What Is a Platform Operating Model and Why Does It Matter?

Platforms don't fail because of the technology. They fail because we run them like just any old tool.

A platform is a shared foundation designed for reuse across teams, divisions, and use cases, and that distinction changes everything about how you need to run it.

Tools like Power BI, ServiceNow, or a reusable data platform aren't single-purpose applications built for one team's workflow, not when they're successful. They become shared infrastructure, and shared infrastructure needs coordination that normal products just simply don't.

At Adair-Major Consulting, we see a platform operating model as the structure that makes a platform actually work at scale. It must cover things like:

• Governance, so there are clear standards around how the platform is adopted and reused

• Teams whose job it is to treat the platform as a product, not a one-off project

• Ways of working that let teams self-serve rather than queuing up for a central team

• Funding that works cross-region and cross-division, but still has clarity to measure value

Without an essential structure like this, platforms fragment. But the magic happens when you do have them, because then they compound in value.

Why Platforms Can't Run Like Normal Products

Most organisations still run technology through project or function-based models, where a team gets funding, builds something, ships it, and moves on. That can work for single-purpose applications with a single dedicated flow of value (though even there, the question of project teams versus enduring "build it, run it" models is worth a whole separate conversation). But it gets much more complicated for platforms, because platforms serve multiple teams and use cases simultaneously, and when done right the value is open-ended with outcomes that aren't always as clear-cut as who pays for what.

You know you have a problem when everyone optimises for their own local outcome and nobody owns the shared foundation. That's when you get the same capabilities rebuilt in different corners of the business, inconsistent patterns, brittle integrations, and what folks end up calling "spaghetti architecture." McKinsey's 2023 analysis on IT fragmentation backs this up, finding that standardising shared platform capabilities can cut duplication by 15 to 20 percent and boost delivery speed two to four times through proper API and microservices governance. A platform's operating model changes that by explicitly assigning ownership for shared capabilities (data, identity, workflows, whatever the platform provides), measuring reuse, and making sure teams contribute to a coherent system rather than just solving their own problem and walking away.

Self-Service With Guardrails, Not Gatekeeping

Platforms only deliver value at scale when people can actually use them without submitting a ticket and waiting three weeks for someone in a central team to get around to it, but "just let everyone do what they want" isn't the answer either, because that's how you end up with 47 versions of the same dashboard, no naming conventions, and you're wasting time and money rebuilding what already exists.

The goal of platform governance is to let teams move fast inside boundaries that keep things compliant, cost-aware, and consistent. A good platform operating model bakes this into the platform itself through standard interfaces, automated controls, and clear policies that are enforced by the system rather than by committee. This is where I see organisations struggle the most, because they either over-govern (and nobody adopts the platform because it's too painful) or under-govern (and adoption happens but quality, security, and cost spiral out). Getting that balance right isn't a one-time decision, it's an ongoing calibration, and it needs someone whose job it is to keep calibrating.

Funding and Platform Team Structure That Actually Sticks

Here's the structural problem: traditional project funding creates short-lived teams optimised for short-term, local outcomes, so the project finishes, the team disbands, and the platform is left without anyone responsible for its health, evolution, or reuse.

A good operating model for a platform creates stable, cross-functional platform teams with shared roadmaps and funding tied to enterprise-wide capabilities rather than individual project timelines, which is what makes reuse seem like the right thing, rather than a nice idea that never happens. When this structure exists, something interesting happens: each team that adopts the platform makes it better for everyone who comes after, because standards get refined, patterns emerge, and the platform compounds in value rather than fragmenting.

One academic paper refers to this as "open-ended value due to recombination," and I just love that idea! But you only get it when reuse is baked in culturally, made easy, and guardrails support it.

How to Tell If Your Platform Operating Model Is the Problem

If your organisation has invested in a platform (or several), and teams are rebuilding the same things, or nobody quite owns the shared layer, the technology probably isn't the problem. The operating model is. It's the kind of thing I help organisations figure out, whether that's designing the model from scratch or diagnosing why the current one isn't working.

Digital Operations & Transformation External: Recombination in the open-ended value landscape of digital innovation