year-old me “you’ll be reading FASB standards for fun,” I’d have laughed.

But Beancount made accounting feel like my kind of problem—text files, version control, scripting, automation. It reframed accounting from “boring data entry” to “interesting systems design.”

A Concrete Proposal

What if we positioned the Beancount community as an on-ramp for career changers? Consider:

  1. Tech workers who want to understand finance — they already have the Python/Git skills; teach them accounting fundamentals through Beancount
  2. Career changers entering accounting — the 120-hour + 2 years experience path means they could work while learning; Beancount fluency makes them immediately productive
  3. Firms hiring for potential over credentials — Alice’s example of hiring a bookkeeper with Python skills is exactly this

The 73-day hiring gap exists because firms are fishing in a shrinking pond. Beancount expands the pond by making accounting accessible to people who would never have considered it through the traditional path.

My Concern

The one thing that worries me: if we say “tech skills can replace CPA judgment,” we’re oversimplifying. I can write beautiful Beancount automation, but I still defer to @accountant_alice on questions like “is this lease an operating lease or finance lease under ASC 842?” That judgment comes from deep training and experience that no Python script replaces.

The answer isn’t “CPAs vs non-CPAs.” It’s hybrid teams where each person contributes their strength. The firms that figure this out first will win the talent war.

4

Adding the Enrolled Agent perspective here, because I think the CPA conversation is missing a credential that’s been quietly gaining ground during this shortage.

The EA Credential Is Having a Moment

I’m an IRS Enrolled Agent with 12 years of experience—including time on the other side of the desk as an IRS auditor. Here’s what I’ve observed:

When clients can’t find a CPA, many of them land on EAs as an alternative. And frankly, for tax-specific work, the EA credential is more focused and arguably more relevant:

  • No 150-hour requirement — pass the 3-part Special Enrollment Exam, done
  • Unlimited IRS representation rights — same as a CPA for tax matters
  • Federal credential — works in all 50 states (CPA licenses are state-by-state)
  • Lower barrier to entry — no degree requirement, just pass the exam and background check
  • Annual continuing education — 72 hours per 3-year cycle keeps us current

The EA exam pass rate hovers around 70-75%, which is actually comparable to individual CPA exam sections. But you can go from zero to credentialed in 6-12 months instead of the 5-7 year CPA pipeline (bachelor’s + extra hours + exam + experience requirement).

Where This Intersects with the 73-Day Problem

Alice’s numbers about time-to-fill are specifically about CPA-required positions. But how many of those positions actually require a CPA versus traditionally prefer one? In my experience:

  • Tax preparation and planning — EA handles this fully
  • IRS representation and audit defense — EA handles this fully
  • Bookkeeping and financial reporting — no credential needed if supervised
  • Attestation and audit — CPA required (this is the real moat)

So the question becomes: of those 124,200 annual openings, how many are truly attestation roles? My guess is less than 20%. The rest could be filled by EAs, bookkeepers with strong automation skills, or hybrid professionals.

The Beancount + EA Stack

Here’s what I’ve built in my practice: Beancount handles the year-round bookkeeping and transaction management. At tax time, I export structured data from my Beancount queries directly into tax preparation workflows. The audit trail is immaculate because every transaction is in Git with metadata tags.

When the IRS asks “show us how you calculated this deduction,” I don’t dig through QuickBooks. I run a BQL query and produce a report with transaction-level detail, dates, sources, and documentation links. My audit defense outcomes have been noticeably better since switching to Beancount.

The Reform That Nobody’s Talking About

Everyone focuses on the 150-hour debate. But here’s what I think would actually move the needle faster:

Expand the work that non-CPAs can do under supervision. If a bookkeeper with Beancount skills and Python automation can produce accurate, audit-ready financial statements—and a CPA reviews and signs off—you’ve effectively multiplied your CPA’s capacity by 3-4x without needing more CPAs.

Texas is doing something smart with their August 2026 changes. But I’d go further: create a tiered credentialing system where practitioners can level up incrementally (bookkeeper → EA → CPA) with each level unlocking specific capabilities, rather than the current binary of “CPA or not.”

The 73-day gap isn’t just a hiring problem. It’s a system design problem. And as people who think in systems (hello, plain text accounting community), we should be proposing solutions, not just documenting the pain.