Studying Platform Strategy at MIT

MIT Media Lab FoyerI recently traveled to Boston’s Massachusetts Institue of Technology (MIT) to attend the executive course, Platform Strategy: Building and Thriving in a Vibrant Ecosystem. The course presented The MIT / Sloan Business School’s latest research on marketplace platform strategy and design. With such platforms now constituting some of the world’s fastest growing, most profitable and disruptive organisations (think Google, Uber, AirBnB, eBay, etc), researchers, executives, managers, designers and entrepreneurs are scrambling to get across this space. From a Collabforge perspective, designing and delivering platforms – especially for collaboration, innovation and consultation – is a central part of our business. So this seemed like a great way to get an injection of the best thinking on the topic.

A natural starting point, and a key theme for the course, was to define what is and isn’t a platform. The more you look, the more everything seems to be one! Therefore, it becomes essential to come to some clear distinctions in order to focus the discourse, research and design. Some key distinctions presented (one or more may apply to qualify as a platform):

  1. There must be at least two sides – “functionality depends on third-party participation”, or, intermediation between (at least) two distinct groups of participants
  2. Inventory is typically not held by the platform or platform owners
  3. The focus is enabling a new high value interaction, as opposed managing delivery of a product, service or management of a value chain
  4. Provision of a technology others can build products and services on
  5. Participants will value it more, the more others use it (The economist term for this being, “demand-side increasing returns”, geek speak being, “network effects”
  6. Generates value by reducing search costs, matching seekers with their needs, or delivering both

Once you know you’re dealing with the specific platform breed, then you can move on to the elements of strategy and design specific to this type of enterprise. And after my MIT experience, I have a renewed appreciation for just how complex these elements can be (and this is after having been in the business of designing and delivering collaboration platforms for the last decade).

Underpinning these elements of strategy, I learned that platform research started for economists with face-to-face marketplaces (probably the first version of a platform). Platforms then continued to ratchet up in pace with technology – from bricks and mortar infrastructure (railroads, power and telecomms infrastructure etc), then with content delivery networks and other deep Internet infrastructure, and finally, with its “appification” and the emergence of the digital giants mentioned above. I have no doubt this is a story for the ages, with cryptocurrencies, blockchain and now distributed autonomous organisations continuing the extension into as yet uncharted territory.

There is a great deal I could talk about, so I will limit myself to a few of the key new concepts that I learned about while at MIT:

1. Tipping refers to attracting users and overcoming chicken and egg issues that plague the launch of all platforms, and specifically, the tipping points that indicate when network dynamics start to kick in. Strategies for tipping include seeding content, “faking it before you make it”, viral strategies, making a deal with an anchor or highly influential lead tenant, building lead user communities, creating stand-alone functionality for one “side” of the marketplace who once they are attracted, can be used as bait for the other side.

2. Coring is the consciously architected outcome of making your platform the core or centre of an existing or new ecosystem. This involves careful upfront design and strategy, as well as often hard-nosed tactics during deployment to ensure your role and position in the ecosystem. Another key part of coring is designing and managing/policing the interactions that are core to your platform and its ability to generate and deliver value.

3. Pricing is key to both tipping and coring, and is a nuanced and complex domain. Determining how you charge and who pays, who is subsidised and why and how to make sure your pricing model is sustainable (even if and especially if you’re a not-for-profit or community endeavour) will often make or break the platform. I learned the term for the clip that AirBnB or Uber takes per drive or stay is called a “rake”. I also learned that different sides of your platform will have different sensitivities to pricing, which need to be accounted for. I also like this nice rule – that you should consider participant-led pricing when quality and supply is uncertain, and when deliberation not speed is important. Lastly, and perhaps most importantly, your pricing model needs to be repeatable for the sustainability of returning customers, on all sides of the platform (e.g. both producers and consumers).

4. Excludability was a fascinating concept presented by the professors (both of whom were economists). It refers to a good or service that can be prevented access to, and thus charged for. Excludability in the context of platforms can be a guide for what you keep secret (think Google’s PageRank algorithm) versus given away for free (think the location of an Uber driver at any give moment). Sharing and open access is central to the value proposition for platforms, as bait to prospective participants, but also to third parties looking to build on your platform and thus helping establish your “core” position in the ecosystem.

5. Lean Startup doesn’t work was an interesting and even controversial topic. The point being that a successful platform is a finely balanced ecosystem of participants whose interactions revolve around a core value exchange. Getting the participants to this state of equilibrium relies on very careful upfront strategic design, followed by a tactically planned sequence of activities (i.e. tipping). So setting up a process whereby the user is at the centre, leading you to the value they will pay for (Lean Startup approach) may lead you away from your platform architecture, and / or, mean you’re constantly re-architecting with every pivot. This may be fine if you’re a community endeavour like Wikipedia was when it was establishing itself, but this will likely be a very different story if you only have so much time, money and reputation comprising your startup runway.

A last tidbit that stuck with me was the following equation:

The network benefits + stand alone benefits – price ≥ adoption decision

I’ll leave you to ponder this – feel free ask questions or pose solutions in comments below!

In closing, one thing of which I am even more appreciative since my visit and learning at MIT, is that the collaboration platform subspecies is even more complex than many of the more transactional versions (and by collaboration I mean platforms where the participants engage in cocreation). For example, when it comes to an Uber or AirBnB, the core interaction and value exchange is very clear, there’s no questions about who is getting what in exchange for what cost. However, it’s not so clear when it comes to examples like Wikipedia, or Collabforge’s mass collaborations focused on policy or strategy cocreation. The producers and consumers can often be the same person or group, their interactions can be much less structured, and there are often many more types of participants. Nevertheless, the understanding of more transactional platforms offer a considerable amount of inspiration and ideas for the design of more collaborative approaches.

Subscribe to our mailing list

* indicates required

4 Responses to Studying Platform Strategy at MIT

  1. Pete Thomas says:

    Hey Mark — interesting read and sounds like you got a lot out of a challenging course! Good process at DFAT by the sound of it

    • retro Mark Elliott says:

      Thanks Pete – it definitely was a challenging (in a good way), primarily in terms of understanding platform design from an economist perspective!

      And yeah, the DFAT ideas challenge as been a great project to be involved in – much learning and great folks to work with.


  2. Prachi says:

    Thanks for sharing your insights Mark.

    On your point “4.Provision of a technology others can build products and services on”.
    Should the relevance of these products and services be monitored to suit the initial intent/value of the platform (like for uber/ air bnb)?

    Or is it seen as a failure if this intent/value keeps evolving? If not, how does one measure the success of a platform on an ever evolving intent?

  3. retro Mark Elliott says:

    Great questions Prachi!

    What really stands out for me is the tensions with needing to have a strong vision for where the platform will go, and thus the types of partnerships and third party products that should be encouraged and supported (as well as actively discouraged). This vision needs to be strong in order to get to a new reality where players in the ecosystem are interacting in a specific way, in order to generate the value premised by your platform. So this is at odds in some respects with a “lean startup” approach to letting users and other ecosystem players lead you to where the value is for them.

    So in a sense, I think for a true, strong, and large scale platform play to be a success, the amount of evolution will be more limited than another types of product or service development that can benefit from more user-led type of approaches. To your last question – these dynamics then seem to limit the amount of variance in the evolution possible, but I wouldn’t see it as a failure if evolution did happen and that led to success!

    Does that answer your question?

Leave a Reply

Your email address will not be published. Required fields are marked *