Continuing from where I left last time, lets talk about ownership considerations for the Infrastructure layer. Infrastructure layer has three broad components – the actual hardware on which the application/s will be set up, the connectivity (with the client data centre & the process execution (BPO) delivery location/s), and the hosting services (data centers).
Hardware: In platforms, even though the service provider is not charging upfront fees to clients, the upfront investment is required to buy the necessary infrastructure on which the application will be set up. Can the service provider sub-contract the hardware provisioning, and pay usage charges to the sub-contractor? If hardware is sub-contracted, can the service provider ensure its scalability? For instance, if the clients’ transaction volumes scale up or down (acquisition, etc.), how will the sub-contracted vendor ensure flexibility to ramp-up or down the infrastructure size?
Connectivity: The treatment of this layer is quite un-ambiguous – the service provider has to ‘buy’ the bandwidth from telecom companies and provide the same to enable platform access to clients as well as the BPO team.
Hosting: Setting up a data centre (both primary & DR data centers) involves huge upfront investment (multi-fold as compared to hardware). Also, a single data center might not suffice for a service provider that has multiple platform clients around the world. Depending on the type of client data being stored in the platform, individual countries impose geographic constraints on where that data can reside geographically. For e.g. for a hire-to-retire business platform, the platform stores the employee data. Some countries have legislations that their citizen data can’t reside beyond their geography. So the service provider needs to set up multiple Primary & DR data centers across the world; in effect, multiplying the investment requirement. Should the service provider own these data centers or sub-contract the hosting services from a hosting service provider? Should ownership strategy be consistently employed for all data centers? Or, should the service provider own data centers in a few geographies with others being covered by tapping onto service provider clouds?
Key Considerations:
Do the clients care about who actually owns the hardware, connectivity and hosting services? Shouldn’t the clients’ primary concern be only on the functionalities & process delivered by the platform, leaving the details to be filled by the service provider? Infact, should the client even try to take a component-wise view of the platform in terms of ownership & service responsibility?
Ideally, the client should treat the platform as a ‘black box’ and disregard the ownership of the sub-components of the service being delivered.
One of the biggest benefits of platform offerings for the service provider is that by bundling a large no. of services together, it can make significantly more money as compared to providing these services on a piece-meal basis. So, if a service provider sub-contracts a large no. of these individual services, its revenue potential & profitability go down. Also, the governance costs (handling multiple sub-contractors with back-to-back SLAs being tracked and monitored) increase significantly. The other side-effect of sub-contracting is that compared to in-house provision, the sub-contracted services come at a higher price (margin of the sub-contractor gets added), effectively increasing the price of the overall platform.
So, for the service provider, it’s essentially a choice between making upfront investment and competitive pricing of its platforms.
Here as well, like the application layer, the service providers need to answer these questions based on their existing capabilities, strengths and depth of their pockets. A platform service provider, who has hosting services as one of its service offerings, and already has data centers across the world, ownership of hosting services is a no-brainer. However for others, ability to invest in setting up data centers plays an important role in this decision.
Essentially, like its true for any manufacturer, the components that go into a product, could either be manufactured in-house or can be sub-contracted to another supplier. The choice needs to be made based on various factors, such as surplus capacity available in-house, end product delivery timeline, cost of in-house operation, price charged by the sub-contractor, margin available on the final product, etc. Here too, the decision for individual service components needs to be based on similar criteria. But the ultimate question is – if the service provider decides to sub-contract most/ all of the individual service components, what role is the service provider playing? Do the clients actually get ‘bundled service’, or have to make do with multiple services being bundled together!
