Is Review Capacity, Not Prep Hours, Capping Your Tax Practice?

Consider a partner named Helen, who runs a twenty-five-person firm and has just closed a good recruiting season. She has taken on two additional preparers and a handful of seasonal staff. By the second week of March, returns are being prepared faster than at any point in her career. Yet the number of returns leaving the firm each week has barely moved, and two of her managers are working until eleven at night.

Helen has not misread her firm. She has found the point where it actually narrows. Preparation capacity went up. Review capacity did not, and the work simply waits in front of the people who sign.

This article looks at what recent sources say about that pattern, how to find out whether it describes your firm, and what a sensible response looks like. The names in the opening are an illustration, not a real client.

Here is what matters most:

• A September 2026 CPA Practice Advisor article states that "production volume can rise quickly, but experienced reviewers do not multiply with it."
• An Inside Public Accounting piece, updated February 2026, argues that "review capacity has not scaled at the same pace as preparation capacity." Its author runs an accounting outsourcing firm, so read it as informed opinion, not measured research.
• The AICPA's 2026 Top Issues Survey of 629 firms put hiring experienced staff first for firms with 11 to 30 employees, and staff workload management in the top four for the four larger size groups.
• Thomson Reuters suggests measuring days from "ready for review" to reviewer pickup, and returns per reviewer, to tell whether the constraint is review or preparation.
• The firms that do this well add capacity in ways that stay reviewable and keep professional judgment with partners and managers.

What the sources actually say

It is worth being precise about the evidence, because this topic attracts confident claims and thin data.

Sergio Emacora, writing in CPA Practice Advisor on September 9, 2026, describes busy season as a capacity gap. His central observation is the one quoted above: production volume can rise quickly, but experienced reviewers do not multiply with it. He warns that "done poorly, however, extra capacity can simply move the bottleneck downstream." His recommended approach is to design the arrangement around review before any work begins, and to measure reviewer time, rework, turnaround and escalations, not hours worked. He notes that "hours worked and tasks completed do not show whether the model is helping."

Maanoj Shah, writing for Inside Public Accounting in a piece updated on February 4, 2026, goes further. He states that "in modern tax practices, review-not preparation-has become the primary constraint." He also argues that temporary staff can increase review effort, because seasonal preparers often lack familiarity with a firm's documentation standards and reviewer expectations, which "leads to additional clarification cycles, rework, and deeper review." He is a co-founder of an outsourcing firm and the piece promotes its model. That does not make the argument wrong. It does mean the claim is an opinion, and the article does not offer measured data on reviewer hours.

The AICPA's 2026 Top Issues Survey, fielded April 20 to May 22 with 629 respondents, adds a more neutral view. For firms with 11 to 30 employees, hiring experienced staff ranked first. Firms with 31 to 100 employees ranked next-generation leadership development first. For the four larger size groups, staff workload management was in the top four issues.

Put these together and the claim I would defend is a modest one. Several credible sources describe review as a constraint that does not scale the way preparation does, and the AICPA survey shows firms struggling with experienced staff and workload. What nobody has published, as far as I can find, is a measured study of how many reviewer hours a typical firm spends per return. If a source gives you a precise figure for that, ask where it came from.

The economics, without invented numbers

A partner's time is the most expensive time in the firm, and review is where it is spent. That is why the arithmetic matters, even when the inputs have to come from your own records.

Consider three quantities.

Reviewer hours per return. What does a typical return cost in senior time, by return type? Most firms have never measured it, and the answer differs sharply between a simple individual return and a multi-state partnership.

Waiting time. Returns that are finished but not yet reviewed are work in progress that cannot be billed or filed. A file that sits for days is not a slow review. It is a queue.

Rework. Every return sent back is review capacity spent twice.

Thomson Reuters Tax & Accounting made this point in an October 20, 2026 article on what the October 15 deadline revealed. It recommends tracking the days from "ready for review" to reviewer pickup and the number of returns per reviewer, along with how many times a file was sent back. Its guidance is that a file waiting eight days before a reviewer opens it signals a queue problem, not a slow-review problem.

That distinction is worth holding on to. If files wait a long time before anyone opens them, the fix is capacity or sequencing. If they are opened quickly and take a long time, the fix is quality of preparation or complexity of the work. The two call for opposite responses, and a firm that has never measured the wait cannot tell which one it has.

A one-week test for your own firm

You do not need software to find out. For one busy week, have each reviewer record, for every return they receive, three things: the day it was marked ready, the day they opened it, and whether it came back to the preparer.

At the end of the week you will have a median wait, a count per reviewer and a rework rate. Compare the wait to the time the review itself took. If the wait is longer than the review, you have a queue, and adding preparers will make it longer.

This also shows which reviewers are overloaded and which return types generate the most rework. The second finding is often the more useful one, because it points at preparation habits that can be corrected cheaply.

What helps, and what moves the problem

Part of review time is not judgment at all. Intuit's practice content, in a September 22, 2025 article by Lexi Bonicard, points to review processes that force "information reconstruction": reviewers lose time switching between documents, emails and prior-year files. Its suggestion is to keep preparation notes and prior-year data in the same place the reviewer is working. That is a cheap change to test before spending on anything else.

On adding people, both CPA Practice Advisor and Inside Public Accounting say the same cautionary thing in different words: extra preparation capacity without review capacity relocates the bottleneck. The remedy they point towards is to choose recurring, well-documented tasks for outside or temporary help, define what is delivered and when to escalate, and test candidates on how reviewable their work is, not only on credentials.

On software, the Inside Public Accounting article makes a point worth repeating. Firms that invest only in automation without strengthening human review often find that bottlenecks simply move further downstream.

Where AI belongs in this picture

The most useful role for AI here is helping the reviewer, not replacing them. The software can read the source documents, compare a drafted return against last year's, and produce a short list of items that look different, so that a reviewer starts with the exceptions instead of hunting for them.

That sits comfortably with how both the profession and the regulator describe AI. The IRS Office of Professional Responsibility's Alert 2026-19, issued June 24, 2026, says due diligence requires verifying the accuracy of what AI produces and that practitioners cannot rely solely on AI. CPA Practice Advisor's goal for extra capacity is "to add reviewable production capacity while partners and managers retain ownership of professional judgment." The same standard applies to an AI tool.

There is a data-handling question that comes with this, and it should be settled first. Client documents read by a tool are tax return information, and the rules in Section 7216 apply. We set out what is settled and what is not in our article on tax preparers and ChatGPT.

A firm of about twenty-five people also has a practical constraint: no spare staff to run a project. The sensible approach is a small, bounded change that fits inside the tools you already use, with your reviewers deciding what the system flags. For where this fits in a firm's wider tax workflow, see our guide to tax document automation for CPA firms, and for what to ask any vendor, the real test for an AI vendor pitch.

FAQ

How do I know if review, not preparation, is my bottleneck?

Measure the wait. Record the days between a return being marked ready and a reviewer opening it, for one busy week. If finished returns wait longer than the review itself takes, you have a queue at review, and more preparers will lengthen it.

Will hiring seasonal staff fix a review backlog?

Not by itself. Sources in CPA Practice Advisor and Inside Public Accounting both warn that extra preparation capacity can move the bottleneck to review, and seasonal staff who do not know your standards can increase the rework reviewers face.

Can AI do the review?

No, and the IRS does not expect it to. The Office of Professional Responsibility's June 2026 alert says practitioners cannot rely solely on AI and must verify what it produces. AI can prepare the reviewer by pointing to differences and gaps, but a qualified person still reviews and signs.

What should we measure besides the wait?

Returns per reviewer, how many times a return is sent back, and reviewer time by return type. CPA Practice Advisor's author suggests tracking reviewer time, rework, turnaround and escalations and notes that hours worked do not show whether a change is helping.

Want to see where your own queue sits?

If you would like help running the one-week measurement, or seeing how a reviewer-assist tool could sit inside your current software, book a free strategy call. We start with your numbers, not a demo.

Read next: AI Automation for CPA & Accounting Firms