The Internet infrastructure services sector has always been a capital-intensive business. Hosters and service providers have built data centres at great expense. Many have moved to a leasing model but still have heavy CapEx requirements. They have to buy the racks, networking gear and servers that data and applications are run and stored on. There is also a shelf life for these investments that must be depreciated over time and then replaced. Planning future capacity can be a tricky proposition and strong growth can put strain on existing resources.
The world of infrastructure services – in its formative years – has been like this because few if any providers were able to achieve meaningful levels of scale. That began to change over the last several years. Massive-scale platforms built by Amazon, Microsoft and Google – running cloud infrastructure and SaaS – have been circling the sector for some time and have quickly reached critical mass. They now possess operating scale and efficiency that will be increasingly difficult to match – save for a select few – over the long-term. It is a story largely built on economics. But scale is not just something that translates into lower prices. It is about accelerating time-to-market, having incomparable geographic scope and reach, and building a critical mass of customers and ecosystem partners. It is also about unprecedented levels of innovation. Scale produces ample resources and it creates the repeated trials and range of use cases that provide unique insight. This drives a pace of innovation – at the infrastructure level – that small and even scaled providers won’t easily replicate.
The rise of massive-scale is creating a permanent shift within the competitive landscape. Already, we are seeing multi-tenant cloud infrastructure services operated by hosters and service providers struggle to gain traction as more workloads head for the massive-scale platforms. The shift is happening at an accelerating rate and it is already becoming apparent that hoster-operated multi-tenant clouds may end up being an endangered species sooner than later. There may yet be more species – not just hoster-operated public clouds – that become endangered.
If this scenario plays out as we think it will, the impact will be far-reaching. And nowhere will this be more evident than in the way service providers operate. Over time, providers will decrease the amount of infrastructure they own and operate – shifting CapEx to OpEx – and take advantage of the scale that massive-scale platforms provide. This will free up resources and simplify operating models.
Providers will take advantage and invest in technologies and capabilities that enable value to be built on top of raw infrastructure. Reselling cloud isn’t exactly a high-margin business and the need to build value will inevitably drive providers up-market.
This is the world we are headed towards. It is a world where the underlying infrastructure won’t matter nearly as much as before. The competition will shift from speeds and feeds to managed services, consulting and features. It will shift from price to value. Making this bet now is crucial as those already taking advantage of a more OpEx-oriented model are building up a lead and gaining competitive advantage.