Digital TransformationAdvanced8 min read

Operating Model Redesign

An operating model is the way work gets done — how you're organized, what you decide where, what you build vs buy, what your processes look like, and what tech and data underpin them. An Operating Model Redesign is the deliberate rewiring of those layers so the company can deliver the strategy it has actually committed to. The classic frame is the POTI model: People, Organization, Technology, Information — every digital transformation that fails has redesigned 1-2 of these and pretended the others would adjust on their own. The real test of an operating model isn't elegance; it's whether a frontline employee can describe how a customer request flows from intake to delivery without naming a single specific person. If they can't, you don't have an operating model — you have a hostage situation.

Also known asTarget Operating ModelTOM DesignOrg RedesignOperating Model TransformationFuture-State Operating Model

Challenge a friend Browse library

The Trap

The trap is starting with the org chart. Reorgs feel decisive, generate a press release, and accomplish almost nothing because boxes-and-lines are the LAST thing you redesign, not the first. Without redesigning decisions (what gets approved at what level), processes (how work flows across boxes), and data (what each box can actually see), the new structure will recreate the same handoff failures inside new department names within 12 months. Bain's 30-year reorg dataset shows ~70% of reorgs fail to deliver intended value — almost always because they redrew the boxes without changing the wiring underneath.

What to Do

Sequence the redesign in reverse: (1) Strategy: what 3-5 outcomes must this operating model deliver? (2) Decisions: build a decision rights matrix (RAPID or similar) for the top 20 recurring decisions. (3) Processes: map 3-5 critical end-to-end value streams across functions. (4) Data & Tech: what shared spine do those processes need? (5) Org structure: ONLY now, design the org around the value streams. (6) People & Skills: identify 5-10 critical roles that don't exist today. Run it as a 90-day design sprint with a small senior team — not a 12-month consulting engagement that ships a deck.

Formula

Operating Model Health = (Decision Speed × Process Throughput × Data Trust) ÷ Handoffs per Customer Journey

In Practice

Microsoft's operating model redesign under Satya Nadella (2014 onward) is the canonical example. The org-chart change (collapsing Windows + Devices + Services into 'One Microsoft') was the smallest piece. The bigger moves: shifting decision rights from product-group fiefdoms to customer-segment teams, rewiring the incentive system away from internal competition, killing stack-ranking, and rebuilding the data spine so a single customer view crossed Office, Azure, and Dynamics. Market cap went from ~$300B to >$3T in roughly a decade — a multiple of the value any individual product change generated.

Pro Tips

01
Ban the phrase 'matrix organization' from the design conversation. Every operating model is a matrix; the real question is which axis is dominant for which decision. Be explicit about that for each decision type — it's the only thing that prevents permanent ambiguity.
02
Write the decision rights document BEFORE drawing the org chart. If you can't describe who decides pricing, hiring, vendor selection, and roadmap on a single page, no org chart will save you.
03
Pilot the new operating model in one business unit for 90 days before rolling it out. The first 90 days will surface 30-40 unanticipated process breaks. Fix them in the pilot, not at scale.

Myth vs Reality

Myth

“A great operating model is a permanent fix”

Reality

Every operating model has a half-life. As the strategy, market, and tech change, the model that delivered for the last 5 years becomes the bottleneck for the next 5. Treat operating model design as a recurring discipline, not a one-time project.

Myth

“The right operating model is whatever Spotify/Amazon/Netflix uses”

Reality

The most expensive transformation pattern of the 2010s was 'copy Spotify squads.' Operating models are tightly coupled to strategy, talent density, and cultural norms. A model that lets Netflix pay top-of-market and fire freely doesn't survive a regulated bank. Steal principles, never structures.

Try it

Run the numbers.

Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.

🧪

Knowledge Check

A new CEO inherits a slow, siloed company and announces a reorg in week 4 that collapses 3 divisions into 1 and renames 200 roles. By month 18, performance has not improved. What's the most likely root cause?

Industry benchmarks

Is your number good?

Calibrate against real-world tiers. Use these ranges as targets — not absolutes.

Operating Model Redesign Outcomes (Bain meta-analysis)

Multi-year studies of 1,800+ enterprise reorgs

Value Created (top quartile)

> 20% productivity lift

Modest Improvement

5-20% productivity lift

Neutral

0-5% lift, organizational fatigue

Value Destroyed

Negative — talent loss + disruption

Source: https://www.bain.com/insights/the-org-of-the-future/

Real-world cases

Companies that lived this.

Verified narratives with the numbers that prove (or break) the concept.

🪟

Microsoft

2014-present

success

Satya Nadella inherited a fortress of competing fiefdoms (Windows, Office, Devices, Server) where teams optimized for internal share over customer outcomes. The visible change was the 'One Microsoft' org collapse, but the durable wins came from redesigning decision rights (segment leaders own customer P&L, not product GMs), incentives (away from forced ranking, toward shared customer goals), the data spine (a single customer ID across cloud, productivity, and devices), and processes (engineering moved to continuous delivery on a shared infrastructure). The org chart was the headline; the wiring was the work.

Market Cap 2014 → 2024

~$300B → >$3T

Cloud Revenue 2014 → 2024

~$0 to >$120B annually

Engineering Cadence Shift

3-year → continuous releases

Stack-Ranking Removal

2013-2014 (precondition for collaboration)

The biggest operating model wins come from redesigning decisions, incentives, data, and processes — not from drawing new boxes. Microsoft's reorg was modest by comparison; the underlying rewiring was the actual transformation.

Source ↗

⚡

GE

2001-2017

failure

GE under Jeff Immelt repeatedly redesigned the operating model — Six Sigma, 'GE Store,' 'GE Digital,' 'simplification' — without ever solving the underlying issue: the conglomerate structure required cross-business coordination that the operating model was actively designed to discourage (each business was a P&L island reporting up to corporate). New initiatives (Predix, FastWorks) were bolted onto an operating model that wasn't designed to deliver them. The eventual outcome was a forced break-up into three companies in 2021-2024 — an admission that no operating model could make the conglomerate work.

Market Cap Peak → 2018

~$600B (2000) → ~$80B (2018)

Major Operating Model Resets

5+ in 17 years

Predix Investment / Outcome

$4B+ / wound down

Final Resolution

Split into 3 companies (2021-2024)

An operating model can only be as good as the strategy it's designed to deliver. When the strategic premise (the value of the conglomerate) is broken, no amount of operating model redesign saves you. Sometimes the answer is structural — split or sell — not redesign.

Source ↗

Decision scenario

The 90-Day Operating Model Sprint

You're the new Chief Transformation Officer at a 12,000-person logistics company. The CEO has approved a $35M, 18-month operating model redesign. You have 90 days to set the design direction. Your team of 8 includes 2 ex-McKinsey consultants pushing for a comprehensive 'future-state operating model' deliverable, and 2 internal operators pushing for a single business unit pilot.

Budget Approved

$35M / 18 months

Headcount

12,000

Current Cross-Function Cycle Time

11 days avg

CEO Patience Window

~12 months for visible results

Team Composition

8 people, mixed consultant/operator

Decision 1

Day 1: The consultants want to spend the 90 days producing a 200-page 'Target Operating Model' document covering all 5 business units and 12 functions. The operators want to pick the highest-friction process in one BU and redesign it end-to-end as a working pilot. You can pick one path or attempt both.

Build the comprehensive Target Operating Model document first — leadership needs to see the full picture before committing to changesReveal

By day 75, you have a beautiful 218-slide deck. Leadership reviews it for 8 weeks. By month 6, you have an approved 'future state' but zero implementation. By month 12, the CEO is asking why nothing has visibly changed. The deck becomes a credibility liability — every executive who sees no results points back to it. By month 18, the project is rebranded and quietly shelved.

Documents Produced: 1 (218 slides)Visible Change at 12 Months: NoneCEO Confidence: High → Low

Pick the single highest-friction process (last-mile dispatch) in one business unit. Redesign decision rights, process flow, data access, and team structure for that one journey. Ship a working operating-model change in 90 days.Reveal

By day 90, dispatch cycle time in the pilot region drops from 11 days to 4 days. You've redesigned 1 of ~40 critical processes — but it's a working proof point. CEO authorizes the next 5 process redesigns in months 4-9. By month 12, you've redesigned the operating model around 6 critical journeys with measurable lifts averaging 50% cycle-time reduction. The 'comprehensive TOM' emerges as a byproduct, not a precondition.

Cycle Time (pilot): 11 days → 4 daysProcesses Redesigned at 12mo: 0 → 6CEO Confidence: High → Higher

Related concepts