This Executive brief was generated using the AI Leadership Strategy tool. It gives you an example of the format and quality of the paid brief.
This is not a technology question. It is a question of position, business model, and where value accumulates in a decade that will restructure professional services.
McKinsey enters the AI decade as the most recognised name in global management consulting — a position built on eight decades of brand investment, alumni network density, and cross-industry pattern recognition that no competitor has yet replicated at scale. Its QuantumBlack AI unit gives it a credible AI capability story, and its proprietary databases of operational and financial benchmarks represent a genuine training advantage for AI models. But the pyramid that funds the firm — thousands of analysts and associates performing work that AI can now replicate — is structurally exposed.
A strategy built for one future is a liability. A strategy built for two is a bet. A strategy built across multiple possible futures — and calibrated against early signals — is what serious organisations build.
AI capability differentiation determines whether McKinsey can charge a premium for AI-augmented work. Client relationship model determines whether the firm captures value from outcomes or hours.
It's 2035. A global bank's CFO opens her Monday briefing. Alongside the usual market data sits a quarterly performance report from McKinsey — not a deck, but a live dashboard connected to the bank's core systems, benchmarked against 340 comparable institutions globally, with three AI-generated strategic recommendations ranked by implementation confidence.
This is what McKinsey's Intelligence Platform looks like at scale. Built on QuantumBlack's proprietary models, trained on two decades of cross-industry engagement data that no competitor can access, it has become the operating system for strategic decision-making at 60 of the world's largest financial institutions. The fee is not hourly. It is a retainer — access to the platform, quarterly partner reviews, and priority access to McKinsey's senior practitioners when the AI flags something that requires human judgement.
The transition was not painless. The analyst class of 2027 was half the size of 2024. The associates who survived were those who could interrogate AI outputs rather than produce them. Partners who could not articulate what they added beyond the model left, or were managed out. The firm became smaller, more senior, and significantly more profitable per capita.
What McKinsey got right was understanding that the scarcest resource in an AI world is not analytical capability — it is the judgement to know when the model is wrong, the relationship to deliver uncomfortable conclusions, and the institutional authority to make recommendations that organisations will actually act on. None of those things can be automated. All of them are more valuable when the analytical scaffolding around them is better than anything a client can build internally.
- Palantir and other enterprise AI platforms that are moving into strategic advisory — they have superior AI infrastructure and are adding consulting talent to close the judgement gap
- Big Four firms that have invested heavily in AI capability and can undercut McKinsey on price while offering integrated audit, tax, and strategy relationships
- Former McKinsey partners who have left to build AI-native boutiques with lower overhead, faster iteration, and credible pedigree
- McKinsey's QuantumBlack wins a competitive pitch against a hyperscaler's advisory arm on an AI strategy mandate — signals proprietary capability is becoming commercially decisive
- A McKinsey client publicly attributes a measurable business outcome to a McKinsey AI recommendation — creates the proof-of-concept the market needs to pay outcome-linked fees
- McKinsey's lateral hire profile shifts toward AI researchers, data scientists, and engineers — the composition of new partners changes visibly
- A major competitor (BCG, Bain, or a Big Four) announces a strategic AI partnership that materially closes the capability gap — triggers McKinsey to accelerate proprietary investment
It's 2035. McKinsey's Strategy Intelligence Platform has 1,200 corporate subscribers. Each pays a base retainer for access to AI-generated industry intelligence, competitive benchmarking, and regulatory scenario modelling. Larger subscribers add on-demand partner access. The smallest are mid-market companies that could never have afforded a McKinsey engagement in 2026.
The volume model was not the original plan. It emerged from a 2028 experiment — a self-serve AI diagnostic tool launched as a marketing initiative that unexpectedly converted 340 paying clients in its first year. The insight was that the addressable market for McKinsey-quality strategic intelligence was far larger than the addressable market for McKinsey-staffed project engagements. Most companies that needed strategic thinking could not afford $500,000 minimum engagements. They could afford $50,000 annual subscriptions.
The platform model has changed what McKinsey looks like internally. Partners have become product managers — their job is to encode their expertise into the platform's models, validate AI outputs, and handle escalations when the system surfaces something the algorithm cannot resolve. The firm has more revenue than 2026, with a third fewer professional staff. Margin per employee has tripled.
The risk in this scenario is that the platform model attracts competition that the project model did not. When McKinsey was selling bespoke advice to Fortune 500 CEOs, no one could compete at that relationship level. When it sells subscriptions to mid-market CFOs, it competes with every enterprise SaaS company that decides to add a strategic intelligence layer. The moat is proprietary data and brand — and both erode if the platform stops innovating.
- Enterprise SaaS companies (Salesforce, SAP, Oracle) that add AI strategy layers to existing enterprise software relationships — they have distribution McKinsey cannot match
- Bloomberg and other data intelligence platforms that add AI-generated strategic analysis to existing financial data subscriptions
- McKinsey alumni who build vertical AI advisory platforms for specific industries, undercutting the general platform on depth while the general platform undercuts them on breadth
- McKinsey's self-serve diagnostic tool exceeds 200 paying subscribers in its first six months — proof of demand at accessible price points
- A mid-market company publicly attributes a strategic decision to McKinsey's platform rather than a McKinsey partner — signals the brand extends to the product layer
- A major tech company acquires a management consulting boutique to accelerate its advisory ambitions — triggers McKinsey to consider whether the platform needs a technology partnership to defend its position
- McKinsey's analyst intake drops below 1,000 globally for the first time — a structural signal that the pyramid is unwinding faster than planned
It's 2035. A procurement team at a European industrial conglomerate runs McKinsey's proposal through its AI evaluation tool. The tool flags that three deliverables in the scope — market sizing, competitive benchmarking, and scenario analysis — can be produced internally using the company's AI platform in approximately 40 hours of staff time. The McKinsey proposal prices those deliverables at $800,000.
The CFO sends the proposal back with a counter: McKinsey can have the engagement if it scopes out the analytical deliverables and focuses exclusively on facilitation, stakeholder alignment, and executive recommendation. The fee is $200,000. McKinsey accepts.
This is the Commodity Adviser scenario — not a crisis, but a slow compression. McKinsey still wins engagements. Its partners are still respected. Its brand still opens doors. But the analytical layer that once justified premium fees has been stripped out by AI tools that clients now own and operate internally. What remains is the relationship, the facilitation, and the recommendation — valuable, but priced as a fraction of the full-service engagement it replaced.
The firms that fare worst in this scenario are those that waited too long to reposition. McKinsey's 2025 and 2026 leadership invested heavily in defending the analytical model rather than rebuilding around the advisory model. By the time the compression became undeniable, the institutional habits — the staffing models, the training programmes, the performance metrics — were all calibrated for a world that no longer existed.
- The client's own AI-enabled strategy team — the most dangerous competitor because it has context, continuity, and zero marginal cost for analytical work
- Specialist boutiques that focus exclusively on the facilitation and recommendation work McKinsey is being reduced to, but with lower overhead and faster decision-making
- Deloitte and Accenture, which have restructured earlier around integrated AI-implementation relationships and have more defensible client lock-in
- A McKinsey client publicly declines to renew a long-term advisory relationship citing AI capability built internally — the first such case becomes widely known in the market
- McKinsey's average engagement size falls below $500,000 for the first time — a structural shift visible in annual revenue per partner metrics
- A major client issues an RFP that explicitly excludes analytical deliverables from scope and prices advisory facilitation separately — signals procurement sophistication has reached the boardroom
- McKinsey's campus recruiting yield at top MBA programmes falls for two consecutive years — early signal that talent is redirecting toward AI-native firms
It's 2035. McKinsey's contract with a global retailer reads differently from any engagement the firm would have signed in 2026. There is no day rate. There is no deliverable schedule. There is a baseline performance measurement, an agreed outcome metric — inventory optimisation, measured in working capital reduction — and a fee structure that pays McKinsey 8% of the documented value created over 36 months.
The Outcome Engine scenario arrived not through McKinsey's strategic planning but through client demand. In 2028, three separate Fortune 100 clients asked McKinsey to share risk on transformation programmes. The first McKinsey refused, citing its independence model. The second it accepted, experimentally. The third became a case study: $340 million in documented value, $27 million in McKinsey fees — less than a traditional engagement would have billed, but more than McKinsey had ever earned on a single client relationship in the firm's history.
The model required McKinsey to build capabilities it had never needed before. Outcome measurement infrastructure. Proprietary AI tools to track leading indicators. Legal frameworks for performance fee structures across 60 jurisdictions. A new career track for implementation-focused consultants who stayed on client site for two years rather than six months. The firm that emerged looks less like a consulting firm and more like a sophisticated principal investor — patient, outcome-focused, and deeply embedded in client operations.
The tensions in this scenario are real. Outcome-based relationships require McKinsey to make judgements that have financial consequences for both parties. The independence that has always been the firm's ethical foundation is harder to maintain when McKinsey's fee depends on a particular strategic choice succeeding. Managing that tension — being genuinely independent in analysis while being financially exposed to the outcome — is the defining leadership challenge of the Outcome Engine model.
- Private equity operating partners who are structured for exactly this model — patient capital, outcome accountability, deep operational involvement — and who are increasingly offering advisory services alongside investment
- Big Four firms that have built implementation arms capable of delivering the integrated advisory-plus-execution model that outcome fees require
- Management consulting firms from emerging markets (particularly India and China) that have built implementation-first models and are now moving upstream into strategy
- McKinsey announces a formal outcome-linked engagement structure — the press release marks the moment the firm publicly commits to the model shift
- A McKinsey performance-fee engagement generates more revenue than a comparable traditional engagement in the same client — provides the internal business case for scaling the model
- McKinsey begins hiring professionals with operational P&L experience rather than exclusively MBAs and PhDs — visible in LinkedIn data before any formal announcement
- A major competitor publicly rejects outcome-linked fees on independence grounds — McKinsey's decision to proceed becomes a market-differentiating position
Across all four scenarios, McKinsey's fee premium is under structural pressure — the question is not whether it compresses but how the firm repositions before compression becomes irreversible. The scenarios where McKinsey sustains leadership (Intelligence Premium, Outcome Engine) both require the firm to build something genuinely new: either a proprietary AI capability that creates a data moat no competitor can replicate, or an outcome-accountability model that transforms the client relationship from advisory to partnership. Neither transition is achievable without cannibalising the pyramid that currently funds the firm. The scenarios where McKinsey struggles (Commodity Adviser, Platform Dominance risks) are those where the firm optimises the current model rather than reinventing it — and discovers that optimisation is not enough when the structural economics of the industry have shifted.
The goal is not to choose one model and commit to it. The goal is to build an organisation capable of operating across multiple futures — one that can learn from early signals and shift before the window closes.
- Accelerate QuantumBlack's proprietary model development now — the window to build a training data advantage from McKinsey's engagement history is closing as clients become more protective of the data they share. The proprietary database must be treated as a strategic asset, with investment commensurate to its long-term value, not as an IT project.
- Restructure the associate talent model before the market forces the decision — voluntary restructuring is more controlled, more humane, and produces better outcomes than reactive downsizing. Define what an AI-augmented McKinsey professional looks like and recruit and train toward that profile deliberately.
- Pilot outcome-linked fee structures on three major engagements in 2025-2026 — not as a revenue experiment but as an institutional learning exercise. The legal frameworks, measurement infrastructure, and cultural adaptation required cannot be built at speed in response to client demand. They must be built in advance.
- Protect the proprietary benchmark database as a non-negotiable strategic asset — do not license it, do not share it in partnership arrangements that dilute McKinsey's exclusive access, and invest in expanding it systematically across every engagement. This database is the training advantage that separates McKinsey from any AI competitor that does not have it.
- Platform versus project revenue mix — the appropriate balance between subscription intelligence products and bespoke project engagements depends on how quickly clients develop internal AI capability. Stay responsive to client behaviour rather than committing to a ratio in advance.
- Geographic expansion of the outcome model — outcome-linked fees work differently across legal jurisdictions and cultural contexts. Pilot in the most permissive markets first and adapt the model before scaling globally.
- Partner-to-staff ratio — the right pyramid structure for an AI-augmented firm is genuinely uncertain and will emerge from practice rather than planning. Avoid locking in a ratio before the work has revealed what the optimal structure is.
- Alliance versus acquisition strategy for AI capability — whether McKinsey builds, buys, or partners its way to AI advantage depends on what comes to market and at what price. Maintain strategic flexibility rather than committing to a single path.
These are not strategic options to weigh. They are decisions that become harder, more expensive, or less reversible with every quarter of delay.
Not rhetorical. These are the questions a leadership team needs to argue about — specifically, uncomfortably, without deferring to the strategy deck.
Generate your own Executive Brief at sofusmidtgaard.dk/ai-leadership-strategy-brief — tailored to your organisation, your sector, and your diagnostic answers.
This brief was generated using the AI 2035 Strategic Brief Generator — part of the AI 2035 Leadership Package. It is a starting point for strategic conversation — not a substitute for it.
The full process can be run with your leadership team.
Not a better version of this document — a different kind of outcome entirely.
When a leadership team builds scenarios together — arguing about which uncertainties matter, naming the worlds, stress-testing their assumptions — the result is not a brief. It is alignment. A room full of people who built the same picture and cannot unsee it.
That kind of buy-in cannot be delivered. It has to be created.
sofusmidtgaard.dk
This Executive brief was generated using the AI Leadership Strategy tool. It gives you an example of the format and quality of the paid brief. You will get a lot of content in your free brief, but unlocking all four scenarios and insight costs $149. You can generate your own free brief here