Launchworks has developed tools and frameworks to help innovative digital businesses scale up and reach critical mass while maximising the value they add to participating communities.

To be successful, digital platforms need strategies for attracting, matching, connecting and enabling transactions between communities.

The Rocket Model for digital platforms by Launchworks & Co


In the 1980s, Michael Porter was one of the first to codify and formalise how firms create and capture value: by transforming inputs of production through a linear set of activities and processes. This traditional linear model of the firm is still the dominant frame of reference in business. Why? Because until recently, most established firms could be described as linear or pipe businesses.

However, Porter’s linear framework is not well adapted for multi-sided platforms where the process of adding value is shared between the firm and customers. It’s not only the transformation of inputs that matters but also the quality of matching, connection, and interaction of communities. Think about eBay connecting buyers and sellers or airbnb connecting hosts and guests.

The rocket model is a high-level functional model of platform businesses based on the core activities of firms serving multi-sided markets. Typically we find that platforms need to:

  • Attract critical mass of customers on each side of the market
  • Match them
  • Enable them to connect
  • Allow them to transact
  • Iteratively optimise their own processes.


Why a rocket?

Launching a multi-sided platform requires a lot of energy, in the same way as launching a rocket into space does. You need to recruit on at least two sides of the market (e.g. buyers and sellers), run development and marketing for each side etc. So in a way it’s similar to launching two companies at the same time. You also need to reach critical mass on both sides of your platform -a challenging hurdle that doesn’t exist for traditional businesses. On the plus side, once the rocket has reached critical mass, it requires less power to propel itself. It has reached escape velocity and is subject to less gravity. And it’s the same when launching platform businesses.

Let’s have a look at the Rocket Model and its associated functional stages:


Attract critical mass (on all sides)

This building block encompasses the characteristics, features and processes by which a platform is able to attract producers and consumers. In management literature, it is also referred as the ‘Magnet’[1] or the ‘Catalyst’[2]. At launch, The Attract Function is primarily focused on acquiring and hooking new customers but as the platform matures, retention starts to play a bigger part. In effect, as the platform matures it is often more appropriate to view this stage of the rocket as a ‘Net Attract Function’, e.g. the net number of active customers joining the platform (difference between new customers and leavers).



In order for both sides to interact, they need to be introduced first. For airbnb, it’s presenting guests with the right properties in the right locations at the right time. For Upwork, it’s matching companies with a specific assignment with available freelancers who have the right skillset. The quality of the matching is critical to the success of the platform. In a world of abundance, the ability to filter and present customers with the right choices creates value. For matching to be effective for participants, results must meet participant needs (relevance), be timely and present the right amount of information or depth (filtered). The latter means that the matching function should be optimised to return enough search results for the user to find a relevant match but not so many as to drown the user with information.



Often, platform participants need to exchange additional information with their counterparty before moving on to the transaction stage. If you think about dating platforms, questions about tastes, interests, etc. are a key part of the process. It is the same with eBay when buyers ask specific questions to sellers before making an acquisition (e.g. has the car had any accident before?). This platform function also increases the trust of the parties and reduces the ‘asymmetry of information’[3] that may get in the way of the transaction.



The transaction stage is at the heart of the platform value proposition. It represents the interaction that creates the most value for participants. A transaction can take many forms depending on the market being served. It could be a physical product (eBay, Alibaba), a rental contract (airbnb), a ride (Uber), a meeting (, a photo or message post (Instagram, facebook), etc. While many platforms crystallize some form of payment for each transaction, this is far from a rule.


Optimise interactions iteratively

This last optimisation stage is an absolutely critical process for continuous enhancement of the platform and is central to the data-driven nature of many of these businesses. Given the dynamic nature of platform businesses, this data driven function allows platform businesses to find the right balance between the two sides of the market and to optimise all the matching, connecting and transacting functions of the platform. Google’s Search algorithm is constantly optimised with several A/B tests[4] a day to ensure the best relevant search results. The concept of “big data” is core to most online platforms and continuous monitoring of potential bottlenecks can unlock growth in near real time.


Platform enablers

Finally a platform is supported by vertical enablers at each point of the rocket model. Here are a few that most platforms share:

  • Trust is what makes people believe that the participants they engage with are reliable, credible and honest. It’s a set of principles, rules, filters, processes and tools enabling participants to interact and transact in a safe environment. High trust encourages interactions and transactions.
  • User experience. It can be online-only for some platforms (facebook) or a mix of online and offline for others (airbnb). Online, the user experience is made of user journeys and touchpoints with the platform and participants. But unlike linear businesses which have control on the user experience end-to-end, a big part of the platform user experience, online and/or offine, is in fact delivered by participants themselves. Platforms have therefore limited control but can nonetheless influence positive outcomes over negative ones. Anyone can list their home on airbnb, but airbnb has the capactity to prioritise hosts with the highest feedback in search results.
  • Brand is also a key enabler which works in tandem with trust. The brand building of platforms is a slightly trickier exercise than for other business models since much of the experience of platform participants is directly influenced by other platform participants. Platforms therefore need to internalize the needs and wants of their communities and capture this in key brand attributes.
  • Payments are often key to the platform function and a critical step to enabling the transaction. A frictionless payment experience is therefore critical to the overall success of many platform businesses.

Together these enabling activities play a critical role in the success of the platform.


Laure Claire & Benoit Reillier are co-founders of Launchworks and help innovative businesses design and execute winning strategies.




[1] Sangeet Choudary, Platform Power, 2013,

[2] Evans David Evans and Richard Schmalensee, The Catalyst Code, May 2007

[3] One side, the seller, knows everything about their products while the buyer knows little. For the transaction to occur the platform needs to facilitate this exchange of information by enabling both parties to communicate. Reputation systems like eBay’s star system have been designed to increase the trust of potential buyers by enhancing the information to the buyer (previous buyer reviews).

[4] A/B testing is jargon for a randomised experiment with two variants, A and B, which are the control and treatment in the controlled experiment. An example is sending two slightly different promotion emails while tracking responses in order to quickly select the best of the two before iterating further. See Kohavi, Ron; Longbotham, Roger (2015). “Online Controlled Experiments and A/B Tests” for a discussion on best practices.


Share this page

Subscribe to our newsletter to keep up-to-date with the latest industry insights.