At Novartis I ran marketing operations for eleven markets across four regions. Americas, Europe, Asia, Oceania. One brand. One team. One framework. Most people who have done this at scale will tell you the hard part is localization. It is not. The hard part is orchestration.
At This Scale, the Operating Model Is the Launch Strategy
Eleven markets means Americas, EU, Asia (with China and Japan demanding separate commercial logic from every other country in the region), and Oceania. That is not one market with regional flavor. That is four operating environments with different regulatory climates, different payer and channel structures, different HCP prescribing behaviors, different patient advocacy ecosystems, and in some cases different therapeutic indications in the pipeline.
The commercial team includes HCPs, patients, payers, and KOLs. It looks materially different from one region to the next. What a KOL advisory board looks like in Japan does not look like what it looks like in Germany. What a patient engagement strategy requires in the EU does not translate directly to Oceania.
When I was running this, the planning question was not "how do we adapt the campaign?" It was "how do we build a model that eleven markets can actually run, and that still produces one brand?" Those are different problems. Most commercial teams plan for the first one. The ones that succeed build for the second.
Co-Creation Is Not a Workshop. It Is the Build.
The first element of a multi-market operating model is co-creation with high-focus markets. Not a workshop. Not a regional advisory call. The strategy and campaign architecture get built with lead markets, not handed to them.
The distinction matters. Most global commercial teams operate on a push-and-explain model. Global builds the strategy, produces the materials, then holds regional calls to explain what was built and field implementation questions. That model generates compliance and confusion in roughly equal measure.
Co-creation means the strategic logic is developed alongside the markets that have the highest complexity and the most commercial stakes. When they have shaped the architecture, they have ownership of it. Implementation friction drops. Deviation from core brand strategy drops.
Hand-Holding at Scale Requires a Product, Not a Process
The second and third elements are paired: deployment with genuine hand-holding, and the infrastructure that makes that hand-holding repeatable across eleven markets.
All eleven markets get the campaign. Not all eleven have the same internal commercial capacity. Some have dedicated brand teams. Some have a single marketing coordinator covering three therapeutic areas. The operating model has to work at both ends.
Hand-holding at scale only works if the hand-holding mechanism is itself a product. The answer I built was two artifacts: Campaign Implementation Guides and Train-the-Trainer. The Campaign Implementation Guide was not a PowerPoint summary of what the campaign was about. It was an operational document. Channel sequence. Approval workflow by country. Asset inventory with usage instructions. Timeline with decision gates. A market without a brand strategist could pick it up and run.
Train-the-Trainer gave regional commercial leads the capability to onboard new local teams without global holding every activation call. Without that transfer mechanism, deployment is just file transfer. Files arrive. Usage is uneven. Results are unpredictable.
Without both, "deployment" means you sent the materials.
Measurement Without a Common Spine Is Just Data
The fourth element is a Common Marketing Performance Framework.
Eleven markets measuring performance against eleven different metrics is not multi-market measurement. It is eleven individual markets doing their own reporting, aggregated into a deck that looks unified but produces no portfolio signal.
The problem is structural. If every market chose its own primary KPI, cross-market comparison tells you nothing. You cannot identify where the commercial model is working and where it is not. You cannot separate markets that look behind because of market dynamics from markets that look behind because of execution gaps.
If every market measures performance differently, the portfolio signal is noise. A common framework does not mean every market tracks identical outputs. It means every market tracks against the same commercial logic, with shared definitions, so the rollup actually means something.
When I built this, the goal was not uniformity. It was comparability. The regional commercial leads could walk into a portfolio review and actually interrogate performance across the full eleven-market picture. That matters when you are making resource allocation decisions.
Localization Is Not Translation
The fifth element is localization, and it lives inside the first one. Translation is the last ten percent. The first ninety is the co-creation work that produces a strategy local teams can actually represent without needing to redesign it on the ground.
When co-creation is done right, localization is the natural consequence. When it is not, what gets called localization is just translation, and translation breaks at the market level. The fifth element does not get its own workstream. It gets earned by how you build the first.
Self-Serve Is Not Abandonment. It Is What Scale Looks Like.
The sixth element is the self-serve library, and it is the one most global commercial teams underinvest in because it only makes sense after the spine exists.
Once the eleven-market operating model is running (co-creation in place, Campaign Implementation Guides distributed, Train-the-Trainer active, common measurement framework live), every other market globally has access to the full asset library without a deployment call from global.
Self-serve is not abandonment. Self-serve is what becomes possible once the eleven-market spine exists. The markets with lighter commercial infrastructure can operate from a fully built campaign system. Global spends its attention on strategic evolution, not onboarding repeats.
The teams that try to build self-serve before the spine is solid end up with an asset library nobody uses correctly. The sequence matters.
The six elements I described are the operating spine. What I have watched AI change is not the number of markets you can serve. It is the speed and fidelity with which every element of the spine runs.