What’s the difference between platform thinking, platform business models, a technology platform and a multi-sided platforM?

When most people say platform, they often mean a technology platform. Think of this as a better version of the pipes that make an existing business run, like an IT platform.

When we say platform thinking, we mean considering how your business can create or engage platforms that enable third-party value, like app stores, to realize network effects.

When we say platform business model, we mean a business model premised on reusable, extensible infrastructure and data. These fall into several types:

  • Technology platforms, which focus on reusable infrastructure (like Nvidia, the graphics card and machine-learning business)

  • IP businesses, which focus on reusable data, either in the form of content (like a media company) or data (like Gartner)

  • Network orchestration businesses, which focus on connecting third parties to each other and enabling their collaboration (like Apple’s App Store, which connects app developers and end users, or eBay, which connects buyers and sellers)

When we say multi-sided platform, we mean platforms as business and operating models. A multi-sided platform is a combination of all three elements of technology, IP and network orchestration business models. Multi-sided platforms usually have some elements of self-service digital offerings for their users in order to access the scale (and thus network effects) of digital technologies.

How is platform thinking different from platform business models?

Platform thinking recognizes the value of taking advantage of network effects through the participation in platforms. Platform-based business models build at least part of that platform as a core way to deliver value, rather than simply accessing others’ platforms.

How do businesses reconcile pipes-based business models with platform-based business models, if they have both?

  • Apps on an app store

  • Featured apps

  • Bringing extension modules into the core, or spinning core modules out to the edge

How do you measure progress towards network effects when building a multi-sided platform?



  • Producer side

  • Consumer side

  • Cocreators

Partners (apps)

Strategic partners/joint ventures (featured apps)


Growth Rate

Rate of growth of key networks nodes (members, etc.)

Rate of growth of engagement


Quality & Utility

Return on engagement (self-reported and inferred from analytics)

Network promoter score/Network peer invites

Number of useful “apps” or offerings

Number of successful peer-to-peer interactions

Number of successful peer-to-superpeer (expert) interactions

Capabilities & Resources

Readiness level (to be quantified) for digital services

Readiness level for self-service and peer-to-peer offerings & features

Agile pipeline

Data compliance, cybersecurity scoring, uptime, % of errors resolved


Percentage of business functions operating inside new infrastructure/microservices approach

Accessibility of resources

API utilization rate

Discoverability of resources on the network (can people find what they are looking for?)


On technical debt

In a services-based business, IT and other technologies serve to enable humans who create value. In other words, they are there to support analog businesses.

In a digital business that includes network orchestration, digital technologies are on a more even footing with humans creating value.

Features that seemed “nice to have” become essential; and if the business is creating or enabling high-scale digital products, the architecture of the technology must be robust and scalable. “Duct-tape” solutions that work for a business unlikely to achieve massive scale on a short timeline are no longer sufficient and in fact present huge risks to the business.

Many “analog” businesses have learned these lessons the hard way through cybersecurity breaches or disruption by small digital businesses with better, easier, more reliable digital user experiences.

How is value created in a network orchestration model?


Matchmaking is a critical element of how network orchestrators create value. For example, the Etsy platform connects artisans and craftspeople with people who wish to purchase or even commission their work.

The quality of the matchmaking (meaning the degree to which the match actually fulfills the needs of the parties) is an important measure of network quality.

Digital tools and thoughtful network architecture can increase both the quantity and quality of matches, as long as the mental model of why people wish to be matched aligns with the algorithms used, and those algorithms continue to learn in both supervised and unsupervised ways.

How is platform thinking different from platform business models?

Platform thinking recognizes the value of taking advantage of network effects through the participation in platforms. Platform-based business models build at least part of that platform as a core way to deliver value, rather than simply accessing others’ platforms.

An MSP gets more valuable to users the more they use it

Like a first smartphone, the MSP may begin as a simple tool used for a few specific functions. But over time the MSP gets to know the customers and can provide truly personalized service. As new partners and services are added to the platform over time, it evolves along with customer needs and new technologies.

An MSP is open to third parties who create customer value

Through the MSP’s shared infrastructure and data, partners can develop new products and tailor existing ones to provide customers with better services and experiences than any one company can provide on its own.

Because it is open to co-developing products and services with external partners through the platform, the MSP model transcends a simple “reseller platform” scenario.

An MSP has a community and marketplace

Community and marketplace are the elements of the MSP that make transactions between partners and customers seamless and scalable. They maintain the customers’ relationship to the platform host’s brand even when using third-party services and products.

Conventional platforms are actually more like pipelines

Infrastructure-only business platforms receive new customers at the “awareness” stage of the customer journey and capture information that makes it easier to move the customer through a pipeline. They basically make incremental businesses work behind the scenes.

Conventional platforms are cost centers

Infrastructure-only business platforms are cost centers which increase the efficiency of other revenue centers in the business. They do not directly create new opportunities for revenue or attract new customer segments on their own; they simply make the existing business model work better.

Conventional platforms are accessible ONLY by insiders

Infrastructure-only business platforms are mostly closed to partners and customers. This means that the host has to be responsible for creating and delivering all value to customers. Resellers and partners contribute effort and resources but do not create new offerings for the customers who pass through the host’s infrastructure or marketplace.

What questions should I consider when determining when and how to bring third-party apps or function in-house vs. spin out core apps/functions out to stand on their own?

  • Is it essential to deliver a core value proposition?

  • Is it essential to execute (reliably, profitably, at scale, etc.)?

  • Can we afford to do it ourselves?

  • How would doing it ourselves affect other areas of the business?

  • Do we have freedom to operate (patents, IP, etc.)?

  • If we did it ourselves, would it be as good, better or worse?

  • What network effect implications are there for in-house vs. partners?

  • What is the experience impact? Does working with partners make it clunky or high friction for users?

  • What are the brand implications? Will it simplify it, differentiate it, or have positive or negative network effects?

What is the process for testing ideas for offerings on a platform?

We recommend using Strategyzer’s Value Proposition Design framework. Their iterative testing process is outlined below and adapted for a multi-sided platform approach:

  1. Craft platform models.

  2. Extract hypotheses about value propositions, infrastructure and members of the platform’s community.

  3. Design tests (such as pilots or feasibility checks).

  4. Enter the learning loop:

    1. Build

    2. Measure

    3. Learn

  5. Capture learnings and next actions. Hypotheses either invalidated (which means the model needs to be adjusted), or validated (which helps progress towards business building).

  6. Measure progress towards:

    1. Customer discovery (is there a market?)

    2. Customer validation (is the market interested?)

    3. Customer creation (will people actually pay for our value proposition?)

    4. Business building (are we increasing the reliability, scale and efficiency of the offering/company supporting?)

Concept credit: Strategyzer/Value Proposition Design