EY Has 150 AI Agents. McKinsey Is Restructuring. Here's How Small Firms Win Anyway.

Published: August 29, 2026 | By: The Crossing Report

---

Summary

KPMG deployed 276,000 employees onto Claude AI. EY announced 150 internal AI agents. McKinsey restructured, cutting roughly 1,400 positions. Most small professional services firm owners read these headlines as a verdict — that the game is over and the big firms have won. That reading is wrong. Big 4 AI deployment compresses the cost of commodity work and frees enterprise firms to compete more aggressively on mid-market clients — but it does nothing to the structural advantages that small, specialized firms have always held: faster deployment, deeper specialization, and client relationships that no enterprise firm can replicate at scale. The firms that will lose to Big 4 AI are the ones competing on the same terms. Here is the three-phase playbook for competing on the terms you can win.

---

The Wrong Conclusion From the Right Headlines

KPMG just deployed 276,000 employees onto Claude AI. EY announced 150 AI agents handling internal work. McKinsey eliminated roughly 1,400 positions and shifted portions of its practice toward outcome-based project pricing. The headlines land every few days now, and most small firm owners read them the same way: as a verdict.

They are not a verdict. They are a signal — but not the one most people draw.

The instinctive response is comparison. How can a 12-person accounting firm compete when KPMG has 276,000 people using AI on every engagement? That framing is the first mistake. You are not competing against KPMG's 276,000 employees. You are competing for clients that KPMG would never take: the $8 million family business, the regional manufacturer that needs one specialist who knows their industry cold, the PE-backed company that cannot afford McKinsey but needs someone who will answer the phone on the second ring. Those clients do not need a global deployment of AI agents. They need someone who knows their business.

The firms that will lose to Big 4 AI are the ones that tried to compete on enterprise terms. The firms that will win are the ones that change the terms entirely.

---

What Big 4 AI Actually Means (and What It Does Not)

When most small firm owners read about Big 4 AI deployment, they process it as a cost story. EY built AI agents — that means EY can now do the same work for less. We cannot compete on cost. That reading is partially right but strategically useless.

Big 4 firms are using AI to reduce the cost of commodity work: document review, initial research synthesis, template generation, compliance checklists, client reporting. These are the categories where junior staff hours were burned and where AI now does the same work faster and cheaper. A KPMG engagement that previously required three associates for a month now requires one associate and a set of AI-assisted workflows.

Here is what that actually means: KPMG freed up capacity to pursue larger, more complex engagements — and became modestly more cost-competitive on mid-market work. That is a real pressure. It is not, however, the existential threat most small firm owners imagine.

The firms at serious risk are the ones doing commodity professional work — basic compliance, generic analysis, template-driven deliverables — without a clear specialization. Those firms are being squeezed from both sides: Big 4 firms are becoming more competitive at the margins, and AI-native competitors (Pilot and Accrual in accounting, Harvey and CoCounsel in law) are building direct-to-client platforms that bypass the traditional firm model entirely.

The firms not at risk are the ones with genuine specialization, with client relationships built over years of specific expertise, and with the operational agility to adopt AI faster than any 340,000-person organization can retrain its people.

The question is not "can we compete with EY's AI budget?" You cannot, and you do not need to. The question is: What can a 10-person firm do in 2026 that EY structurally cannot do? The answer is specific, achievable, and exactly what the Big 4 displacement narrative obscures.

---

Three Unfair Advantages Small Firms Have in 2026

1. You Can Deploy in Days. They Need 18 Months.

EY has 150 AI agents. It also has 395,000 employees spread across 150 countries, partner compensation structures built over decades, client engagement standards requiring months of compliance review before any new tool gets approved, and a global training rollout calendar that cannot accommodate rapid iteration.

When EY deploys a new AI workflow, the implementation timeline looks like this: internal testing, IT security review, risk and compliance approval, regional partner signoff, training design, global rollout. Eighteen months minimum for anything substantive. And that is assuming the first attempt works.

A 10-person accounting firm can test a new AI tool on Monday, train the team on Wednesday, and deploy it to client work on Thursday. Not as a hypothetical — as a literal calendar event. That speed is not a consolation prize. It is a structural advantage that no amount of headcount or budget can replicate.

The firms winning right now are not the ones that have matched Big 4 AI capabilities. They are the ones that have adopted AI faster than Big 4 firms can retrain their own people. The 18-month enterprise deployment cycle is exactly what makes boutique AI adoption valuable — not despite being a small firm, but because of it.

2. You Can Specialize So Deeply That AI Cannot Commoditize You

McKinsey does strategy for pharmaceutical companies, financial institutions, healthcare systems, technology platforms, consumer goods companies, government agencies, and forty other sectors. They have expertise in all of them. They have deep expertise in none of them.

That is not a criticism — it is the structural reality of a firm that needs to deploy 45,000 consultants onto engagements globally. Breadth is the business model.

A boutique firm does not have that constraint. A 10-person law firm that does nothing but employment litigation for mid-sized manufacturing companies in the Southeast knows something no Big 4 firm can replicate: what every comparable case in that specific geography with that specific client type has looked like for the last fifteen years. That institutional knowledge is not in a database. It lives in the heads of three partners who have been doing this work since 2009.

AI can synthesize case law research faster than any paralegal. It cannot replace the partner who knows that the OSHA inspector assigned to your client's region has a specific pattern in enforcement actions, or that the mediator on the case has a consistent settlement preference. That specificity is the product. Big firms cannot manufacture it. You cannot outsource it. Clients will pay for it.

The move: stop describing your firm in generalist terms and double down on what makes it unreplicable. If you are an accounting firm with 80% of clients in commercial real estate, become the accounting firm for commercial real estate investors in your region — not a generalist accounting firm that happens to serve some real estate clients. That shift is in the claim first, not in the work. You may already be doing the work. The competitive advantage is in the specificity of what you say you are.

3. You Can Be the Person Who Answers the Phone

EY has engagement partners, client service teams, escalation chains, and relationship managers. None of them are the person who did the work.

A client with a $4 million problem does not want to talk to an AI interface or a junior associate who will escalate to a partner who will respond within 48 hours. They want to talk to the person who has been managing their account for six years and who knows their situation without needing to be briefed.

That relationship is not transferable. It is not scalable. And no amount of AI investment changes it.

The Clio 2026 Legal Trends Report found that 93% of mid-sized law firms report using AI in some capacity, while only 10% of small law firms have adopted AI extensively. This gap is often framed as a problem — small firms falling behind. The more accurate framing: it is an adoption opportunity. The small firms that close that adoption gap in 2026 will have AI-compressed cost structures and senior relationship depth simultaneously. That combination is structurally unavailable to the Big 4.

Every hour AI saves on commodity work is an hour that can go to a client relationship the Big 4 structurally cannot build. The question is not whether to adopt AI. The question is whether you use AI to free up the right capacity.

---

What Small Firms Are Getting Wrong

The most common response to Big 4 AI news is a reflexive pivot toward looking like a bigger firm. Small accounting firms adding "AI-enabled" to their service descriptions without changing anything. Law firms announcing "AI-assisted document review" to signal sophistication. Consulting firms adding AI to proposal language without actually integrating AI into any workflow.

These moves fail for three reasons.

First, they concede the frame. If you present yourself as a smaller, cheaper version of a Big 4 AI firm, you will lose that comparison. You cannot win a scale contest against organizations with nine-digit AI budgets. The only winning move is to not play that game.

Second, they signal insecurity to the exact clients you want. Sophisticated mid-market buyers who have real options can tell the difference between defensive positioning and genuine capability. "AI-enabled" in your pitch deck reads as a response to Big 4 news, not as a real answer to their problem.

Third, they delay the adaptation that actually matters. The time spent polishing marketing language is time not spent redesigning one workflow that would save 10 hours per week and let you serve two more clients at your current headcount.

The other common mistake: chasing clients the same size as Big 4 clients. A 12-person firm pursuing Fortune 500 mandates that EY is also pursuing will lose on brand, bench depth, and now increasingly on AI infrastructure. The correct move is down-market in client size and up-market in specialization depth — smaller clients who need more specific expertise than any Big 4 firm would deploy senior talent to develop.

---

The 3-Phase Differentiation Playbook

This is a sequenced set of actions, sized for a firm with 5 to 25 people, executable without a dedicated operations team or a technology budget.

Phase 1: Automate the Commodity Layer (Days 1–60)

Big 4 firms are using AI to reduce the cost of commodity work: document synthesis, research aggregation, compliance checklists, first-draft client communications, standard reporting. These categories were always the most time-intensive and least differentiated work in any professional services firm. Automating them does not change your competitive position relative to Big 4 firms — but it compresses your cost structure and frees senior capacity for the work that actually differentiates you.

The starting point: identify the three tasks in your firm that consume the most hours relative to the value they create. These are almost always administrative or commodity professional work — not the specialized judgment calls clients are actually paying for.

For accounting firms: client document intake and organization, initial tax return data entry, standard reconciliation, status update client communications.

For law firms: discovery document review, contract clause comparison, case law research, standard form and pleading generation.

For consulting firms: project kickoff documentation, meeting note synthesis and action item tracking, market research aggregation, client status reporting.

Tools that handle these tasks cost between $0 and $200 per month and are available to any firm regardless of size. The real investment is not money — it is two days to redesign the workflow once. A 10-hour workflow that becomes a 2-hour workflow creates 8 hours of recovered senior capacity per instance. If that workflow runs 20 times per month, you have created 160 hours of available capacity without a hire.

Big 4 firms are spending billions to achieve the same compression. You can accomplish it with a focused workflow redesign sprint and a $20/month subscription.

Phase 2: Deepen One Specialization That AI Cannot Replicate (Days 30–180)

Commodity automation creates time. Phase 2 determines what you do with it.

The winning move: use that recovered capacity to deepen expertise in one specific niche that Big 4 firms cannot credibly serve with the same depth. This is not adding a new service — it is becoming more specific in the service you already offer.

An accounting firm serving 40 different industries in a general practice model is difficult to distinguish from any other accounting firm. An accounting firm serving specifically cannabis operators in licensed states has built expertise in the specific tax treatment of Schedule I businesses, the banking constraints for cannabis, the state-by-state licensing accounting requirements, and the audit exposure that comes with federal illegality. No Big 4 firm deploys senior partners to build that specialization for a market that size. But dozens of mid-sized cannabis operators need exactly that expertise — and they will pay a premium to the firm that has it.

The specialization does not need to be that dramatic. It needs to be specific enough that you can make a credible claim that no other firm in your region has made more investments in it than you have. The test: if you described your firm's specialization in one sentence to a prospect in that niche, would they say "that's exactly what I've been looking for" — or "yes, but what do you do specifically?"

Phase 3: Position the Specialization (Days 90–Ongoing)

Specialization that clients do not know about does not differentiate you in a competitive situation. Phase 3 is making the claim visible and consistent.

This is where most small firms underperform. They develop genuine deep expertise, then describe themselves in the same generic terms as every other firm in the directory: "We are a full-service accounting firm serving businesses of all sizes." That language communicates nothing and converts no one.

The alternative: "We are the accounting firm for commercial real estate investors in [region]. We have [X] clients in the sector and have handled [specific transaction type] for [Y years]." That claim cannot be made by EY — not because EY lacks accounting expertise, but because EY cannot make a credible specificity claim about a regional niche. You can.

The distribution channels for this positioning: client referral language (train your clients to describe you the way you want to be described), LinkedIn content targeting the specific sector, presence at the sector-specific association conference, and a one-paragraph specialization statement in every engagement letter and proposal. None of this requires a marketing budget. It requires clarity about what you are and the discipline to communicate it consistently in every context.

For related context on how specialization intersects with pricing model — including the valuation gap between generalist and specialist firms — see How to Move Your Consulting Firm Off Hourly Billing.

---

FAQ

What are the actual competitive advantages a 10-person professional services firm has over EY or McKinsey in 2026?

Four advantages, each defensible and specific.

Speed of deployment. A 10-person firm can evaluate a new AI workflow on Monday and run it on client work Thursday. EY's internal AI deployment, by contrast, requires IT security review, risk and compliance sign-off, regional partner approval, and a global training rollout — typically 12 to 18 months from evaluation to widespread adoption. For any category where AI tools are evolving rapidly, the small firm's deployment speed is a structural competitive advantage that no amount of enterprise AI budget can eliminate.

Specialization depth. Big 4 firms need to staff tens of thousands of engagements globally per year. That requires generalist bench strength — consultants and accountants who can work across industries and engagement types. Boutique firms are not constrained by staffing scalability. A 10-person law firm can build expertise so narrow that no Big 4 firm would invest in developing it for a market of that size. That depth of specialization is precisely what sophisticated clients in specific niches pay a premium for — and what no AI deployment can simulate if the knowledge does not exist in the firm to begin with.

Trust proximity. Research across professional services consistently finds that the senior relationship — the partner who knows the client's business — is the primary driver of both retention and referrals. At a 10-person firm, that senior relationship is present on every engagement, accessible for every question, and cannot be replaced by an engagement team structure or escalation chain. Big 4 firms have relationship managers; small firms have the person.

Pricing flexibility. Big 4 firms carry overhead structures — global real estate, technology infrastructure, partner distributions — that establish a minimum pricing floor. A small firm with lower overhead can price engagements strategically when the relationship warrants it, offer hybrid structures that Big 4 billing frameworks cannot accommodate, and move immediately on pricing decisions without partner committee approval.

The Clio 2026 data point that surprises most small firm owners: 93% of mid-sized law firms report using AI in some capacity, while only 10% of small law firms have adopted AI extensively. The standard framing is that small firms are behind. The more strategic reading: the small firms that close that adoption gap in 2026 will have AI-compressed cost structures and senior relationship depth simultaneously. That combination is structurally unavailable to any firm large enough to require an 18-month AI deployment cycle.

How are big 4 firms using AI in 2026, and does it actually threaten small professional services firms?

The specific deployments, as of August 2026:

KPMG deployed Claude AI to 276,000 employees through a partnership with Anthropic — the largest enterprise AI deployment in professional services. Primary use cases include internal research synthesis, audit documentation support, and client reporting.

EY announced 150 AI agents handling internal knowledge management, training content generation, and engagement support tasks.

McKinsey has restructured portions of its workforce — reducing approximately 1,400 positions — and shifted practice areas toward outcome-based project delivery, in part because AI has compressed the time required for deliverables that previously justified large analyst teams.

What this means for small firms depends entirely on firm type:

Small accounting firms: The more immediate threat is AI-native competitors — Pilot, Accrual, and similar platforms — that are building direct-to-SMB accounting services at price points below what most small accounting firms can reach. KPMG's AI deployment is primarily targeted at larger clients. The risk for small accounting firms is the commodity compliance layer eroding over 2 to 3 years, which makes specialization even more urgent.

Small law firms: BigLaw AI tools (Harvey, CoCounsel) compress the time required for discovery, contract review, and legal research. This creates pricing pressure on the commodity end of legal services — firms whose value proposition is speed and volume of document work. Firms whose differentiation is relationship-based, deeply specialized, or contingency-driven are substantially less exposed.

Boutique consulting: Margin compression from faster project delivery is already visible. Clients who previously accepted 6 to 8 week research-heavy engagements now expect faster turnaround. The risk is not replacement — it is that firms that have not also compressed their own delivery costs are losing margin to the same AI they are not using.

The conclusion: the threat is real and specific at the commodity layer. For any firm that has built genuine specialization and is using AI to enhance that specialization rather than substitute for it, the direct threat from Big 4 AI deployment is minimal.

What workflows should a small professional services firm automate first to compete?

Start with the workflows where Big 4 firms are gaining the most cost advantage — because those are the same workflows costing you the most time relative to the value they produce.

Document review and first-pass synthesis. AI can read, organize, and summarize large document sets faster than any associate at any experience level. For accounting firms, this is client intake documentation, bank statement organization, and prior return analysis. For law firms, this is discovery review, contract redline comparison, and case law research. For consulting firms, this is competitive research aggregation, interview note synthesis, and initial findings organization.

Client communication drafts. Status update emails, meeting follow-up summaries, proposal first drafts, and standard client responses are the category where AI delivers consistent quality with minimal oversight once templates are established. Build the template, refine it over three to four actual uses, and the time savings compound quickly across every client.

Compliance checklists and standard procedures. Any workflow that runs on a standardized checklist is an automation candidate. Payroll compliance reviews, quarterly reporting, audit preparation checklists, standard contract language reviews — AI can handle the checklist execution while your senior people focus on the exceptions and judgment calls.

Meeting documentation. AI transcription and summary tools (Otter, Granola, Fireflies) are the lowest-friction automation entry point for most professional services firms. Setup is a browser extension or calendar integration, the quality of output is usable immediately, and the workflow change required is minimal. The time savings are visible in the first week.

The prioritization logic: automate the commodity layer first, then redirect the recovered capacity toward specialized, relationship-intensive work that no AI deployment and no Big 4 firm can replicate at your depth. The goal is not to automate more than your competitors — it is to automate the right things so that senior capacity is available for the work that actually differentiates your firm.

Should a small firm try to use the same AI tools as big 4 firms — Harvey, CoCounsel, Microsoft Copilot — or different tools?

Tool choice is secondary to workflow design. The question to ask is not "which tool is most impressive" but "which tool will my team actually use consistently, in the workflows that matter most."

Many of the same tools available to Big 4 firms are accessible to any professional services firm:

• Law firms: Harvey and CoCounsel are available to firms of any size. Clio Duo integrates directly into practice management. Claude and ChatGPT handle research synthesis, drafting, and client communications at a monthly cost comparable to a professional association membership.
• Accounting firms: Intuit Assist and similar tools are integrated into QuickBooks and major tax platforms. Karbon AI is available for workflow management. Claude and ChatGPT handle client communication drafting, research synthesis, and explanation generation.
• Consulting and advisory firms: Perplexity for research synthesis, Claude for drafting and analysis, Notion AI for documentation and knowledge management — all available without enterprise licensing.

The actual differentiator between KPMG's AI advantage and yours is not access to Claude. It is that KPMG has built the internal workflows, training programs, quality-control checkpoints, and output review processes that make AI output usable at scale across 276,000 people. That infrastructure is what makes the deployment valuable — and a small firm can build a version of it for one team in one workflow in two weeks.

Recommendation: pick two tools. One for research synthesis. One that integrates directly into your primary workflow platform — practice management, document management, or CRM. Use both consistently for 60 days before evaluating whether to add anything else. Tool breadth does not create competitive advantage. Depth of workflow integration does.

How do I tell clients that my small firm is as capable as a big 4 firm?

Do not make a direct capability comparison. In any direct comparison between a 10-person firm and KPMG on the basis of general professional quality, the 10-person firm loses. That is not a criticism — it is a positioning reality.

The winning frame is not "we are as capable as the Big 4." It is "we are more specialized than the Big 4 for your specific situation." These are categorically different claims with different competitive dynamics.

EY can say: "We have 395,000 professionals with expertise across every industry and service line in 150 countries." You cannot match that claim, and you should not try.

You can say: "We are the accounting firm for family-owned manufacturing companies in [region]. We have handled [X] succession transactions, [Y] sale processes for manufacturers in your revenue range, and we know the specific tax structure your industry uses for ESOP transitions. EY has a team that can serve you. We have specialists who have only served businesses like yours for the past twelve years."

KPMG cannot make that specificity claim. That claim is only credible coming from a firm that has actually made the commitment — in staffing, in continuing education, in client selection, in the cases they have taken and the ones they turned down — to one specific niche.

The practical application: audit your current client base. If 60% of your clients share an industry type, geography, or business situation, you may already be specialized and not communicating it. Rewrite your website headline, your LinkedIn summary, and the first paragraph of your new client proposal to lead with the specialization claim. Specificity converts better than quality claims every time — because only one firm in your market can make your specific claim, and any Big 4 firm can match any general quality claim.

What does the timeline look like — how long before small professional services firms that don't adapt to AI will lose to big 4?

The timeline varies by firm type, and the mechanism of pressure varies too — which matters for deciding which adaptations are most urgent.

Accounting: 2 to 3 years before AI-native competitors (Pilot, Accrual, and similar platforms) commoditize standard compliance work at price points below what most small accounting firms can sustainably match. Big 4 AI deployment is a less immediate direct threat to small accounting clients — KPMG is not pursuing your $3 million manufacturing client. The more pressing risk is AI-native firms offering bookkeeping, payroll, and standard compliance at dramatically lower price points, targeting the clients small accounting firms rely on for volume.

Law: 3 to 5 years before BigLaw AI efficiency gains create meaningful pricing pressure on commodity legal work at the small firm level. The near-term risk is not losing clients to BigLaw — it is losing price-sensitive clients to AI-augmented solo practitioners and alternative legal service providers who use AI to offer standard services at substantially lower prices. Firms whose differentiation is relationship-based or highly specialized are substantially less exposed on a 3-year horizon.

Consulting: Margin compression from faster project delivery expectations is already visible in 2026. Clients who previously accepted 6 to 8 week strategy engagements now expect faster turnaround on analysis and synthesis work. The firms losing ground now are those that have not also compressed their own delivery costs — meaning clients are seeing slower work at the same price from a consultant who is not using AI, compared to faster work at competitive prices from one who is.

Staffing: The most acute timeline across all firm types. Analysis from Aqore in 2026 suggests AI is handling 80% of transactional staffing tasks in firms that have adopted it. Transactional placement — the commodity end of the staffing market — is under significant competitive pressure now, not in three years. For a deeper look at the strategic reset for staffing firms, see The AI Staffing Pivot.

The firms at risk across all categories are those with no clear specialization and no AI workflow adoption — running generalist practices at manual-process cost structures in a market where both of those attributes are becoming liabilities simultaneously.

The firms that will come through this cycle are the ones using AI to deliver more of the commodity layer faster, then redirecting that recovered capacity toward the specialized, relationship-intensive work that no AI deployment — Big 4 or otherwise — can replicate.

---

The Action This Week

Pick one workflow in your firm that runs more than five times per week and requires more than one hour of non-specialist time. Write one sentence describing it: "Every [time period], someone on our team spends [X hours] doing [specific task] before a senior person can do [the actual valuable work]."

That sentence is your automation brief. Spend two hours this week testing one AI tool on that specific task — not the most impressive tool, the one your team will actually use on Tuesday morning. Run it on three real examples from the past month. Measure the time saved against your current process.

If it saves more than 50% of the time, implement it as the standard workflow.

That recovered capacity is the raw material for everything else in this playbook. You do not need a Big 4 budget to compete with a Big 4 firm. You need one automated workflow that frees three hours of senior time per week — and a specific, defensible answer to the question every client in your niche is eventually going to ask: "Why you, specifically, for this?"

The Big 4 can spend billions building AI infrastructure. They cannot manufacture a genuine answer to that question for your clients. You already have the answer. The work is making sure it is visible.

For the pricing model that pairs with this differentiation strategy, see How to Move Your Consulting Firm Off Hourly Billing: A 90-Day Transition Framework — the same logic that applies to how you price applies to how you position.

---

Related Reading

• How to Move Your Consulting Firm Off Hourly Billing: A 90-Day Transition Framework — The valuation gap between generalist and specialist firms, and the 90-day roadmap for making the pricing transition
• The AI Staffing Pivot: From Transactional Placement to Workforce Advisory — The specific reset for staffing firms facing AI displacement of the transactional placement model
• Measuring AI ROI in Professional Services Firms — How to measure the return on AI workflow adoption before making the business case internally

---

The Crossing Report helps professional services firm owners navigate the AI transition. For weekly intelligence on what's changing and what to do about it, subscribe here.