Marsh: One Platform, Four Kinds of Expertise?
Download case PDF (4 pages)Request teacher’s pack
Classroom case | Strategy, professional services, and organization design
Information cutoff: October 4, 2026, 11:59 p.m. Eastern Time
This case presents a hypothetical next-cycle decision from public information. It does not describe an actual internal assignment or disclose any unreported program budget, deployment, adoption, or result.
The decision
In October 2025, Marsh McLennan announced a new Business and Client Services unit (BCS) to centralize investments in operational excellence, data, artificial intelligence, and analytics. The company also began a brand transition that took effect in January 2026. A hypothetical leadership team is now preparing its next cycle of operating priorities.
How should the team decide which capabilities BCS should own and scale across the firm, and which workflows should remain primarily within specialized businesses? A shared platform could improve consistency, reuse, and coordination. Centralizing client-facing workflows could also create friction where business needs, professional judgment, client relationships, and local requirements differ. Business-led implementation could preserve fit but duplicate tools and limit cross-business learning. A staged, federated approach could test selected workflows while building common foundations.
Assume the team may recommend priorities, decision rights, implementation sequence, and evidence gates. It is not deciding whether the company changes its legal name or whether its client businesses should merge. No public BCS-specific budget, rollout schedule, adoption data, client outcome, savings attribution, or return is disclosed in the sources available by the cutoff.
One company, distinct professional businesses
The registrant remains Marsh & McLennan Companies, Inc. The company uses the Marsh brand, and its NYSE ticker changed from MMC to MRSH effective January 14, 2026. The company said its legal name and CUSIP remained unchanged.
It reports two reportable segments. Risk and Insurance Services (RIS) includes Marsh Risk and Guy Carpenter. Marsh Risk brokers insurance and advises on risk, insurance programs, analytics, and alternative risk financing. Guy Carpenter advises on reinsurance, capital, actuarial work, and related analytics. Consulting includes Mercer and Marsh Management Consulting. Mercer works across health, wealth, career, investment, and related advice and services. Marsh Management Consulting provides management, strategic, economic, and brand consulting.
These four businesses are not four reportable segments, and they do not deliver identical work. Marsh Risk’s brokerage and risk-advisory activity, Guy Carpenter’s reinsurance and capital advice, Mercer’s people and investment offerings, and Marsh Management Consulting’s advisory engagements involve different client problems and workflows. BCS is a corporate operating-services unit; it is not a third reporting segment or a fifth client business.
The 2025 announcement described a staged brand transition. In January 2026, the corporate brand became Marsh; the business brands included Marsh Risk and Marsh Management Consulting, while Mercer and Guy Carpenter continued under their existing brands during transition. Do not infer that every legal subsidiary or operating name changed with the corporate brand.
What the company says BCS is meant to do
The October 2025 announcement said BCS would bring technology, data, and operations teams together under the Chief Information and Operations Officer. Its stated remit was to accelerate innovation and centralize investment in operational excellence, data, AI, and analytics. The company said it intended to build a technology and data ecosystem to enhance client outcomes and efficiency.
The Q2 2026 Form 10-Q described BCS within the broader Thrive program, launched in Q3 2025. The filing said Thrive focused on brand strategy, client value, growth, and efficiency, and that BCS was formed to accelerate innovation and centralize investment in operational excellence, data, AI, and analytics. It reported approximately $500 million of estimated program costs over three years, primarily severance, technology, and outside services, and approximately $400 million of annualized savings. The company expected savings and costs to be distributed over the program period and said its plans remained subject to refinement.
The filing reported $239 million of cumulative Thrive program costs through June 30, 2026, including $52 million in Q2 and $89 million in H1. Total restructuring costs were $58 million in Q2 and $103 million in H1; those broader figures are not the same as Thrive charges. The Q2 investor presentation described approximately $400 million in savings “over the next three years,” a different phrase from the filing’s annualized-savings estimate. Neither disclosure gives BCS-only costs or realized BCS savings.
The presentation also described a BCS and Oliver Wyman partnership with AWS to reimagine mid- and back-office processes using AI-driven automation and process redesign. It listed client-facing analytics offerings and described management’s view of AI and automation. These disclosures show an announced direction and partnership, not how many workflows or employees had adopted a system or what effect it had on service, efficiency, quality, or client outcomes.
Recent results show different business conditions
For Q2 2026, Marsh reported consolidated revenue of $7.404 billion, up 6% as reported and 5% on the company’s underlying-revenue measure. GAAP operating income was $1.899 billion, up 4%; adjusted operating income was $2.166 billion, up 5%. Adjusted measures and underlying revenue are company-defined non-GAAP measures and should be kept separate from GAAP results.
RIS revenue was $4.823 billion, up 4% as reported and 3% underlying. Marsh Risk revenue grew 6% as reported and 4% underlying, reflecting higher new business and renewal revenue, partly offset by declining insurance premium rates. Guy Carpenter revenue declined 2% as reported and 2% underlying; the company cited lower reinsurance rates. RIS also includes fiduciary-fund interest, which declined year over year as average interest rates fell.
Consulting revenue was $2.602 billion, up 10% as reported and 8% underlying. Mercer revenue grew 7% as reported and 5% underlying; Q2 underlying growth varied across Wealth, Health, and Career. Marsh Management Consulting revenue grew 15% as reported and 13% underlying. For the segment, the reconciliation reports a $25 million combined Acquisitions/Dispositions/Other Impact adjustment without isolating acquisitions. In H1, Mercer Career’s U.S./Canada project-related decline offset growth in products and services.
These different results do not measure BCS performance. Underlying revenue adjusts for foreign exchange and certain comparability items; acquisitions/dispositions and other items can also affect reported comparisons. New business, renewals, rates, project mix, capital markets, investment flows, interest income, acquisitions, foreign exchange, and restructuring may all contribute to observed results. Public results provide no clean untreated counterfactual for BCS.
Marsh’s July Global Insurance Market Index reported global commercial insurance renewal rates down an average 6% in Q2, after a 5% decline in Q1; property was down 12%, casualty up 2%, and the U.S. composite down 2%. This is Marsh’s proprietary, rounded index reflecting its client portfolio. It is market context, not an independent census or a measure of any one client’s premium, Marsh’s placement volume, or the effect of BCS.
The design choice
The team can consider several operating models:
- Common enterprise platform. Centralize shared data foundations, analytics and AI tools, technology standards, and selected operational workflows. Keep expert judgment and client relationships close to the businesses, with common measures across deployments.
- Business-led implementation. Let each business choose tools and workflow priorities, using BCS for limited infrastructure and coordination. This preserves specialized control but may lead to duplicated investment and less reuse.
- Federated, stage-gated adoption. Centralize basic platforms, security, data standards, and service support; test workflows separately in Marsh Risk, Guy Carpenter, Mercer, or Marsh Management Consulting. Expand only after evidence shows acceptable service quality, adoption, client safeguards, and value for that workflow.
These paths can be combined. A data foundation may be common even if client decisions or expert review stay local. A repeatable internal process may be a better initial test than a judgment-heavy client engagement. The case evidence does not establish which tasks should move first or which model would produce better results.
What evidence would help?
To decide where to centralize, the team would need more than company-wide savings estimates. Useful evidence would include:
- a BCS service catalog, business participation, decision rights, planned rollout, investment and cost allocation;
- workflow-level baselines for cycle time, quality, rework, uptime, user adoption, client experience, and service cost;
- task differences among insurance brokerage, reinsurance, health/wealth/career consulting, and management consulting;
- client consent, confidentiality, data residency, professional conflicts, privacy, cybersecurity, and human escalation controls;
- systems retired, transition costs, workforce training and retention, and costs of exceptions;
- savings actually realized against a baseline and any incremental revenue or retention linked to a specific workflow; and
- comparable pilots by business, geography, client type, and process before scaling a shared capability.
These are proposed data requests. The public record does not disclose them at the level needed to assess BCS outcomes.
Assignment
Prepare a recommendation for the hypothetical next-cycle leadership team. State:
- what BCS should centralize, what should remain business-led, and why;
- how the team should sequence the work across Marsh Risk, Guy Carpenter, Mercer, and Marsh Management Consulting;
- what measures and safeguards should govern pilots and scale-up; and
- what evidence would cause the team to expand, redesign, or pause a capability.
State assumptions. Attribute all company aims and estimates. Do not describe the brand transition as a legal rename, treat BCS as a reportable segment, infer a completed AI rollout, or claim savings, adoption, client outcomes, or returns not disclosed by the cutoff. No Q3 or later results may be used. Financial modeling and Excel are out of scope.
Selected sources
- Q2 2026 Form 10-Q
- FY2025 Form 10-K
- BCS and brand announcement
- Q2 2026 results release
- Research detail: integrated literature review