In this blog, we explore why so many companies are moving their data and analytics platforms to the cloud and why others are resisting. We provide insights on current cloud technology, along with our tips for migrating to the cloud that will minimize the risk of disrupting your business.
Simply stated, cloud computing companies handle components of IT infrastructure better than most organizations could do themselves. They are often better in terms of maintainability, scalability, security, and cost. In the same way that a retailer may outsource distribution of their products to a company with a core competency in distribution, many companies outsource some of their IT infrastructure to companies with a core competency in IT infrastructure.
It is difficult and costly to manage and maintain a server network. By outsourcing these tasks, you trade capital expenses (servers, data centers, etc.) for a variable expense and only pay for what you use. While data volumes continue to flux, cloud-computing platforms allow you to dynamically scale both storage and computing power to quickly meet your needs while paying only for the resources you need.
A simplistic definition of a Capital Expenditure (CapEx) is an expense incurred now to realize future benefits, while an Operating Expense (OpEx) is an expense required to support a company’s current day-to-day operations. These very basic definitions explain why companies’ Accounting departments typically prefer to realize computing expenses as OpEx to CapEx, but there are additional benefits.
CapEx requires assets to have utility beyond the tax year in which they were purchased, and only a fraction of expenses can be deducted each year – the amount of which is variable (and often confusing and controversial!) depending on the depreciation method(s) employed by a company’s Accounting department. Conversely, the entirety of an OpEx can be deducted immediately. This simplicity of accounting for OpEx is another reason subscription-based cloud services are attractive to companies’ Accounting departments.
No matter how sophisticated and cutting edge a piece of server hardware is today, it will be obsolete in the near future. By outsourcing server infrastructure to a cloud company whose raison d’etre is to provide the most attractive server hardware, companies can be assured they’re always paying for the most current and performant server environment available.
Security breaches are an everyday occurrence. Nefarious hackers snatch valuable data from companies, regardless of size, industry, and geography. Moving to the cloud can help mitigate your security risk. The leading cloud platforms offer robust security frameworks, maintained by a small army of the industry’s best security professionals armed with the latest training and tools to help keep hackers at bay. Some providers often take shared responsibility for any breaches.
Cloud providers are constantly releasing new features and functionality. Service integration, performance, and security continually improve so that building and managing your cloud environment becomes easier and more efficient. Moving to the cloud allows you to be at the forefront of new releases to get ahead of your competitors.
With so many cloud providers, methods, and models, it can be easy to be overwhelmed by the migration process. PaaS vs. IaaS vs. SaaS, and Public vs. Private vs. Hybrid are some of the many choices you will have to make, but you are not the first to do so. Companies of all sizes and complexity are successfully migrating to the cloud with well thought out strategies. You could consider getting assistance from a company who specializes in creating data strategies for customer initiatives, including cloud migration.
Moving your data management and analytics platforms to the cloud is a major project that can take months or even years, depending on project scope, company regulations, planning (or lack thereof), and much more. However, migrating to the cloud is a sound long-term investment that will eventually help you act fast and increase efficiency.
As mentioned above, most companies’ Accounting departments prefer to realize computing expenses as OpEx instead of CapEx, but there are situations where the opposite is preferred. Sometimes Accounting departments want to spread expenses across years via depreciation, sometimes the budgeting process favors CapEx to OpEx, or maybe the accountants want to “tell a story” on the balance sheet to investors who yearn for a reduction in day-to-day operating costs and investments in the future.
Cloud computing is designed to reduce many of the burdens of IT management, but specific types of skills are required to successfully manage cloud environments and fully leverage their power. If the cloud stays internal – developed, hosted, and managed by the company itself – they need to hire a specific skillset. For organizations who use third-party providers, they still need people who know what services to pick, can negotiate service level agreements, and can integrate off-site offerings with on-prem data.
If you’ve made the decision to move to the cloud, planning your move is just as important as the migration itself. To avoid a migration rife with delay, first gain a clear understanding of some following technical considerations.
When moving to the cloud, you need to choose your deployment model: Public, Private, or Hybrid.
As companies move to the cloud, CIOs are faced with an overwhelming number of options for technology providers and services. And to make it more confusing, CIOs often don’t just use one provider – many times they are using more than one cloud provider for their business. We see that the big cloud companies are still dominating, but they all have different strengths and priorities. Here is a breakdown of a few of the big cloud companies and why you may use each one:
Amazon and Microsoft currently dominate the cloud-computing space with their comprehensive, dependable cloud platforms, Amazon Web Services (AWS) and Microsoft Azure. These products include the following service model capabilities straight out of the box:
These storage and service options are priced on a usage basis, and many services can be turned on/off or scaled at will to save costs. Why pay for computing power or storage that you are not using?
Microsoft AWS and Microsoft Azure also offer numerous built-in tools for data profiling, data integration, data science, and data visualization. While these services are newer to their platforms, they continue to release enhancements and features to rival existing on-prem tools. One benefit of using the major cloud platforms is that they offer SaaS options that allow you to integrate with existing software. For example, Informatica, SSIS, and other data-integration tools are compatible with AWS and Azure; and data visualization tools such as Qlik, Tableau, and Power BI can tap into data stored in the cloud.
Google Cloud (GCP) is quickly becoming another viable alternative. The GCP suite grows daily (the reporting tool, Looker, is part of GCP as of June 2019), and it already offers the same base features and services as AWS and Azure. However, GCP uses the FASTER Cable System, which can allow for up to 10Tbps, compared to 100Gbps for AWS and 263Gbps for Azure. This means you have up to 100 times the network throughput, which is great for real-time data and high traffic web applications. With Live Migration, GCP will migrate your applications to new instances automatically so there’s no downtime.
Snowflake is not a full stack cloud platform; it a data warehouse built specifically for the cloud. It is able to offer instant scalability and availability and true automatic elasticity. Behind the scenes, it is built on top of AWS and Azure (and GCP in Q4 2019). If you’re already storing data in the cloud, it’s quite easy to migrate it into Snowflake. Once your data is loaded, you can then create a virtual warehouse (the computing power of Snowflake) with a few clicks and start analyzing your data.
While both Azure and AWS have their own proprietary data warehouses, Snowflake offers some key benefits that Azure Data Warehouse, AWS Redshift, and Google BigQuery haven’t offered yet:
If your organization owns sensitive data, Snowflake has levels for HIPAA support, PCI compliance, and even dedicated virtual servers.
Migrating to the cloud frees you to focus on analyzing your data and spend less time managing data storage. However, moving to the cloud can be a significant business decision and expense, so before you begin migrating, it’s important to think through the following points:
How does the cloud fit into your BI strategy? Do separate business units have their own data? What insights are you looking to gain, and how will the cloud help you reach those goals?
There can be a significant initial investment to set up your cloud environment. Does all your data need to be in the cloud, or can you start with a single source as a proof-of-concept? Is it easier to get OPEX or CAPEX approval to fund a cloud migration? Will you continue to fund cloud and analytics in the future?
What’s already in the cloud, and what’s on premise? How will you structure your data in the cloud, and how will different data sources be stored? How will your data be ingested into the cloud, and how will it be consumed from the cloud?
What components of your data and analytics will be in the cloud? Will it be hosted, managed, or SaaS? Are your users in the same part of the world, or do you need to consider several availability zones? Understanding your current situation will help you optimize your use of cloud storage and compute resources.
Who’s driving the effort and who will use it? How will your applications and data be managed? Is this driven by the business, or IT? The number of users (concurrency) and the compute resources they use need to be considered to fully leverage the benefits of a cloud migration.
Does your data come from a variety of sources? Does the data need significant grooming and transformation prior to end use? Is your data structured, semi-structured, or unstructured? What volume of data will you be storing in the cloud?
Data security concerns are more prevalent than ever. Do the cloud platforms meet your data security needs? Does your industry have specific security compliance? Will their SLAs satisfy your data availability needs? What are your backup strategies?
Especially in Hybrid environments, user accounts may be stored and managed in a variety of places. For example a user’s account may exist in an on-prem active directory server, it may exist in a cloud active directory server, and it may exist in both places. Consider user management and access before trying to map out an architecture.
Applications may be dependent on each other, and so can their data. Does historical data need to be migrated? If so, in what timeframe, and how will that impact business operations? Cloud migrations are more successful with a plan.
Meet with an Analytics8 expert who can listen to your company’s questions and needs to help with your move to the cloud.
Collin is a Staff Consultant based in our Chicago office. Collin graduated from DePaul with a degree in mathematics and has since been developing and delivering technology solutions on many BI platforms, including Tableau, Microsoft BI, and SharePoint. Outside of work, Collin enjoys golf and is a craft beer enthusiast.
To thrive with your data, your people, processes, and technology must all be data-focused. This may sound daunting, but we can help you get there. Sign up to meet with one of our analytics experts who will review your data struggles and help map out steps to achieve data-driven decision making.