More than 80% of accounting firm leaders expect succession challenges within five years. Fewer than half have a formal plan to handle them.
That gap is not an administrative oversight. It is a structural risk that shows up as client attrition, onboarding delays, and revenue tied to relationships that exist only in one person’s head. When your senior CPA retires after 20 years, the institutional knowledge they carry does not transfer automatically. It has to be captured, organized, and made accessible before they walk out.
This article covers what the research and accounting community actually say about this problem, what the data shows, and what a realistic four-part response looks like.
The Accounting and Tax Problem
Accounting firms sell expertise and continuity. A client who has worked with the same CPA for two decades expects that person to know their business history, understand why certain elections were made, and remember the context behind prior-year decisions. That knowledge is rarely written down.
When the senior partner leaves, the new staff member inherits a client file with tax returns and workpapers but no narrative. Why did the client take that position in 2019? What conversation happened before that accounting method was selected? Which issues did the firm flag internally but decide not to document formally? None of that lives in the engagement software.
This is the knowledge problem that nobody talks about publicly, because firms are embarrassed to admit how much depends on one person. But the data is clear on how widespread it is.
A 2020 succession survey found that fewer than half of multiowner CPA firms had a succession plan in place, even as more than 80% of those same firms expected succession challenges within five years. That means most firms are heading into a known crisis without a documented response.
Deloitte’s research on organizational knowledge management found that knowledge management ranked among the top three issues influencing organizational success, but only 9% of organizations felt ready to address it. Nine percent. In a profession where the product is expertise, that is a catastrophic readiness gap.
What Industry Professionals Are Actually Saying
The accounting community has been working through this problem in practice-management forums, journal articles, and professional association guidance. A few clear patterns emerge.
First, the firms doing this well start three to five years before the retirement date. Mark Gallegos of CliftonLarsonAllen recommends beginning exit planning that far in advance, with a documented plan that covers client handoffs, responsibility transfers, knowledge-sharing protocols, milestones, and post-retirement boundaries. Earlier guidance from the Journal of Accountancy similarly calls for roughly two years’ notice to give clients time to build comfort with the successor and give the firm time to reassign the relationship if the match fails.
Second, the transfer needs to be gradual and visible to clients. Firms that move an entire book of business at once at retirement see more attrition than firms that break handoffs into stages over two or three years. Julie Welch’s guidance in this area specifically recommends confirming that each client is comfortable with the successor and placing important relationship context into the permanent file.
Third, documentation has to move from personal folders into shared systems. AGN’s succession roadmap advises firms to document standard operating procedures, map the outgoing leader’s client and internal influence, assess stakeholder impact, and define how the transition will be communicated. This shifts knowledge from an individual partner’s memory into repeatable firm processes.
Fourth, equity is increasingly used as a retention mechanism. CPA Australia describes giving key staff initial equity stakes of 5% to 10%, with incremental increases, as a way to lock in future leaders and align them financially with the firm’s continuity.
The hardest part is execution. Firms know what the right approach looks like. Most still have not done it.
By the Numbers: Industry Benchmarks
The quantitative case for treating knowledge transfer as a capital investment, not an HR task.
| Metric | Benchmark | Source |
|---|---|---|
| Firms with formal succession plans | Fewer than 41% | 2020 Succession Survey |
| Firms expecting succession challenges within 5 years | More than 80% | 2020 Succession Survey |
| Organizations ready to address knowledge management | 9% | Deloitte Global Human Capital Trends |
| Replacement cost for senior roles | 50% to 200% of annual salary | Secondary reporting |
| Typical onboarding time for complex client situations | 6 to 12 months | Practitioner estimates |
| Equity stake range used to retain successor talent | 5% to 10% initial | CPA Australia guidance |
| Estimated lost economic output from retiring knowledge workers | $6.9T to $9.6T | Deloitte |
Deloitte estimates that more than 30 million Americans will reach age 65 over a four-year window, potentially creating $6.9 trillion to $9.6 trillion in lost economic output if institutional knowledge is not transferred effectively. That figure covers the broader economy, but accounting firms are among the most exposed categories because they have high concentrations of senior expertise in small teams with long client relationships.
The replacement cost figures in secondary sources consistently land between 50% and 200% of annual salary, with higher estimates for senior and specialized roles. That range is not just about recruiting. It includes the onboarding period, the delayed productivity of the replacement, and the client-relationship disruption that occurs during the gap.
Strategy 1: Solving the Problem of Knowledge Trapped in One Person’s Head
The core issue is that senior CPAs become the single point of contact, the single point of memory, and the single point of trust for their client portfolios. When they leave, all three disappear at once.
The structural fix is to deliberately reduce that concentration before it becomes a departure event.
The strongest model from the research combines four controls, described in the community research as relationship transfer, process transfer, capability transfer, and governance. In practice, this means:
- Introducing the successor to clients two to three years before the retirement date, not two months before
- Having the retiring partner participate in client meetings alongside the successor throughout that period
- Moving critical client context out of emails and personal folders into a centralized file that any authorized staff member can access
- Writing down the decision history: not just what was decided, but why
Deloitte’s structured approach to this uses a three-role model they call Expert, Next’pert, and Practitioner. The expert explains and demonstrates the work. The successor validates and applies it. The broader team reviews and institutionalizes it. That progression works in accounting because it mirrors how client knowledge actually transfers: observation first, then supervised application, then independent execution.
Firms that skip the middle step, where the successor applies knowledge under the expert’s oversight, tend to see the most client-retention failures during handoff. The client notices the gap because the new CPA does not have enough context to ask the right questions.
Strategy 2: Cutting the 6-to-12-Month Onboarding Problem
New staff taking 6 to 12 months to reach full productivity on complex client situations is not a training failure. It is a documentation failure. The information exists. It just lives in the retiring partner’s head and in years of unstructured emails.
The PICPA’s guidance on using AI for succession planning points toward centralized, searchable knowledge repositories as the operational answer. Instead of the new CPA spending weeks reviewing prior-year files and asking senior staff for context, they query a system that already has the relevant history indexed and retrievable.
This is where RunFrame’s approach becomes relevant. RunFrame builds AI knowledge bases from a firm’s existing emails, work papers, and client notes. A new staff member can ask the system why a particular client uses a specific accounting method, and the answer comes from the firm’s own historical files rather than from a senior partner’s memory or a two-hour onboarding meeting. The knowledge gets captured once and becomes accessible indefinitely.
The distinction that matters is between storage and retrieval. Most firms store documents in practice-management software. The problem is that stored documents are not searchable in a useful way. A new CPA cannot search for “why did we recommend the S-corp election for this client” and get a clear answer from a document repository. An AI knowledge base built on the firm’s own data can answer that question because it understands context, not just filenames.
Firms that embed knowledge capture into normal workflows rather than treating it as a separate documentation project see better adoption. Deloitte’s research supports requiring reusable deliverables, client insights, and decision rationale to be captured at engagement close rather than relying on voluntary after-the-fact contributions.
To understand how this fits into a broader operating structure, see how RunFrame deploys AI for professional services firms.
Strategy 3: Documenting Why Decisions Were Made
The most common complaint from managing partners during transitions is not that the files are missing. It is that the reasoning is missing.
The workpapers show what was done. They rarely explain why the client chose that position, what alternatives were considered, what the partner’s read on the client’s risk tolerance was, or what verbal agreements exist alongside the written engagement.
This is a different documentation challenge than procedure manuals or workflow templates. It requires capturing the interpretive layer, the context that makes a file readable to someone who was not in the room.
A few practical approaches from the research:
- Structured client-history interviews with retiring partners, recorded and transcribed before the exit date
- Case debriefs after major engagements that capture the decision logic, not just the deliverable
- Permanent file notes that explain relationship history and key prior decisions in plain language
- Job-shadowing requirements where successors attend client meetings with the retiring partner for at least 18 months before the handoff is formalized
The Journal of Accountancy’s 2016 research on CPA firm succession identified client and internal influence mapping as a critical step that most firms skip. Knowing which clients trust the retiring partner most, and which relationships are most at risk, allows the firm to prioritize its documentation and handoff effort rather than treating all 200 clients identically.
For firms exploring what a systematic AI deployment looks like for their accounting practice, the RunFrame accounting industry page covers the specific use cases.
Implementation Roadmap
A realistic four-phase approach based on the research.
Phase 1: Assess the exposure (months 1 to 2)
Map every senior partner within five years of retirement. For each one, identify the client revenue tied to that relationship, the percentage of those clients who have no secondary relationship with another firm member, and the current state of documentation for each portfolio. This produces a risk register, not a retirement plan.
Phase 2: Start relationship transfers now (months 2 to 18)
For every high-risk client relationship, assign a named successor and begin joint coverage immediately. The retiring partner does not need to announce their retirement timeline. They just need to start bringing the successor into client meetings and communications. This is the most time-sensitive step because client comfort with a new relationship owner takes 12 to 24 months to build.
Phase 3: Capture and centralize knowledge (months 3 to 12)
Begin structured interviews with each retiring partner, focused on their client portfolios and the decision history behind major positions. Move that information into a centralized system with search capability. Require engagement teams to document decision rationale in permanent files going forward, not just deliverables.
This is the phase where technology choices matter most. A practice-management system with document storage is not sufficient for this purpose. The system needs to make that knowledge retrievable by someone who does not already know what they are looking for.
Phase 4: Build governance and contingency (months 6 to 24)
Formalize retirement notice requirements in partner agreements. Create practice-continuation agreements for emergency scenarios. Establish equity pathways for successor development. Review the plan annually rather than treating it as a one-time project.
For firms that want a structured assessment of where they stand before building this roadmap, the AI Readiness Scorecard at RunFrame is a useful starting point. It identifies where the firm’s knowledge infrastructure has gaps before those gaps become departure events.
How RunFrame Approaches This
RunFrame builds AI knowledge bases from the documents and communications your firm already has: emails, work papers, client notes, engagement files, and prior-year returns. The system is designed for accounting firms where the knowledge problem is not a shortage of documents but a shortage of retrievability.
The deployment process starts with ingesting the firm’s existing data and building a searchable knowledge layer on top of it. Staff ask questions in plain language. The system returns answers drawn from the firm’s own historical files, with source references so the answer can be verified.
For managing partners facing near-term retirements, the most immediate use case is onboarding acceleration. Instead of a new CPA spending weeks piecing together a client’s history from prior-year files, they spend 20 minutes querying the knowledge base and arrive at the first client meeting with actual context.
The longer-term use case is institutional memory. As more of the firm’s knowledge gets captured in the system rather than individual partners’ heads, the firm becomes less dependent on any single person’s continued presence. That is the structural shift that makes knowledge transfer sustainable rather than a crisis to manage every time someone retires.
RunFrame’s AI operating system deployment and ongoing AI operations service cover both the initial buildout and the long-term management of the knowledge infrastructure.
For managing partners who want to see where their firm stands today, start with the AI Readiness Scorecard or book a discovery call to talk through the specific retirement timeline and client-concentration risks your firm is carrying.
Frequently Asked Questions
How long does AI deployment take for accounting and tax companies?
Most accounting firms can deploy a basic AI knowledge base within 4 to 8 weeks. The ingestion phase, where the system processes emails, work papers, client notes, and prior-year files, typically takes 2 to 4 weeks. Staff training and workflow integration add another 2 to 4 weeks. Firms with more complex data environments or stricter data-governance requirements may take 10 to 12 weeks for a full rollout.
What does AI cost for an accounting and tax firm?
Costs vary by firm size and scope. A mid-size firm deploying an AI knowledge base typically invests between $1,500 and $6,000 per month depending on the number of users, data volume, and integration requirements. That range is substantially lower than the cost of a single senior CPA departure, which can run 50% to 200% of annual salary when you account for recruiting, onboarding, and client-relationship disruption.
What ROI can accounting and tax companies expect from AI?
The clearest ROI comes from three areas: faster onboarding for new staff, which reduces the 6-to-12-month ramp time; reduced search time for client history and prior-year decisions; and stronger client retention during partner transitions. Firms that document and systematize partner knowledge before retirement also tend to retain more of the revenue tied to those accounts. Specific ROI depends on how many senior partners are within five years of retirement and how much billable revenue is concentrated in those relationships.
Do I need technical staff to use AI in my accounting and tax business?
No. The AI knowledge systems used for accounting firms are designed for non-technical users. Staff ask questions in plain language and receive answers drawn from the firm’s own documents, client notes, and work papers. Setup is handled by the deployment provider. Your team does not need to write code, manage servers, or understand machine learning to use the system effectively.
