Overview
McKinsey & Company is the oldest and one of the largest of the so-called Big Three strategy consulting firms (alongside Boston Consulting Group and Bain & Company). Founded in 1926 and headquartered in New York, the partnership operates from more than 130 offices in over 65 countries. Industry estimates place fiscal 2024 revenue at approximately US$16 billion and project growth toward roughly US$17.8 billion by end of 2025. Global Managing Partner Bob Sternfels has led the firm since 2021. McKinsey remains privately held as a partnership.
McKinsey delivers AI and machine learning work primarily through QuantumBlack, AI by McKinsey. QuantumBlack was founded in London in 2009 as a Formula 1 analytics specialist and acquired by McKinsey in 2015 with around 45 employees. As of 2025 the practice operates from more than 40 offices, supports hundreds of client deployments per year, and is the firm's centre of excellence for advanced analytics, machine learning engineering, and agentic AI. Internally, McKinsey reports that more than 75% of its workforce uses its proprietary Lilli generative AI platform monthly, with around 150 centrally developed agents and around 17,000 employee-built agents in production.
Buyers typically engage McKinsey for AI strategy, executive alignment, high-value use-case identification, and lighthouse build work. The firm is not a low-cost engineering shop and rarely competes on staff augmentation. For pure model build, MLOps, or platform engineering at scale, hyperscaler-aligned system integrators and specialist AI engineering firms are often more cost-effective alternatives.
Services Offered
- AI strategy, value identification, and use-case portfolio prioritisation
- Generative AI and agentic AI transformation programmes
- Machine learning model development and productionisation (QuantumBlack Labs)
- Data platform strategy, architecture, and analytics modernisation
- Responsible AI, model governance, and regulatory alignment (EU AI Act)
- Enterprise-wide AI rewiring and operating-model redesign
- AI talent strategy, capability building, and academy programmes
- Industry-specific AI for banking, insurance, pharma, and retail
- Open-source AI tooling (Kedro, Brix) and MLOps reference patterns
Typical Engagement
| Engagement Type | Model | Typical Range |
|---|---|---|
| AI strategy & use-case sprint | Fixed-fee project | $500K–$2M (6–10 weeks) |
| AI lighthouse build (single use case) | Outcome-aligned project | $2M–$8M (12–20 weeks) |
| Enterprise AI transformation programme | Multi-phase programme | $10M–$50M+ (12–36 months) |
| Responsible AI / governance review | Fixed-fee assessment | $400K–$1.5M (6–12 weeks) |
| QuantumBlack engineer (build phase) | Daily bill rate | $3,500–$8,500/day blended |
Pricing ranges verified May 2026 from reference checks with 11 enterprise buyers, public statements of work, and market intelligence from competing MBB-tier firms. McKinsey is rarely the lowest-priced bidder and does not typically discount on rate card.
Strengths
- Executive-grade strategic credibility — board and CEO-level access is materially higher than at most system integrators
- QuantumBlack engineering bench combines data scientists, ML engineers, and designers under a single delivery model
- Deep proprietary IP, including the Lilli generative AI platform and open-source Kedro ML pipeline framework
- Strong responsible AI capability aligned to the EU AI Act, NIST AI RMF, and ISO/IEC 42001
- Global research engine (McKinsey Global Institute, QuantumBlack research) produces frequently cited benchmarks
- Track record of moving from strategy into build via QuantumBlack rather than handing off to external integrators
Limitations
- Premium pricing — partner-level rates can exceed $8,000 per day and senior engagement managers exceed $5,000 per day
- Limited willingness to take on pure staff augmentation, MLOps run-the-business, or commodity engineering work
- Junior-heavy delivery teams once engagements move past the strategy phase, which buyers should account for in contract scoping
- Variable depth on hyperscaler-specific reference architectures compared with cloud-native AI specialists
- Reputation overhead — public scrutiny of McKinsey's broader practices (opioids, sanctioned-state work) remains a procurement risk for some boards