Data for Financial Services
Data is a hot topic, not only for the Financial Services industry, but the business world as a whole. You may not notice it, but data is everywhere. One of the issues with data, though, is that it’s really only useful with context. The difference between Data and Information can be confusing but the distinction is key for data to serve any purpose.
Data is typically raw, unfiltered, and often unstructured in nature; more is not necessarily better than less, and complexities increase when the data increases. When you filter, sort, organize, or otherwise give data context; you now have information. Information is the key to the usefulness of data and the ability to gain actionable insight on your business.
Data has been an afterthought for years, often due to the cost of compute resources and storage. But thanks to advances in hardware and economies of scale, it doesn’t need to be an afterthought any longer.
What is the Cloud?
These are just two improvements in cloud infrastructure, we will talk a bit more about cloud in a later post. To simplify, “The Cloud”, in its most basic form, is a computer that exists somewhere else. Allowing a business to offload the need to have resources on site, and allowing a cloud service provider to handle backups, updates, upgrades, and maintenance, allowing a business to focus on their core service area.
The Storage Problem
Increases in density and commoditization of storage has led to a reduction in the overall cost of storing data. This is important for a number of reasons internal and external to your business. To give you an idea of the scale, 90% of all data (by storage volume) has been created in the last 2 years. The mass reduction in costs, and accessibility of resources allows businesses to track a broader range of historical data across a variety of reference points.
In 1990, storage cost around $4,400 USD per Gigabyte (GB) (Disk Drive). To give a better understanding, this is about 1 hour of High Definition video, 250 songs, or 600 photos. Fast forward to 2019, the average storage cost per GB is around $0.02. This is a supposed price floor for the time being. It is important to note that the price of drives fluctuate based on supply due to weather issues in production regions, last seen at the end of 2011 with the Thailand flooding crisis driving the price of hard drives up to 300% in some cases.
We have seen mass reduction in the newer, and faster, flash storage, allowing even faster calculations and processing of data. In 2000 flash storage was around $1,255/GB and is roughly $0.10/GB today. Analysts are expecting further price cuts in 2019 to as low as $0.08/GB. Due to the prevalence of cheap storage options, businesses can capture more data, and gain much more information on their businesses.
Storage is only one part of the data puzzle. Being able to process, filter, and sort that data into actionable insight requires more power. The more data you have the bigger the problem is to solve.
The Compute Problem
Compute resources have not only reduced drastically thanks to some of the optimizations found though Moore’s Law, but how we manage these physical resources digitally have changed too. Improvements in virtualization software have allowed more cost-effective server resources to be distributed at different scales. Virtualization software, like VMware, allow you to run an Operating System (OS) like Linux or Windows on an assigned block of hardware resources. On the micro this means that you could run Windows 10 on your Macintosh computer, but on the macro, it means a single enterprise server can host thousands of virtual computers that are isolated from one another.
Part of the massive growth of cloud has come from tech giants like Microsoft, Google, and Amazon. One of the largest pieces of Amazon’s business is cloud resources. Due to their immense needs, they started to lease extra compute power to other businesses. This provided an opportunity to capitalize on under-utilized resources. It also increased purchasing power to boost their own capacity, along with providing the resources for businesses across the world.
The Regulation Problem
One of the major reasons many businesses retain on-premise resources is due to compliance. By owning the servers and technology, data is held at an arm’s length and is believed to be more secure. The two key issues here are the rapid depreciation in capacity of owning hardware, and the know-how and capability to maintain these systems through patches and updates needed to prevent the latest vulnerabilities. Many OS’s will only run on specific hardware, and often most hardware will only support older, unmaintained software.
Companies like Amazon have adapted to the changes in requirements from a hardware and security standpoint. Many cloud service providers are starting to offer more regulated offerings. Industries like Financial Services and Government must abide with some of the toughest regulations around security, storage location, and backup. We’re now at the tipping point where owning physical compute resources is a disadvantage in the rapidly changing business environment, and actually cost more than their cloud-based counterpart.
The Knowledge Gap
We have covered storage and compute power, but there is one missing piece in the process. Data Science. More specifically the ability to take complicated data, apply the appropriate (and computationally efficient) algorithms, and gain statistically significant insight. Data science is a bit of an umbrella term including disciplines such as Machine Learning (ML), statistics, databases, and coding.
Until recently, many businesses couldn’t afford, or have the need for, having a full-time data scientist on payroll. The demand, and cost, increased rapidly in 2015 and through 2016 with the emergence of the big data boom. What once were individual academic research roles have been optimized and improved by the way of abstraction layers and compelling new software platforms that can do some of the work that a high value data scientist could do.
With advances in technology, companies like Salesforce are making it easier to understand data. They are not the only provider in the multi-billion-dollar data space but through improvements to their Einstein platform small businesses can leverage the power of AI to gain insight on their business. AI is everywhere and as we move forward, it will become more accessible and empower SMBs, without needing the resources for dedicated scientists, to better serve the needs of their customers.
Enter 2019, the cost of both cloud resources and storage are at a point where it is reasonable for most businesses to leverage data and the entry of abstraction layers that allow you to find insight without needing a degree in data science.
The first step in gaining insight on data and information is by first asking the question “What data do I have and where is it stored?”.
In the next instalment we will talk about the current state of data for financial services businesses and what you can implement today to build for tomorrow. As always if you have any questions on data in the financial services industry and how it might impact you, please get in touch with us here.