The Big Firms Are Struggling to Become What You Already Are

Published: March 17, 2026 | By: The Crossing Report | 8 min read

---

Summary

McKinsey, Starmind, and half a dozen consulting research firms have identified the same structural shift: AI is forcing the traditional professional services pyramid to collapse into an "obelisk." Wide base of junior staff doing rote production work. Narrow apex of senior partners doing judgment work. AI automates the base. The structure gets flatter, smaller, and more senior-heavy. The problem: large firms are struggling to execute this transition while managing hundreds of clients, partner compensation systems built around hourly billing, and large cohorts of junior associates whose work is being automated. Small professional services firms — 5 to 20 people — are already closer to the obelisk than any large firm will be for years. The transition that's painful and expensive for the Big 4 is a practical upgrade for a firm that never built the pyramid in the first place.

---

The Story Everyone Gets Backwards

The dominant narrative about small professional services firms and AI goes like this: large firms have resources, technology teams, and innovation departments. They'll adopt AI first, get more efficient, undercut your pricing, and take your clients. You're behind, and falling further behind every quarter.

There's some truth in that framing for certain use cases. Big firms are buying enterprise AI tools at scale and deploying them on complex engagements first.

But there's a different story in the structural research that almost never makes it into the conversation aimed at small firm owners — and it goes in the opposite direction.

The Big 4, the AmLaw 100, and the top management consulting firms are spending billions of dollars trying to become what you already are.

---

The Pyramid Problem They Have That You Don't

The traditional professional services business model is shaped like a pyramid. Wide at the bottom: large numbers of junior associates and staff doing high-volume, lower-judgment work. Research synthesis. Document review. First-draft contracts. Reconciliation runs. Initial screening work. The work that has clear procedures and can be taught to someone in their first year.

Narrow at the top: senior practitioners and partners doing the judgment-intensive work that actually requires expertise. Complex client problems. Strategic advice. High-stakes decisions. Relationship management.

The pyramid works because the junior-level work — though relatively low-margin — creates volume leverage and a training pipeline for developing senior talent. Junior associates do the first pass. Partners review and advise.

AI automates the base.

Not eventually. Now. The document review that used to require a team of first-year associates can be handled by an AI model in hours. The reconciliation runs that occupied three staff accountants can be automated with the right tool. The research synthesis that took a junior consultant two days can be drafted by an AI in an afternoon.

When the base of the pyramid becomes machine-executable, the pyramid is no longer the right shape for the organization.

---

The Obelisk: What the Research Says

Starmind, McKinsey, and multiple consulting research organizations have published analyses identifying the structure that replaces the pyramid. The working term is the "obelisk."

Flatter. Narrower overall. Three distinct roles rather than a broad junior tier.

The first role is the AI facilitator: junior and mid-level professionals whose job shifts from producing first drafts to directing AI tools, reviewing AI output for accuracy and fit, managing the workflow between AI systems and human review, and catching the errors and blind spots that AI produces. This is not a demotion from the old junior role — it's a different skill set that makes the junior professional a multiplier on the firm's output rather than a production unit.

The second role is the engagement architect: senior practitioners who frame the problem before AI works on it, interpret what the AI has produced, identify what's missing or wrong, and translate it into client-ready work. This role exists in the current pyramid — it's the manager or senior associate layer. In the obelisk, it becomes more important and requires better judgment, because the AI is doing the production work and someone needs to supervise quality at a level that matters.

The third role is the client leader: the partner, principal, or senior professional who holds the relationship, manages trust, and makes the calls that require institutional credibility. This role is unchanged by AI. Clients don't want to talk to a model. They want to talk to a person whose judgment they trust and whose professional reputation is on the line.

That's the obelisk: AI facilitator, engagement architect, client leader. Three distinct roles. Fewer total headcount. Higher output per person. More expensive per employee. More demanding of everyone on the team.

---

Why Large Firms Are Struggling With This

Building an obelisk when you're currently a pyramid is painful.

A Big 4 accounting firm has thousands of junior associates whose primary work is being automated. You can't eliminate them overnight — you have client commitments, professional development obligations, and institutional inertia. You can't easily retrain them all into AI facilitators in a single cohort — that requires curriculum development, management culture change, and a wholesale revision of how work gets assigned and reviewed. And you can't revise partner compensation systems that have been built around billable hour volume without a political fight that takes years to resolve.

The result: these firms are spending enormous sums on AI tools while simultaneously running the old staffing model that assumes humans do the production work. They're adding AI costs without yet capturing AI efficiencies. The transition period is expensive and organizationally disruptive.

And they're doing it in public, with clients watching.

---

The Small Firm Structural Advantage

Here's what a 10-person accounting firm, law firm, or consulting practice already has by default.

You have no junior associate bench to dismantle. You've never had 30 first-year associates doing document review on rotation. Your team has always been senior-weighted relative to firm size. The production work you do has always required more judgment per hour than a BigLaw associate rotation, because you couldn't afford to have people on projects who weren't pulling weight.

You have no partner compensation system built around billable hours at scale. When you implement AI and complete the same client engagement in 20% less time, you don't have to convince a 50-person partnership to revise its compensation methodology. You make a business decision about how to price the outcome versus the hours.

You have no organizational change management problem. You have a management team of roughly one. If you decide to change how your firm uses AI, you change how your firm uses AI. The implementation timeline is weeks, not years.

You are, structurally, already much closer to the obelisk than any large firm currently is. You have a client leader (you and your senior partners). You have engagement architects (your experienced professionals). You need AI facilitators — junior or mid-level staff who direct AI tools and review their output — and you may not have thought about hiring for that role yet. But that's a hiring frame, not a structural dismantling.

---

What the Transition Looks Like for You

The firms that are capturing this advantage right now are doing three specific things.

They're deploying AI for first-draft production. Every deliverable that currently starts with a human at a blank page should now start with an AI draft that a human reviews and improves. For accounting firms: first-draft financial analyses, first-draft client communications, first-pass anomaly identification in reconciliation. For law firms: first-draft contracts, first-draft briefs, first-pass contract review flagging issues for attorney review. For consulting firms: first-draft research synthesis, first-draft recommendation frameworks, first-draft client decks. The senior professional's time shifts from producing to evaluating and improving — which is a better use of their expertise.

They're reframing junior roles as reviewers, not producers. When a junior professional's job shifts from "prepare the reconciliation" to "review the AI's reconciliation for errors, edge cases, and client-specific context the AI doesn't have," the role becomes more cognitively demanding and more valuable, not less. This is the AI facilitator in practice. The firms doing this well are making it explicit in how they describe the role and how they give feedback.

They're using freed capacity to take on more clients or more complex work, not to reduce headcount. The firms that try to use AI efficiency to cut team size often discover they've eliminated important institutional knowledge and relationship context — and then can't scale back up when growth requires it. The smarter use of AI-generated capacity is expansion: more clients, deeper client relationships, more complex engagements that command better margins.

---

The Counterintuitive Positioning Opportunity

One more thing worth naming: this structural reality is a positioning asset, and almost no small firm owner is using it.

When a CFO at a 100-person company asks why they should use you instead of a Big 4 or a large regional firm, you probably talk about responsiveness, attention, lower fees, and sector expertise. All of those are legitimate.

But there's a deeper answer available to you now that you couldn't have given three years ago.

Large firms are in the middle of a structural transition that is complicated and expensive and still unresolved. They're running AI tools while managing the organizational costs of moving from the pyramid they built to the obelisk the market needs. That transition is visible to sophisticated clients in the form of mixed AI output quality, inconsistent staffing on engagements, and senior partners who are selling work that junior staff are executing.

Your firm doesn't have that transition problem. Your structure is already compatible with how AI-enhanced professional services work gets delivered best. You're senior-weighted. Your client leader is also your engagement architect. You have direct accountability on every file. When you implement AI tools, you implement them into a structure that was already doing the work the hard way — and they make the hard work faster.

The big firms are spending billions to become what you already are. That's not a threat. That's a competitive position.

---

What to Do This Week

Two things.

First, implement AI for first drafts. Pick the type of deliverable your firm produces most often — client analysis, contract review, financial summary, research memo — and this week use an AI tool to generate a first draft before a human starts writing. Review it. Edit it. See how much time you save and what the quality gap is. That's the experiment that tells you where to invest next.

Second, tell clients about your structure. Not in technology jargon — in terms of how their work gets done. "Every engagement is led by a senior professional from start to finish, and we use AI to ensure our senior professionals can focus on judgment and analysis rather than production work." That's the obelisk in plain English. It's a differentiator. Most large firms can't say it honestly right now.

You don't need a transformation. You need to recognize what you already have — and use it.

---

The Crossing Report covers AI adoption for owners of professional services firms — accounting, law, consulting, and staffing — with 5–50 employees.