TubeReads

How McKinsey Plans to Survive AI (and Reinvent Consulting)

McKinsey is navigating a pivotal moment: technology that promises to automate much of what consultants traditionally sold is already reshaping the economics of advisory work. As generative AI commoditizes analysis and insight, the firm faces a fundamental question — what will clients actually pay for when they can do so much themselves? Meanwhile, the firm is still reckoning with reputational damage from high-profile missteps in client selection and oversight. Can a century-old partnership model reinvent itself fast enough to stay relevant in an agentic, outcomes-driven future?

Harvard Business Review2 People mentioned4 Glossary terms
Video length: 31:37·Published Feb 9, 2026·Video language: English
6–7 min read·5,154 spoken wordssummarized to 1,347 words (4x)·

1

Key Takeaways

1

McKinsey already employs 20,000 AI agents alongside 40,000 humans and expects to reach one agent per person within 18 months — a threshold originally projected for 2030.

2

The firm is migrating from fee-for-service advisory to outcomes-based partnerships, with roughly a third of revenues now tied to underwriting client results rather than delivering reports.

3

AI will commoditize much of what McKinsey historically charged for, forcing the firm to climb the value chain toward questions like «how do we double market cap» that require judgment, aspiration-setting, and discontinuous creative leaps.

4

The firm invested $1 billion to overhaul compliance and client vetting after missteps in opioid work and South Africa, bringing in leadership from Apple and Walmart to impose public-company-grade standards on a private partnership.

5

McKinsey overhauled talent selection to prioritize resilience, teamwork, and learning aptitude over perfect grades, discovering through internal data that candidates who had experienced setbacks were more likely to succeed long-term.

In a Nutshell

McKinsey is betting its future on a radical shift from selling advice to underwriting outcomes, building a hybrid workforce of humans and AI agents, and moving upmarket to tackle problems so complex that even superhuman tools can't solve them alone.


2

The Agentic Workforce Is Already Here

McKinsey now deploys 20,000 AI agents and expects one per human within 18 months.

Current AI Agents Deployed
20,000
Up from 3,000 agents just 18 months ago, now working alongside 40,000 human employees
Projected Timeline to One Agent per Human
18 months
Originally estimated to take until 2030; firm accelerated adoption dramatically
Annual Investment in Innovation
Over $1 billion
Funds proprietary IP development, research institutes, and new capabilities
Outcomes-Based Revenue Share
~33%
Roughly a third of total revenues now tied to underwriting client results rather than advisory fees

3

The Client Tension: CFO vs. CIO

CEOs face internal conflict between technology spending and measurable enterprise value from AI.

Do I listen to my CFO or my CIO? My CFO is in my ear that we're spending a lot of money on technology, but we're not yet seeing enterprise-level value from this. CIO's saying, are you crazy? This is one of those moments. And if we're not in the lead, we're going to get disrupted.

Bob Sternfels


4

Why AI Transformation Takes Longer Than Expected

Half or more of the challenge is organizational redesign, not technology.

Sternfels reports that CEOs worldwide are torn between enthusiasm for AI's potential and frustration with the pace of value realization. While technology leaders push for aggressive adoption to avoid disruption, finance chiefs question the return on massive AI investments. McKinsey's research points to a surprising culprit: organizational structure.

The firm finds that successful AI implementation requires fundamental rewiring — flatter hierarchies that eliminate middle management layers, consolidated workflows that break down departmental silos (like collapsing four or five mortgage process departments into one), and new operating models that can actually capture productivity gains. Banks, for example, have traditionally organized around process steps — origination, credit scoring, servicing — but AI-enabled workflows make those boundaries obsolete.

This organizational half of the equation is proving harder and slower than the technology half. Companies that focus only on deploying models without reimagining structure are the ones whose CFOs and CIOs remain at odds. Those that tackle both in parallel are beginning to see the enterprise-level value that justifies the investment.


5

What AI Can't Do (and What McKinsey Will Sell)

🎯
Aspiration & Leadership
AI models don't set goals or inspire organizations to stretch. Great leaders help define what's worth pursuing — a durable human skill in a post-AI world.
⚖️
Judgment & Truth
Models produce inference, not truth. Humans must impose parameters, context, and ethical boundaries that algorithms cannot self-generate.
💡
Discontinuous Creativity
AI excels at linear problem-solving but struggles with novel, non-obvious leaps. McKinsey is returning to liberal arts hiring to source discontinuous thinking.
📈
Market Cap Multiplication
«They're going to pay us to find ways to double their market cap» — problems so complex and high-stakes that superhuman tools alone aren't enough.

6

From Advisory to Outcomes: The Business Model Shift

McKinsey is moving away from fee-for-service advice toward underwriting client results.

OLD MODEL
Advisory & PowerPoint
Traditional consulting delivered insights, frameworks, and recommendations — but clients bore all implementation risk. If results didn't materialize, the failure was attributed to poor execution, not bad advice. This model separated thinking from accountability.
NEW MODEL
Outcomes & Co-Risk
McKinsey now co-creates business cases with clients and underwrites the outcomes, tying revenue to realized impact rather than delivered slides. The firm shares implementation risk and stays engaged until targets are hit. Roughly a third of revenues already operate this way; Sternfels aims for a majority by the end of his tenure.

7

Overhauling Talent Selection for Resilience

Internal data revealed McKinsey screened out the wrong candidates for 20 years.

1

The Discovery McKinsey analyzed 20 years of hiring data to identify skills predictive of partnership success. The firm discovered systematic bias: it over-indexed on perfect academic records and under-indexed on resilience, teamwork, and learning aptitude.

2

Resilience Over Perfection Candidates who had experienced setbacks and recovered proved more likely to succeed long-term than those with unblemished transcripts. The firm now actively screens for resilience in the application process.

3

Teamwork & Human Skills Real-world collaboration experience — team sports, retail jobs, group projects — became a priority. McKinsey is fundamentally about helping clients change, which requires human-to-human engagement skills.

4

Learning Aptitude in Novel Environments The firm now assesses how candidates perform in situations with zero pattern recognition — environments where no one has an advantage and pure learning ability matters most.

5

Opening 500 Pathways to Thousands McKinsey historically drew from only 500 elite academic and professional tracks worldwide. The new framework diversified sourcing to capture talent the old system systematically excluded.


8

Reckoning with Opioids and South Africa

High-profile missteps forced a $1 billion investment in compliance and client vetting.

Sternfels acknowledges that McKinsey's work advising opioid manufacturers and partnerships in South Africa were failures that demanded humility and structural change. The firm's traditional model gave individual partners broad autonomy to commit the enterprise, with minimal central oversight — a structure that scaled poorly as the firm grew and regulatory scrutiny intensified.

In response, McKinsey invested roughly $1 billion to bring in compliance and audit leaders from Apple and Walmart, building a client selection framework that evaluates country risk, topic sensitivity, institutional integrity, individual actors, and operating environment before engagement. Partners no longer commit the firm alone; they work alongside risk professionals. Sternfels framed the goal not as remediation but as setting the standard for professionalism in consulting.

At the same time, the firm pushed back on criticism it considered unfair — particularly attacks on its work with hard-to-abate sectors on climate transition. Sternfels argued that solving climate change without engaging high-emission industries is «naive,» and that McKinsey would not retreat from difficult work simply to avoid controversy. The dual posture — humble accountability for genuine failures, courage to defend legitimate but controversial choices — reflects the firm's attempt to navigate heightened institutional scrutiny without abandoning its mission.


9

What CEOs Are Obsessed With Right Now

🤖
Extracting AI Value
«How do I get value from this technology? And what is everyone else doing?» Leaders are eager but uncertain, caught between the promise of productivity and the reality of slow, hard organizational change.
🛡️
Institutional Resilience
In a world of continuous shocks, CEOs want enough buffer to withstand unforeseen hits while maintaining capacity for bold offensive bets. «Can I play offense and defense at the same time?»
🏢
Organization Model Frustration
No CEO thinks their org structure is perfect. Matrixed models born from 1959 HBR thinking are too slow, too cumbersome, and poor at reallocating resources or navigating complexity.

10

The 10-Year Aspiration

Sternfels wants McKinsey known as an impact partner, not an advice shop.

💡

The 10-Year Aspiration

In a decade, Sternfels hopes McKinsey is still the world's leadership factory — but no longer remembered for smart slide decks that may or may not get implemented. Instead, he envisions clients saying: «We designed a business case together. These guys underwrote the same outcomes I took to the board, and we kept at it until we got to someplace I didn't think I could get to.» The shift from advisor to impact partner is the existential bet.


11

People

Bob Sternfels
Global Managing Partner, McKinsey & Company
guest
Adi Ignatius
Editor-in-Chief, Harvard Business Review
host

Glossary
Agentic workforceA hybrid labor model combining human employees with AI agents — autonomous software tools that perform tasks traditionally done by people.
Outcomes-based modelA consulting engagement structure where the firm underwrites client results and ties revenue to realized impact, sharing implementation risk rather than charging fees for advice alone.
Hard-to-abate sectorsIndustries like steel, cement, and chemicals with high emissions that are difficult to decarbonize but essential to address for meaningful climate progress.
Institutional resilienceAn organization's capacity to withstand exogenous shocks while maintaining the flexibility to pursue bold strategic bets — playing offense and defense simultaneously.

Disclaimer: This is an AI-generated summary of a YouTube video for educational and reference purposes. It does not constitute investment, financial, or legal advice. Always verify information with original sources before making any decisions. TubeReads is not affiliated with the content creator.