By Harvepino / Shutterstock.com
This article aims to express what platform businesses are and how they intend to partner and collaborate with other platform businesses to deliver solutions to customers. It also deals with the limits of the current standardisation efforts, and how they can be counteracted by the deployment of artificial intelligence components in the system. It also proposes why artificial intelligence (AI), robotics and leveraging the established notion of intent-based ordering can accelerate the formation of platform ecosystems as weakly defined intents support both greater decoupling between partners and allow greater differentiation between platform providers.
What is a platform business?
Firstly, for those wishing to understand in-depth what a platform business model is, I would recommend reading “The Platform Revolution” by Parker, Van Alsteyne and Choudary.
In essence, a platform business delivers agility and growth by digitising the business innovation elements of an organisation. This enables the real-time adjustment of key pillars of a business such as pricing, infrastructure deployed and products offered. While the majority of platform businesses depend on a high degree of automation being present in the operating processes of the organisation, this is not always true. A notable exception to operational automation is Uber, where flexibility is achieved by outsourcing its mobility providers to a massive pool of partners instead of automation. Uber’s interest in autonomous driving cars reflects the automation gap in the mobility industry.
In either case, a platform business achieves its competitive edge by near real-time flexing of fundamental business factors.
What is a Platform Business ecosystem?
Most of the platform business examples quoted are those of organisations that have achieved remarkable growth by controlling all the aspects of a market. While this may be effective in some spaces, this is not a guaranteed success formula for all economic activity. There are many successful niche players out there, and displacing them is neither an economic possibility nor, often, regulatory permissible.
Nonetheless, the flexibility and effectiveness of the platform paradigm would serve these niche enterprises well. In this evolution, an organisation would be able to pivot between multiple partners in real-time to deliver an ever-better customer experience. This would probably play to greatest advantage where the partnership delivers non-core or commodity value to the customer. A simple example of this could be an online payment partner for a telco.
In this concept, we would see a mesh of platforms-operated service providers co-delivering value to a customer. In each case, the partnerships would be optimised for that transaction, and the partners would be linked to each other in real-time by interfaces that are interoperable. This net becomes the platform ecosystem, where no one organisation delivers all the components of value, but each partner in the system delivers what they can do most effectively.
The Bitcoin ecosystem is a significant platform ecosystem.
The landscape of platforms, and why size matters
I particularly like this graphic from Statista Market Outlook, it is their intellectual property but it conveys a very important concept very well. You can gain a deep understanding of the different dynamics of the platform business landscape if you contact them. 
Copyright: statistica
The point is that while the highest, and globally accessible platform opportunities have been taken, there is a massive – orders of size more massive – market of platform opportunities to be had where individual enterprises are limited either geographically or by sector, and we have yet to see the peak growth of this sector.
At the hyperscale end of this continuum, the Googles and Apples have created a monolithic ecosystem girdling the globe, often arbitraging between national jurisdictions to maximise their returns.
At the swordfish level, vertical players have taken on multiple, often regulated industries across many countries. These players, while they have established dominance within a target vertical, have been unable to achieve the same depth across multiple sectors.
That brings us to the swarm of piranha platforms, niche in terms of geography and sector, and the kind of strategies they would adopt to maximise their returns. This group includes highly regulated industries such as banking, telcos and utilities. In each of these licensed cases, an operator is defined by a sovereign issued licence, and possibly constrained to very specific geographical boundaries. In addition, each of these industries is transforming to meet these challenges via digital deployments, for instance fintech, 5G and VNF and smart grids. For these platforms, the best way to scale new revenue opportunities is to partner to either new verticals, or across national boundaries.
The key competitive advantage of each form of player
For the largest white shark players, sheer numbers engaged in the ecosystem network deliver the competitive advantage. Apart from Apple, most of these players have to be content with relatively slim portions of the total revenue flows, often wholly funded by indirect revenue such as advertising to sustain income. Apple’s income, while massively profitable, has seen growth constraints delivered by the presence of low-margin competitors in significant verticals such as smartphones.
The swordfish platform layer bases its competitive advantage on delivering exceptional customer experience within a specific vertical, often while relying on crowdsourcing operation capital to enable scaling. In this space revenue shares are larger, often pitched at 15% of the total flow for the platform, but the need to deeply understand the customer’s view of value has constrained most of these platforms to a relatively narrow home range of products.
The piranha platform group is much more diverse than the others, but clear groupings of core competitive advantage are visible. In the licenced industries such as banking, telcos and utilities, competition is limited by regulators controlling the number of competing licence holders. In this diverse space, organisations that adopt platform-to-platform partnering approaches to the market can access additional revenue streams, at varying levels of revenue share and capital investment to balance their offerings portfolio. In this platform ecosystem, there is a clear advantage to being able to efficiently establish and pivot off multiple partnerships to reach and curate an optimal product mix. In an online world, this mass partnering can be a major source of complexity, cost and delay if each partnership requires a unique technical interface infrastructure and solution. There is thus a competitive advantage to the enterprise that has the lowest technology cost of partnering.
Standards: Building a path for the Piranha Platform players
While the piranha platform set is incredibly diverse, the licenced platform subset is used to a collective or crowdsourced response to systemic challenges.
Some of the responses have been:
- ONAP – network level standardisation to allow fluid network exposure;
- Fiware – open API solutions, favoured in smart city and IoT market;
- Open Banking API’s – openly available API’s to enable banking transactions, such as payments, deposits, account management and account enquiries;
- TM Forum API’s – growing set of capabilities to enable product, service and customer management.
In each case, these standards are as open for use as possible, with each of the participating sets of organisations betting that the network effects of mass adoption can be attained if the cost, complexity and friction to adopt the interfaces is minimised. To this end the interfaces are instantiated as RESTful API’s across multiple open developer platforms such as Apogee, BlueMix, GitHub and the like. Furthermore, the interface specifications and code are crowdsourced, and the solutions are crowd tested in series of hackathons. Some of these standards also carry investment support, providing start-up financing to early adoptees.
These industry body centred efforts are positioning their members as the central players as potential enablers of ecosystems of multiple platform business operators. The logic is simple and powerful. The standard interfaces aim to reduce the friction of partner integration by maximising the potential that the organisation’s next partner will use the same interfaces as the previous one. With this re-use, massive partnering is envisaged, and single product specialists can collectively deliver solutions.
In this sense these standards deliver to machines what we humans developed in our natural languages.
The limit of standards
The challenge of partnering with another organisation to deliver to a customer is fraught with the challenges humans face when they attempt to discuss collaboration. If the context is well constrained, and all participants are trained and used to each other, the requestor knows in detail what to ask for to receive a known response. But, as anyone who has ever held responsibility for a group output can affirm, the words we use are not always interpreted as intended by the requestor. This effect is heightened if the team has not worked together before, or if participants have an imperfect agreement as to the language in use. While these things are solvable over time, we also all know that constant adaption to new circumstances will keep an element of instability in the system.
A digital ecosystem will experience the same forces and challenges. Vendors will have implemented different versions of the same standard, or may differentiate to deliver compelling offers. This means it is unlikely that a consuming platform will always have an accurate view of what information a providing platform needs to respond to a request. This inter platform dynamic places the complexity of ecosystem integration at orders of magnitude higher than ESB integration where both consumer and provider process are within the same governance boundary.
Of course, this challenge is not limited to B2B integration, and we can take lessons from other digital implementations.
The first challenge of providers needing to train consumers how to request services can also be automated in a B2B digital platform ecosystem. Chatbots solve an equivalent problem for humans who need to interact with a system capability occasionally.
A linear automated system can only accept a limited, predefined set of inputs. Where a human interacts regularly with the system, the human becomes trained over time to provide the specific system inputs to get the desired outcomes from the system. Most markets are bigger than the number of humans that are able or willing to be trained on the needs of a specific system, so most systems require a trained customer service representative to manage the untrained masses. The digital version of the trained customer service agent is the chatbot. Current chatbots like IBM’s Watson Conversation component are trained in natural language by ingesting the whole of Wikipedia. While I would suggest that a platform ecosystem chatbot would need to be trained in a set of machine calls instead, the same concept holds. We can deploy chatbots to interpret consumer interface requests automatically. An example of this capability would be a provider’s ability to recognise a requested service address within a message of variable structure.
The second challenge is that robots have solved the problem of managing system behaviour to meet provided intents. Every quadrotor toy flies to an operator provided intent such as “Go Left”. This contrasts with previous generations of radio controlled aircraft that would take instructions such as “Left elevator up” and the operator would have to counter the instruction once the desired result was achieved. The quadrotor robotically adjusts its local controls to counter effects such as wind, temperature and speed transitions to achieve the intended result without operator counteractions.
It thus stands to reason that any organisation wishing to participate and differentiate in a digital platform ecosystem as a producer will be wise to utilise robotics to translate consumer platform provided intents into internal explicit and exact requests that can be delivered by the provider operations systems. The alternative of exposing detailed request formats and options is either too expensive for the consuming partner to onboard, or will expose the provider’s IP and unique value proposition to competitors. An example of this type of intent resolution would be the ability to localise a service address provided in a range of formats.
A third ecosystem capability would be that a consumer platform would be well served with a capability to learn which formats work best with which providers. This would be applicable where the providing platform responds to the consumer platform’s request with a structured order. If the consumer platform has the chat and robot listening capabilities these can be used to continuously align the intent request formats to minimise order rework in the future.
In conclusion
If we are to see the rise of a dynamic B2B platform business ecosystem, we will see system behaviours that parallel human society patterns. The pressure to constantly evolve and differentiate platforms will drive a continuously shifting disconnect between consumer platforms’ capability to request services and provider platforms’ capability to respond to service requests.
Judicious deployment of cognitive chatbot and robot capabilities in the platform ecosystem message stream will deliver resilience across these ecosystems in much the same way as these technologies have digitised human system interactions in untrained human interaction centric processes.
I have yet to see this implemented, but think that this is a function of the simplicity of current integrations and this will become necessary once the numbers of participants in ecosystems soar.
What do you think?
Recent Comments