S12.1 · Business & Professional Services
Project-priced strategy advisory where agentic AI collapses the associate-research cost line that generates the fee.
Management and strategy consulting sells advice on strategy, operating model and performance through staffed, project-based engagements — almost nothing here is subscription. Sizing depends on where the line is drawn, and estimates vary by roughly a factor of four: core strategy work runs roughly $180-200B (Source Global Research, 2024), while vendor estimates that fold in digital and IT consulting stretch the category to $400B-$900B by the early 2030s. What matters more than the sizing is where the profit actually comes from. It comes from the associate and analyst layer — research, benchmarking and synthesis billed at a staffed day rate — and that layer is precisely what agentic AI attacks first.
Enterprise cost discipline held growth to low-single-digits in 2024-2025 (Source Global Research), whatever the 6-11% CAGR built into forecasts on the wider digital-consulting definition. North America takes ~40% of demand, Western Europe a large share of the rest, and delivery increasingly runs out of India regardless of where the client sits. At the large-account end the field is MBB, the Big Four's strategy arms and Accenture; below them, an enormous boutique and independent long tail.
Nearly every dollar is priced by project. Contracted recurring revenue is minimal — the weakest recurring profile of any segment in the sector — and that is the commercial problem, because premium multiples across business and professional services accrue overwhelmingly to contracted, compliance-linked work, not episodic advisory. In isolation the economics look fine: 20-30% margins, capital-light, no regulatory gating, talent supply the only real moat.
Expansion runs in two directions, and only one adds revenue. MBB firms building implementation capability push into delivery work traditionally owned by systems integrators — new dollars. The Big Four folding strategy into a "one firm" delivery model, bundled with audit-adjacent and IT-advisory capacity, defends share as standalone strategy work commoditizes; no new budget is opened.
The cost line agentic AI collapses — associate and analyst time on research, benchmarking and first-draft synthesis — is also the billing line. A client that can generate AI-produced output in hours, rather than commissioning a staffed team for weeks, breaks the billable-hour, per-head model that underwrites the entire pyramid; there is no indirect substitution effect softening the blow. The moat this segment has actually run on — information asymmetry between what a client could produce internally and what a research-trained team could deliver — is the moat AI erodes fastest.
Benchmarking and research-heavy work is where the damage already shows. Broader deconstruction of the staffing pyramid sits on a 2-5 year horizon.