CPA firms have more ways to add capacity than ever before.

They can recruit locally. They can hire remote employees. They can use offshore professionals. They can engage freelancers. Increasingly, they can automate parts of tax and accounting workflows with AI.

Yet having more ways to add capacity has not made the talent problem disappear.

The 2026 AICPA PCPS Top Issues Survey shows firms continuing to navigate significant change while technology, AI and the skills required in an AI-enabled profession remain important priorities. AICPA's broader workforce research points in a similar direction: technical accounting knowledge remains essential, but firms increasingly need professionals who can connect that knowledge to business context, communicate effectively, exercise technological judgment and adapt as the work changes.

That makes the talent conversation bigger than one question:

Where do we find another accountant?

There are other questions worth asking.

Who trains that accountant? How long before that person understands the firm's way of working? Where does the knowledge they accumulate live? What happens to that knowledge if they leave? Who keeps the workflow moving when capacity changes?

The capacity problem is real.

But behind it sits another problem: continuity of capability.

In-House Talent: Integration Comes With Responsibility

For many CPA firms, an in-house team remains the natural model.

The firm selects its people, develops them, manages them directly and builds a culture around them. Staff learn not only tax and accounting, but the firm's clients, partner preferences, documentation standards and review expectations.

That accumulated context has considerable value.

The challenge is the investment required to create and retain it.

Recruiting, compensation, training, supervision, career development and capacity management all remain with the firm. As qualified accounting talent has become harder to secure, firms have increasingly looked beyond traditional hiring models.

Offshoring is one of those responses.

It expands the available talent pool and can provide meaningful cost efficiencies. But changing where someone works does not, by itself, eliminate the management problem.

Offshore Staffing Expands the Talent Pool. But Capacity and Continuity Are Different Things.

A common offshore model is straightforward.

A CPA firm works with a provider. The provider allocates a dedicated professional. That person becomes an extension of the CPA firm's team.

For many firms, that model works well.

The CPA communicates directly with a known person, assigns work, provides feedback and progressively teaches that professional how the firm operates.

But consider what happens during that process.

At first, the offshore professional may be learning the firm's software, workpapers and workflow.

Then they begin learning reviewer preferences.

Then recurring clients.

Then industry-specific issues.

Then the small details that rarely appear in an SOP but make someone substantially more productive inside a particular firm.

Over time, the CPA is no longer working with the same capability they received on day one.

The professional has accumulated firm-specific knowledge.

That creates an important question:

Where does that knowledge actually live?

If most of it lives with one individual, turnover means more than replacing available hours.

Part of the learning curve may leave with them.

The replacement may be capable, and the position may be filled quickly. But someone still has to understand the firm's systems, preferences, clients and recurring issues again.

This is not unique to offshore teams. The same thing happens when an experienced in-house employee leaves.

The difference is that external staffing is sometimes purchased as though adding headcount has also outsourced the responsibility for continuity.

It has not necessarily done so.

The lesson isn't that dedicated offshore staffing fails.

It is that staff allocation and knowledge continuity solve different problems.

The Hidden Cost Is the Learning Curve

When an experienced professional leaves, their salary or capacity is not the only thing that disappears.

Over the relationship, that person may have learned how work enters the firm's process, what constitutes a return ready for review, which issues should be escalated, how particular reviewers want workpapers documented, which clients have recurring complications, and where risks tend to appear within the firm's niches.

A new professional can learn those things.

But the firm pays for that learning in time, questions, supervision and review.

If the knowledge is never captured outside the individual, the learning curve can partially reset whenever the individual changes.

That is why we think a more useful question than:

How many people have been added?

is:

Is the delivery capability becoming stronger as the relationship continues?

 

From Staff Allocation to Workflow Ownership

This distinction has shaped how we think about the CredTax Pod model.

A Pod is not intended to function simply as a group of people allocated to a CPA firm.

The unit of responsibility is the workflow.

In a traditional dedicated-staff relationship, the structure may largely look like:

CPA Firm → Dedicated Professional → Work

The professional becomes closely integrated with the firm, while the CPA firm directly manages priorities and day-to-day assignments.

There is nothing inherently wrong with that model. It is also why CredTax offers dedicated staffing for firms that specifically want an identifiable professional working as an extension of their internal team.

A Pod is designed differently:

CPA Firm → Defined Workstream → CredTax Pod → Internal Assignment, Preparation, Escalation and Quality Control → CPA Firm

The difference is operational.

The CPA firm still defines its requirements, controls the client relationship, maintains appropriate oversight and knows who is authorized to access its information.

But deciding how CredTax organizes its people to keep the agreed workflow moving becomes CredTax's management responsibility, rather than another staffing problem for the CPA to solve.

If an individual becomes unavailable or eventually leaves, the person still matters and some transition may still be required.

But the workflow is not supposed to belong to that individual.

Knowledge Has to Move From the Individual Into the System

For workflow ownership to mean anything, simply putting several people into a Pod is not enough.

Knowledge has to be captured.

A reviewer correction should not only fix one return. It should teach the system something.

A technical question should not only produce an answer. Where appropriate, the underlying research should become reusable knowledge.

A recurring client issue should not surprise the team every year.

And a CPA firm's preferences should not have to be rediscovered by every person who touches the engagement.

The progression should increasingly look like:

Work performed → Issue encountered → Question or correction → Knowledge captured → Procedure or training improved → Learning applied to future work

Over time, the objective is for capability to become increasingly institutional rather than individual.

That does not eliminate turnover.

People pursue new opportunities, career advancement, different compensation and different roles. A resilient delivery model should acknowledge that rather than pretend otherwise.

The objective is not to create a system in which nobody ever leaves.

It is to reduce how much of the knowledge developed through the relationship leaves with them.

This Is Why Niche Knowledge Matters Too

The same principle applies when a CPA firm has concentrations in industries such as real estate, healthcare, construction or fitness.

At the beginning of a relationship, an external team may understand tax preparation but know relatively little about the recurring patterns within that firm's niche.

That should not remain true indefinitely.

As the relationship develops, recurring technical questions, reviewer feedback, client patterns and research should progressively improve what the team knows to look for. What begins as experience on individual engagements should gradually become part of the knowledge available to the wider delivery team.

The objective isn't merely to have someone who has prepared a particular type of return before.

It is to build a delivery system that becomes better prepared for the next one.

In other words, experience should accumulate, not continually restart.

Where AI Fits

AI adds another source of capacity.

It can assist with document processing, research, data extraction, reconciliation, anomaly identification and other parts of tax and accounting workflows. Used appropriately, that can make trained professionals substantially more productive.

But AI does not change the underlying question of this article:

Who owns the capability and the responsibility for the workflow?

IRS guidance on responsible AI use in federal tax practice recognizes the potential benefits of these tools while emphasizing risks such as inaccurate output, fabricated information or citations, confidentiality concerns and the need for practitioners to review and validate AI-generated work.

For us, that makes AI an enabling layer rather than the operating model itself.

AI and automation can support the work. Trained professionals apply judgment. Documented systems preserve what is learned. The CPA firm maintains the professional oversight required of it.

Technology can increase capacity and improve how work gets done.

But the system around that technology still determines whether knowledge accumulates, issues are escalated appropriately and responsibility remains clear.

The Bigger Question: Who Owns the Capability?

CPA firms will continue to use different combinations of talent.

Some will maintain predominantly in-house teams. Some will integrate dedicated offshore professionals directly into their organizations. Some will use managed outsourcing. Most will increasingly use automation and AI somewhere in the process.

The appropriate combination depends on the firm.

But across all of those models, we think one question matters:

Who is responsible for ensuring that the capability gets better over time?

Adding another person creates capacity.

Technology creates leverage.

Neither automatically creates institutional knowledge.

That requires a system that captures what is learned, develops the people involved, incorporates technical changes, improves procedures and maintains continuity when individuals change.

That is the problem the CredTax Pod model is designed around.

Not replacing the CPA.

Not replacing the CPA firm's judgment.

And not pretending people are interchangeable.

The objective is to take responsibility for a defined part of the workflow, build knowledge around it, develop the team supporting it, and make delivery progressively less dependent on any one individual as the relationship matures.

A CPA firm should not have to start the same learning curve again every time the person performing the work changes.

The relationship should compound what has already been learned.