The pyramid cracks. What agentic AI does to the consulting leverage model.

📊 Full opportunity report: The pyramid cracks. What agentic AI does to the consulting leverage model. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.

TL;DR

AI is fundamentally altering the consulting industry’s leverage model by commoditizing analysis and boosting deployment work. Firms focused on analysis face margin pressure, while those specializing in implementation benefit. The industry is splitting, not shrinking, with long-term talent pipeline impacts.

Generative AI is significantly disrupting the traditional consulting leverage model, with firms experiencing varied impacts based on their core value propositions. The industry is splitting into distinct segments, with analysis-focused firms under margin pressure and deployment-centric firms expanding.

The consulting industry has long operated on a pyramid structure, where a large base of junior analysts and associates perform document-heavy, repetitive work that is highly commoditized by AI. This base fuels the profitability of top-tier partners through leveraged billable hours. Recent developments show that AI tools, especially generative models, are reducing the need for human analysis, leading to headcount reductions at firms like McKinsey and KPMG. Conversely, firms that focus on large-scale implementation, change management, and AI deployment—such as Accenture—are experiencing growth, with record bookings and expanding AI and data teams.

This divergence indicates a reallocation of value within the industry rather than an overall contraction. Firms with a strong emphasis on analysis are facing margin compression and a broken talent pipeline, as AI commoditizes the work that traditionally trained analysts. Meanwhile, firms that excel at deploying AI at scale are capturing new revenue streams, as deployment work is now more valuable and less replaceable by AI itself. This shift is leading to a structural split in the industry, with the traditional pyramid model under threat and new opportunities emerging in AI deployment services.

The Pyramid Cracks — Thorsten Meyer AI
BILLABLE
● DISPATCH / MAY 2026
THORSTEN MEYER AI · ENTERPRISE REORG · § 02
ENTERPRISE REORG · 02
CONSULTING / COMPRESSION
Essay · Professional-Services Structural Reading · 2026-05-22

The pyramid cracks.
What agentic AI does
to the consulting
leverage model.

Consulting’s profit was always the spread on a base of juniors doing exactly the work AI now does. The base is the most AI-exposed structure in professional services.
The consulting business is a leverage pyramid: a few partners over a wide base of billable juniors, billed out at a multiple of cost. The base does the document-heavy analytical work — research, synthesis, modeling, slides — which is exactly what generative AI does best. McKinsey’s own research puts the compression at 30%+ on a typical engagement; the firm has pulled headcount from 45,000 toward 40,000, KPMG cut ~400 advisory jobs and ~10% of US audit partners. But the compression is not uniform — that is the whole story. Pure-strategy MBB grows at 5-6% while execution firms grow at 11-12%: Accenture booked a record $22.1B with 85,000+ AI professionals. The structural argument: AI does not shrink consulting so much as split it by DNA — compressing the firms whose product was analysis, feeding the firms whose product is deployment, squeezing the labor-arbitrage IT tier between them. And the base of the pyramid was never just a billing layer. It was the machine that made the partners.
30%+
Research-synthesis compression
per McKinsey’s own Quantum Black
45K→40K
McKinsey headcount · ~10% more
non-client-facing cuts coming
$22.1B
Accenture record quarterly bookings
85,000+ AI & data professionals
5-6 / 11-12
MBB growth % vs execution-firm
growth % — the compression, visible
THE PYRAMID CRACKS· THE LEVERAGE MODEL MEETS THE AGENT· 30%+ RESEARCH COMPRESSION· MCKINSEY 45K → 40K· ~10% NON-CLIENT-FACING CUT· KPMG ~400 ADVISORY + 10% AUDIT PARTNERS· ACCENTURE RECORD $22.1B BOOKINGS· 85,000+ AI & DATA PROFESSIONALS· MBB 5-6% VS EXECUTION 11-12%· 3 ASSOCIATES + AI = 10 ASSOCIATES· THE LEVERAGE RATIO INVERTS· TCS $29B · INFOSYS $19B · WIPRO $11B· 20-30% LOWER PRICE POINTS· ANALYSIS COMMODITIZED · DEPLOYMENT NEW· THE 1:6 RATIO COLLAPSES AND RE-FORMS· THE BASE IS THE PARTNER PIPELINE· SPLIT BY DNA · NOT A CONTRACTION· GARTNER AI SPEND +44% TO $2.52T· THE PYRAMID CRACKS· THE LEVERAGE MODEL MEETS THE AGENT· 30%+ RESEARCH COMPRESSION· MCKINSEY 45K → 40K· ~10% NON-CLIENT-FACING CUT· KPMG ~400 ADVISORY + 10% AUDIT PARTNERS· ACCENTURE RECORD $22.1B BOOKINGS· 85,000+ AI & DATA PROFESSIONALS· MBB 5-6% VS EXECUTION 11-12%· 3 ASSOCIATES + AI = 10 ASSOCIATES· THE LEVERAGE RATIO INVERTS· TCS $29B · INFOSYS $19B · WIPRO $11B· 20-30% LOWER PRICE POINTS· ANALYSIS COMMODITIZED · DEPLOYMENT NEW· THE 1:6 RATIO COLLAPSES AND RE-FORMS· THE BASE IS THE PARTNER PIPELINE· SPLIT BY DNA · NOT A CONTRACTION· GARTNER AI SPEND +44% TO $2.52T·
FIG. 01 — THE LEVERAGE PYRAMID
The profit is the spread on the base, multiplied by the size of the base
The leverage ratio — juniors per partner — is the single most important number in the firm’s economics
PartnersJudgment · relationship · origination
Bill 1, oversee 10
Managers / PrincipalsPackage · oversee · QA
Mid-leverage
AssociatesRefine · model · structure
Billable
Analysts — the baseResearch · synthesis · modeling · slides
Most automatable
A partner overseeing ten associates bills out eleven people’s hours while personally working one person’s. The profit is not the partner’s billing rate; it is the spread on the base, multiplied by the size of the base. The dirty secret of the model: much of what the base produces is not irreplaceable insight — it is the structured labor of turning information into a presentable analysis, the layer with the highest ratio of process-to-judgment and therefore the highest exposure to automation. The pyramid concentrates a firm’s billing in precisely the layer whose work is most automatable.
FIG. 02 — THE BASE UNDER ATTACK · THE LEVERAGE-RATIO MATH
The brutal arithmetic that makes consulting partners nervous
The technology that makes the partner more productive makes the base redundant — and the base was the profit engine
10
Associates needed
before AI
3
Associates + AI tool
for the same output
If three associates plus an AI tool produce what ten associates used to produce, the engagement needs three associates. Multiply across hundreds of engagements and tens of thousands of staff, and the leverage ratio that funded the pyramid inverts from an asset into a liability. The hiring signal confirms it: job postings that once asked for Excel modeling now ask for prompt design and AI-output validation — roughly one in four entry-level consulting/finance postings now require AI fluency, up from fewer than one in twenty two years ago. The junior job is being redefined from “produce the analysis” to “direct and validate the machine,” which needs far fewer people.
FIG. 03 — THE CUTS ALREADY LANDING · SAME TECHNOLOGY, THREE PAYROLL OUTCOMES
The compression has moved from forecast to payroll
Cut the back office and lower-performing base, redefine the rest, frame it as realignment
FIRM
WHAT HAPPENED
DIRECTION
McKinsey
17K → 45K → ~40K · ~10% non-client-facing cut over 18-24 months · 200 tech cuts late 2025 · revenue flatlined
Cutting
KPMG
~400 US advisory jobs (half lower-performers, no partners) · ~10% of US audit partners (~100) · “strategic realignment”
Cutting
Deloitte / EY / PwC
All rolled out AI assistants, trimmed back-office · PwC abandoned hiring target · PwC Office-of-CFO unit + 30K certified on Claude
Hedged
Accenture
Record $22.1B bookings (+6%), 41 deals >$100M · 85,000+ AI/data professionals · “use AI to be promoted” · exiting non-retrainable staff
Hiring
What is consistent: cut the base and the back office, redefine the survivors around AI, frame it as realignment. What differs is the DNA underneath. McKinsey cuts because the work it sells is the work AI commoditizes; the Big Four trim selectively because their audit-and-execution mix is hedged; Accenture hires because the work it sells is the work AI creates demand for. The headcount numbers are the surface; the DNA underneath them is the story.
FIG. 04 — THE SPLIT BY DNA · THE THREE-TIER COMPRESSION MAP
Stop treating consulting as one industry · it is three businesses with three relationships to AI
The compression lands in inverse proportion to execution capability
Tier 1 · Most exposed
Pure strategy advisory
McKinsey · BCG · Bain
Product is analysis — exactly what AI commoditizes. Economics depend most on the leverage pyramid. The “tell us what the data says” engagement compresses.
5-6%Growth · the compression visible
Tier 2 · The winners
Execution & implementation
Accenture · Deloitte · EY
Product is deployment — data cleanup, integration, change management, AI scaling. New work AI cannot do for itself. GenAI bookings <5% of a $200B+ market: long runway.
11-12%Growth · capturing deployment
Tier 3 · Squeezed both sides
Labor-arbitrage IT
TCS · Infosys · Wipro · Capgemini
AI deflates the bodies-in-seats model from below; premium players take high-value AI work from above. TCS $29B / Infosys $19B / Wipro $11B · 20-30% lower price points.
±0%The vise · pivoting to managed AI
The same technology, applied to three different business models, produces compression, growth, and a vise. Reading the industry as one business is the error that makes the headcount numbers look contradictory. Reading it as three makes them obvious. The pure-advisory pyramid (analysis is the product) compresses hardest; execution (deployment is the product) grows; labor-arbitrage (bodies are the product) is squeezed between AI taking the commodity work and premium players taking the premium work.
FIG. 05 — THE TALENT-PIPELINE RUPTURE · THE COST THE NUMBERS HIDE
The base of the pyramid is not just a billing layer — it is the partner pipeline
The headcount cuts are visible · the pipeline rupture is invisible · which is exactly why it is more dangerous
The pyramid is an apprenticeship machine · nobody is hired as a partner · a partner is an analyst who survived a decade of base work, learning judgment by doing it
The mechanism
AI eliminates the analyst work · the firm hires fewer analysts · but the analyst job was where future partners learned judgment by grinding through the analysis
First-order
The validation paradox · the surviving junior job is to validate AI output — but validating output well requires the expertise that used to come from producing it
The catch
A thin manager class, a thinner future-partner class · you cannot hire a ten-year-experienced partner who never existed · the gap surfaces and cannot be quickly repaired
2030s
The firms are optimizing the first-order cost — fewer juniors, higher margin now — and deferring the second-order cost — fewer trained seniors later. The pyramid is an apprenticeship machine disguised as a billing machine, and hollowing out the base to capture the margin gain quietly disables the machine that produces the people the firm cannot function without. That cost is real, large, and absent from every quarterly number.
The compression is a reallocation, not a contraction. The demand for help migrates from analysis — which AI commoditizes — to deployment — which AI creates demand for. The pyramid that monetized analysis-by-juniors compresses. The firm that monetizes deployment-at-scale grows.
Thorsten Meyer · The Pyramid Cracks · Enterprise Reorg 02

Impacts on Industry Structure and Talent Pipelines

This transformation matters because it signals a fundamental shift in how consulting firms generate revenue and develop talent. The erosion of the analyst base threatens the long-term pipeline of partners, risking a hollowing out of the top leadership over time. Additionally, the industry split creates winners and losers: firms focused on analysis face margin squeeze and talent shortages, while those specializing in AI deployment gain strategic advantage and market share. The long-term implications include potential industry fragmentation and a reevaluation of traditional consulting value propositions.

Amazon

AI-powered consulting analysis software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Industry Evolution and AI’s Role in Reshaping Consulting

Historically, the consulting industry relied on a pyramid structure where junior analysts performed high-volume, document-heavy work, which was then leveraged by senior partners to generate profits. Recent advances in generative AI have commoditized much of this work, leading to headcount reductions at major firms like McKinsey, which has cut non-client-facing roles by roughly 10%. Simultaneously, firms like Accenture, with a focus on large-scale AI deployment, are expanding their AI and data teams, signaling a shift in industry focus from analysis to implementation. This evolving landscape is driven by AI’s ability to perform tasks that once required human labor, fundamentally altering industry economics.

“The leverage pyramid that defined elite consulting is the most exposed structure in professional services, because its economics depend on billing out a large base of juniors doing exactly the work AI now does.”

— Thorsten Meyer

Your AI Survival Guide: Scraped Knees, Bruised Elbows, and Lessons Learned from Real-World AI Deployments

Your AI Survival Guide: Scraped Knees, Bruised Elbows, and Lessons Learned from Real-World AI Deployments

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Long-Term Industry and Talent Pipeline Risks

It remains unclear how deeply the industry will fragment over the next decade, and whether firms can successfully pivot from analysis to deployment at scale. The precise long-term impact on partner development and firm profitability is still uncertain, as is the full extent of talent pipeline disruption caused by the erosion of the analyst base.

Amazon

AI document automation tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Industry Reorganization and Future Growth Opportunities

Firms will likely continue to adjust their structures, with analysis-focused firms shrinking or diversifying, while deployment-centric firms expand. The industry may see increased specialization, with new roles and revenue streams emerging around AI deployment and scaling. Monitoring these shifts will be critical as firms seek to sustain growth and talent pipelines in a rapidly changing environment.

Amazon

AI change management software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

How is AI reducing the need for analysts in consulting?

Generative AI automates research, synthesis, modeling, and document production, tasks traditionally performed by analysts, leading to headcount reductions and margin compression in analysis-heavy firms.

Why are some consulting firms growing while others shrink?

Firms focused on large-scale AI deployment, change management, and implementation are expanding because these services are less commoditized and more valuable in the AI era, unlike analysis work which AI can perform more efficiently.

What are the long-term risks for the consulting industry?

The erosion of the analyst pipeline could lead to a shortage of future partners and leadership, potentially destabilizing traditional firm structures and profitability over time.

Will the industry fully split or consolidate?

Current trends suggest a split into distinct segments—analysis-focused and deployment-focused firms—though future consolidation or further fragmentation remains uncertain.

How might firms adapt to these changes?

Firms may need to diversify their service offerings, invest in AI deployment capabilities, and rethink talent development pipelines to remain competitive in the evolving landscape.

Source: ThorstenMeyerAI.com

You May Also Like

The citation. Why generative engine optimization rewards the same brand on the least stable ground.

An analysis of how GEO favors established brands, the instability of citations, and what this means for publishers and SEO strategies.

The queue. Why the grid, not the chip, is the binding constraint on AI.

US interconnection queues have become the primary bottleneck for AI data-center expansion, shifting the constraint from chip supply to grid access.

Liquid vs Air Cooling for 24/7 Inference Rigs

Comparing liquid and air cooling for continuous AI inference systems, focusing on reliability, cost, and long-term performance for unattended operation.

The clause. How a contractual definition of AGI met the capital built on top of it.

OpenAI’s contractual AGI clause was systematically defused from a doomsday trigger to a verification step amid capital restructuring efforts in 2025-2026.