Wide adoption of serverless will be led by tomorrow’s giants, not today’s
One day soon, serverless will be seen as a primary way to develop new applications. In this article, Erez Berkner explains why it’ll take some time for the knowledge and expertise to accumulate.
The advantages of serverless are clear. IT operations are much easier when a company doesn’t have to worry about maintaining its own infrastructure. No more patching or working to keep server or container environments running smoothly. Significantly lower costs and the ability to gracefully scale up and down in response to dynamic use patterns.
One day soon. But not today. It’ll take some time for the knowledge and expertise to accumulate, so smart developers can learn to tell powerful serverless design patterns from enticing anti-patterns. But within a decade – maybe half that time – serverless will be a choice that every new app designer will at least consider.
Even then, it won’t be the only choice or even necessarily the default. Some application workflows will still make more sense to develop in containers or to just take as fully-managed SaaS solutions. However, as the limitations of serverless are solved and cloud providers produce new offerings, serverless will become a viable design choice for more and more companies beginning their digital journeys.
I say ‘beginning’ because serverless is easiest to adopt from a clean sheet. Today, we’re seeing two types of companies dive fully into serverless.
Who adopts serverless?
The first type is startups who have no technical debt and are able to design from scratch without being constrained by existing systems. Examples include retail platform Sqquid and the self-explanatory Vacation Tracker. These companies were able to dive head-first into serverless architecture, reducing their upfront costs and enabling them to scale gracefully.
The other sort of companies going all-in on serverless aren’t new; they’re traditional larger firms who undergo ‘digital transformation’ to software companies in addition to their core business. Legendary guitar company Fender adopted serverless to deliver its instructional content and manage Industrial IoT in its factories, and Roomba maker iRobot went through a “greenfield” re-architecture when it moved to the AWS cloud. Similar to startups, these companies weren’t constrained by tech debt and could benefit from the cost reductions and long-term scalability of serverless.
Many other established enterprises are using serverless somewhere in their stacks. Media company Hearst, for example, built a serverless data analytics pipeline. Another common use, such as at CapitalOne, is for data migration and processing. These companies aren’t adopting serverless for everything like the startups or greenfields; breaking up a legacy monolith system into event-driven microservices small enough for a serverless infrastructure can be complicated, expensive and scary. Like replacing a gas-guzzling car with a hybrid, it could take years to feel the benefits so it’s understandable that many companies are sticking with what works.
The future of serverless adoption will be found in the future of software companies, not the present.
SEE ALSO: “In a Serverless world, sophisticated orchestration tools aren’t just nice-to-have, they’re non-negotiable”
Today’s tech giants may never go serverless. But the startups of today are the tech giants of tomorrow. It is these companies that will be the digital pioneers of serverless, just as the current wave of tech giants like Netflix and Airbnb pioneered public cloud and microservices.
Tomorrow’s tech giants are today’s growing startups. It’s these companies that will drive demand for new ways to use serverless, and encourage new technology options that don’t currently exist. They will generate the knowledge, expertise and best practices that help other companies to develop and operate serverless applications and to overcome any technical challenges. They will remove the roadblocks that currently prevent pure serverless adoption and break down the artificial boundaries inside companies, merging business financial decisions and tech design choices into one.
These companies will expect their partners and vendors to have the same agile, scalable, low-touch approach. The desire for familiarity and flexibility, combined with their purchasing power, will drive serverless adoption further into the ecosystem that exists around them.
The developers of tomorrow’s tech giants will have the training, experience, and mindset of designing “serverless first”. It’s these software engineers who’ll go on to found the next generation of startups. Many of them will never have known a world in which serverless wasn’t the preferred design choice, and will be more familiar with serverless than any other architecture. They will have entered the promised land of serverless and they won’t want to go back.