Platform-as-a-service: it's here and it’s the future of analytics
X-as-a-service models have already hit plenty of industries, but it’s only now that they are mature enough to meet all the business intelligence and analytics needs of modern organizations. Along with the most forward-thinking IT service providers, we are now entering the era of platform-as-a-service for BI – and the future looks bright.
Why buy (and manage, upgrade and replace) when you can simply use?
For many businesses, the “traditional” approach to computing is all about installing software tools on servers that have specific amounts of space and power.
“Certainly, these servers are increasingly moving to the cloud, but virtual servers operate along the same lines – your IT team will have to spend plenty of time managing the software you’re hosting,” explains Pieter Hebben, analytics solution lead at delaware.
But just a few years ago, a scalable, flexible and agile way of implementing organizational IT began to emerge: platform-as-a-service, or PaaS. Pieter: “Instead of servers, select companies began purchasing services. So, instead of buying licenses for software A and B and installing and managing the tools on your on-premise or cloud servers, you simply buy access to that software according to the needs of your business.”
PaaS transcends the entire expensive, hassle-filled process of supplying your firm with hardware and operating systems, upgrading and replacing them, and keeping patches up to date. “Just think of PaaS from an IT asset management perspective: not only do businesses win time and money, they automatically eliminate a range of cybersecurity risks that come with hardware,” Pieter continues.
Every BI service is now covered
The concept of platform-as-a-service may sound simple, but it took a range of developments to make it possible within the context of business intelligence applications.
“Just one year ago, PaaS offerings specifically for analytics were just not robust, varied and customizable enough to be worth it,” Pieter states. “But now, PaaS is the de facto standard when starting a new business intelligence implementation project.”
The difference today is that every possible element of a BI system is available as separate components of a PaaS that your business can design. Trusted vendors now offer an increasing number of services, and they are getting better and easier to use.
“Full analytics architectures offering every service you need can be added to a custom package. Instead of buying more computing power to your entire server to accommodate the needs of a single tool, you can configure every puzzle piece independently. The cost savings add up quickly.”
Cheaper – but only when approached the right way
IT teams following the “traditional” approach spend a large portion of their time managing their hardware. A mindset shift is needed to move to PaaS, as effort will go mainly to scaling services up and down and ensuring that they are running at their optimal capacity.
“If businesses don’t spend time developing a well-considered plan for PaaS, they won’t be able to reap the benefits,” Pieter asserts. “IT teams have to monitor the use of purchased services and ensure that the performance they are buying matches perfectly with their needs. This can be automated by writing scripts and configurations that adjust service levels by the day or even by the hour.”
Still, this automation doesn’t mean that IT teams will have extra time on their hands. Pieter: “That’s definitely false – it's important to stay vigilant to remain agile and ensure that nothing you are buying is going to waste.”
For businesses of all shapes and sizes
Apart from this mindset shift that Pieter describes, moving from traditional to PaaS-driven business intelligence is simple.
“It’s a low-effort move,” he goes on to say. “Vendors like Microsoft and SAP ensure that the same code from older solutions can be easily modified for use in a PaaS setup. Interestingly, the incentives for PaaS are also huge for smaller companies with proportionally smaller IT departments – they don’t want to buy all that hardware. With PaaS, your team has less to manage, and everything to gain.”
When switching from server-based IT services to PaaS, Pieter recommends implementing a series of robust naming roles to ensure that all of the different components of the platform fit smoothly together. “You need an overarching approach to stave off chaos. It’s important to establish rules and best practices before you get started.”