Jun 26, 2018
Your company needs a data analytics platform. Or, you’re considering alternatives to your current platform. Either way, you’re reading this blog because you want more insight from more data, and you want it faster, easier and with the highest value. Here’s the first question you must ask: On-premises or SaaS?
If you’re leaning toward an on-premises solution, first consider the following questions:
Does my organization have the expertise and capacity to run our own analytics platform?
Your IT team is no doubt highly qualified. But do they really have the experience necessary to deploy an on-site analytics platform from the ground up? In addition to understanding SQL databases, they’ll also need a strong background in data architecture.
Even if your team has the right mix of skills and knowledge, do they actually have the capacity to manage the intricacies of an on-premises solution? If so, is that really the best use of their time? What other higher-value activities could they pursue instead? Database and application administration, security and troubleshooting can prevent you from focusing on delivering advanced analytics that separate contenders from competitors.
How important is instant elasticity?
How rapidly does your company plan to scale its storage and compute requirements? Like most organizations, will you access the broad and deep insights that lay waiting in the exponential growth of data generated by your website activity, mobile apps and other log and machine-generated data? Today’s data requires organizations to fluctuate at a moment’s notice.
Adapting an on-premises analytics platform to your increasing data needs can be a struggle. Scaling storage and compute up, down and out (concurrency) brings all sorts of headaches with on-premises solutions and is nearly impossible to accomplish on the fly.
By contrast, a modern and instantly elastic SaaS solution can make fluctuations in compute and storage – by the hour, minute or second – standard practice and a viable opportunity for your business. Monday morning usage peaks and end-of-month reporting periods require greater data and processing power than most nights and weekends. Rather than investing in additional infrastructure as data needs change, the SaaS model provides scalability that aligns with the rhythm of your business. Pay only for the compute and storage power that you need, when you need it.
What is the true cost of deploying an on-premises solution?
On-premises systems also come with a number of upfront expenses. Obviously, there’s the initial cost of the hardware and software. Someone from your staff must then install, test and deploy the app. There’s also the opportunity cost associated with diverting your team from other pressing matters. And, if they’re unable to do the work themselves, you may need to bring on consultants to help achieve liftoff. A good data architect can be rather pricey these days.
You should also factor in the ongoing cost of support, maintenance and training. If and when things break, it will be your team’s responsibility to provide a fix.
Compare the on-premises model, which has significant upfront and ongoing financial considerations, to the pay-as-you-go model that’s native to a SaaS data solution. But don’t settle for standard subscription-based pricing. Save even more money with usage-based pricing. For example, our Snowflake Standard Edition starts at only $40/TB/mo for storage and $2.00/hour for compute and charged by the second. Check out our pricing here.
Are our data sources already in the cloud?
Stop and think about your data pipeline. Does most, if not all, of your data already reside in the cloud? Let’s assume a sizable percentage of your data rests in your CRM solution. You’ve been utilizing a cloud-based CRM for years, which means your customer, lead, opportunity and engagement data is already stored online. In this scenario, it makes little sense to pull that data from the the cloud and into an on-premises system – especially when you consider data has mass, which means it also has inertia and gravity. The longer data resides in a siloed system, the less feasible cross querying becomes.
A better approach would rely on a SaaS analytics tool built for the cloud. Your IT team could spend less time manually exporting and importing data files, and your users could extract the insights they need faster.
Leverage SaaS data analytics
SaaS data analytics platforms solve many of the headaches associated with on-premise solutions. If you’re ready to give SaaS data analytics a try, continue reading about Snowflake’s built-for-the-cloud data warehouse. Snowflake is easy to configure and use, offers a pay-by-the-second billing and can scale on the fly.