In the first article of this series on cloudification, we dispel common myths and misconceptions of migrating your SAS workload to the cloud. This second instalment discusses how you can field support for your migration initiative. The support of different stakeholders in your organisation is a key factor in ensuring a project’s success.
This article helps you identify the different stakeholder types in your organisation and how SAS in the cloud can make life better for them on a day-to-day basis. We discuss the different apprehensions other users have that you may not have considered. Finally, we discuss how moving SAS into the cloud could address some of their concerns, and perhaps help them to be more productive.
At MySASteam, we observe that the stakeholders in most organisations, regardless of different titles, can be categorised in 4 groups. They are: Visionaries, Practitioners, Enablers, and Sponsors.
Figure 1. Personas Framework
Practitioners are responsible for performing analyses of an organisation’s data. Some of the roles of Practitioners include data analyst, data scientist, analytics manager, business analyst, data quality analyst, quantitative analyst, data engineer, statistician, and machine learning engineer. Some roles are specific to the business units that benefit from their work. For such jobs, the titles can include marketing analyst, operations analyst, and transportation logistics analyst, among many others.
This section discusses how Practitioners benefit from migrating SAS to the cloud.
SAS9 and SAS Viya have a complete suite of products and solutions that simplify the data analytics lifecycle.
Because the tools are so well integrated, practitioners spend less time uploading, downloading, transferring and converting files from different formats. Authorised users from different business units can access the data without routing it through IT. The data analytics process can proceed unhampered since the necessary users are empowered with self-service capabilities through a cloud implementation.
Figure 2. Data Analytics Lifecycle
Many organisations have already shifted their data storage to cloud service providers. By moving the compute processes required for data analytics where the data is, performance is increased.
Data preparation takes up most of the time of practitioners when they could be focusing on analysis work. Most data preparation can be repetitive and can thus be automated.
SAS Data Preparation has built-in AI that can minimise the effort in basic and repetitive tasks. The AI suggests transforms based on problems found in your data set. A Practitioner can choose to accept the suggestions or not.
SAS Visual Analytics automates many data exploration functions including determining distributions of variables, correlation between variables, and segmentation of data sets. These functions can be performed using an intuitive graphical user interface and that does not require any coding at all. With the automation of data exploration, Practitioners can proceed to modelling faster than ever before.
Sharing regular reports in SAS Visual Analytics can also be automated in SAS Viya 3.5 and up. Practitioners can share the report once and recipients who would like to receive updates can subscribe to be notified by the system every time the new report has been released. This means activities that don’t add much value, like producing regular reports, saving, attaching and emails, are lessened.
SAS Viya 3.5 and up allows Practitioners to create scheduled jobs with time-based or condition-based triggers. Practitioners can quickly and easily, schedule repetitive tasks, review results and monitor progress.
SAS Visual Analytics, SAS Data Preparations and SAS Data Management have condensed the most used functions for data analytics into drag and drop functions that require minimal coding. In doing this, they allow users with minimal SAS experience get up to speed in no time.
Practitioners can use SAS Quality Knowledge Base, to help with standardising of customer and product related fields. Transforms can be created using AI driven suggestions.
The SAS Data Quality product is useful for cleansing, standardisation, and profiling data coming from multiple sources. Newer data sources such as Twitter and Facebook feeds, Google Analytics, YouTube, Google Drive, ESRI, and local files can only be processed through a cloud-version of SAS Data Quality.
Viya 4, has data lineage integration for Text Analytics, Forecasting, Data Mining and Machine Learning, and Visualisation products. This allows an organisation to be accountable for transformations or analytics performed on a dataset. Practitioners can quickly and easily investigate where the data is coming from, how it has been transformed and which reports rely on it. Less time is spent on making sense of many lines of code.
In SAS Viya, Practitioners can experience the seamless integration of their programming language of choice, typically R or Python. Practitioners can focus on analysis data using their preferred tools rather than being bothered with data wrangling or copying datasets from one system to another just to be able to take advantage of their preferred tool of trade.
SAS in the cloud empowers data scientists with the latest machine learning tools and algorithms. With these tools, Practitioners can uncover trends and patterns in Big Data, which was previously limited by on-premise resources. Some of the machine learning methods that are available only on a cloud implementation are listed below:
Table 1. Updates in SAS Viya Machine Learning and Statistics Libraries
Machine learning methods are not the only ones that have been upgraded. Econometrics, optimisation, and forecasting now run more efficiently with in-memory processing available in SAS on the cloud.
Table 2 Updates in SAS Viya Advanced Statistics Features
Creating sandboxes for data scientists is made easy with SAS on the cloud. Practitioners can perform compute-heavy processes without affecting other users. With the cloud’s scalability, practitioners benefit from auto-scaling resources for RAM, I/O, CPU etc based on their application demand. When the analysis is over, these resources can be returned or scaled down. As a result, analysis is done faster and more efficiently.
For organisations to convert observed trends and patterns into actionable insight, Practitioners should be able to present their results to the relevant business units in compelling and easily understood ways. SAS Visual Analytics on Viya has been massively upgraded to enable data scientists to do this.
Table 3. Updates in Visual Analytics Reporting Capabilities
Further, partners offer reporting tools that simplify the creation of visualisations. Instead of building reports from scratch, Practitioners can use existing templates and leverage their partner’s intellectual property to facilitate the creation of the necessary reports.
Data analytics should be performed on accurate, complete, and consistent data, otherwise the results can be misleading. Moving SAS to the cloud helps Practitioners with:
Enablers ensure that the Practitioners have the right resources for the task. They may include IT Managers, Cloud Service Providers (for outsourced initiatives), HR & Admin Managers.
Enablers are most concerned about the following issues:
Migrating SAS to the cloud addresses these concerns and brings many benefits to your organisation.
On-site hardware can be considered a big-ticket purchase and may take a longer time to obtain approval. Because of the high costs of a server, companies need to be absolutely sure of their requirements. In-depth analysis for this can take a while to produce.
Once the requirements have been identified, suppliers must arrange for the units to be delivered to a client’s site.
Before your company can turn on the unit, a physical space must be secured with the proper temperature, humidity conditions, and access, in which case a room fitout may be in order.
The necessary software to run your applications must also be procured. Managing different suppliers for this can also take some resources.
Before all of these can be put together, it can take weeks or sometimes months. Whereas with a cloud deployment, securing a server can happen almost instantly.
During capacity planning, IT practitioners are faced with a choice, whether to prioritise performance over costs. Traditional IT infrastructure requires companies to commit to large resources even if their workload only demands these resources intermittently. Data analytics workloads are often performed in batches at fixed intervals. When these batches are not being run, resources are under-utilised.
Conversely, if a company prioritises costs and chooses lower specifications, they can experience poor performance. It will take longer for a job to run and the efficiency of Practitioners can also go down because of longer response times. Furthermore, constant overutilisation of resources may decrease the lifespan of your on-premise hardware.
Moving SAS to cloud solves this with elastic and scalable hardware deployment. Server performance including disk usage, memory, network, I/O and CPU utilisation can be easily monitored with interactive GUI. Adding or removing resources can be done in a click of a few buttons. Instead of reducing costs by decreasing the overall capacity of your hardware, you reduce the cost by limiting the usage of the hardware that you need.
As seen in the Figure 3 below, SAS batch processes can perform faster with the support of added resources, only when necessary. The cloud’s elasticity addresses Enablers’ main concern of efficiency.
Figure 3. Resource Provisioning in On-Premise vs. Cloud-Based Models
Most companies have already shifted some of their operational systems to the cloud. By moving the compute component of SAS to the cloud, resources are used more efficiently. Reducing network and I/O requirements can also result in a reduction in cloud expenses.
Soe projects with on-premise resources face problems with testing. Applications may run differently even with only slight changes in a testing environment. Elastic resource allocation allows Enablers to develop and test their solutions in a replica of their production environment without extremely bloating the costs to maintain such resources.
Moving to the cloud, especially a public one, raises legitimate concerns about the security of data and its processing.
Ironically, by moving your SAS deployment to the cloud, your data and applications can be made more secure. Data centres are rated from Tier 1 to 4, ensuring a minimum uptime of 99.671% or maximum downtime of 28.8 hours per annum. Depending on the sensitivity of your processes, a more responsive level of service may be secured.
Tier 4 data centres boast security measures including: 24/7 CCTV, security personnel, biometric controls, data redundancies, 96 hours of independent power for power, data storage, network links, and cooling.
Most SMBs do not have the resources to arrange enterprise-grade security measures unless it is through a cloud provider.
Organisations normally operate under rules, or treatises to maintain a certain level of standards within the industry that a company works in. By moving your SAS workload to the cloud, your organisation implicitly abides by many data and online privacy principles adopted in the development of SAS’ cloud-native solutions.
In a white paper, SAS reports that SAS Solutions OnDemand adheres to data and online privacy principles of TRUSTe. Vulnerability testing is performed periodically, and third part manual penetration tests are conducted regularly. Further, SAS Solutions OnDemand has undergone audits and complied with the following:
While compliance remains under the purview of each cloud service provider customer, Azure and AWS provide clear pathways on compliance with the necessary regulating body (by industry, function, or geography) through its wide array of security tools, frameworks and alignments.
Without moving the SAS deployment to the cloud, compliance with the pertinent regulating bodies is done manually, unassisted by the frameworks shared by the cloud service provider.
With virtualisation, software fixes can be deployed faster and easier, with fewer dependency issues. Propagating the fixes to all users is made simpler through containerised deployment. This DevOps approach to releasing upgrades and fixes removes bottlenecks in improving current software and tools.
Moving to the cloud may require a different skillset from what is currently available in your company. There are many partners that can help you augment the skill gap in an as-a-service format.
For SMBs that have a limited number of staff that fall under the Enabler category, migrating to the cloud can reduce the workload significantly. This helps the team focus on tasks that bring more value to a company.
Organisations normally begin with a bright idea. The ones behind this are Visionaries who see the future and how to make it better. They set the direction in which an organisation should head towards. Roles that fall in this category may include CEOs (or executives in the C-Suite), Directors and Department Managers.
Visionaries are primarily concerned with the following:
Analysis of big data can give your company an edge over your competitors. To give you an idea of the impact of new data sources, here are two examples.
Consider a call centre where call logs are recorded. SAS on the cloud allows you to take those audio files and conduct analytics on it. Before the cloud, on-premise servers would have a difficult time keeping up with the volume of data and the resources needed to process it. With the scalability of a cloud implementation, your organisation can glean valuable insights from unstructured data such as calls.
SAS Viya has been used to perform real-time analytics to extract call topics, sentiments and emotions, call intent using SAS Viya’s audio translator, Visual Text Analytics, and EM Text Mining products. The results of the analysis have lead to faster and better recommendations by an agent to a caller, thereby increasing customer satisfaction.
On the sales and marketing front, brand perception is being shaped by newer channels every day. Moving to the cloud helps your marketing team keep up with your customers. Cloud-native SAS Viya helps you get a pulse of what your customers feel about your company and its offerings. Imagine being able to conduct analytics on data from YouTube channels, forums, Facebook and Twitter feeds, and other social media sources.
Cloudification of SAS allows Visionaries to enlist the entire organisation to come up with the
next new thing. It’s widely recognised that innovation can stem from anywhere, not just from the top. Data analytics can link seemingly disparate information. To quote Professor Lynn Wu of the Wharton School of Business, “What I find is that analytics can really drive the creation of recombinations, or combining a diverse set of existing technologies in a new way. Each individual technology already exists…(how do we) reuse something that we know solved one problem, but apply it to a different domain?”
The connectivity and flexibility that SAS in the cloud offers empowers all business units to be data-driven for more inclusive solutions and improvements to current business processes. Instead of innovation coming from the gut feel of a selected few, generating insight becomes more methodical. With this discipline deeply engrained in company culture, better solutions can be put forward at a faster rate.
By moving to the cloud, you enable your business units to pinpoint inefficiencies themselves.
For example, applied on production lines, analytics allows you to pinpoint causes for lower productivity. Specifically, one can identify machines that require preventive maintenance before it becomes a bigger problem. Analytics can reveal bottlenecks in supply chains. It can also reveal quality issues that can ruin the reputation of a company. For organisations that are scattered geographically, a cloud-based analytics platform helps to aggregate the data needed so that data scientists can unearth insights, regardless of where the data sources are.
Further, expenses for external management consultants can be reduced as the business units are empowered to self-diagnose.
During events when working on-premises is necessary, a cloud implementation of SAS can help with business continuity. Data scientists, working from different locations, can still access the data and conduct analyses regardless of where they are. Most cloud providers offer an of 99.671% uptime as a minimum, it’s not easy to match that performance for an SMB essentially trying to run its own data centre.
A holistic view of the company’s current status has the ability to shape where it goes. With this kind of perspective, Visionaries can steer an organisation in the proper direction faster and better than the competitors. This is what Visionaries stand to gain by migrating SAS to the cloud.
Sponsors main concern is the maximisation of an organisation’s assets. They foot the bill for a cloudification initiative. The roles that belong to this category depend on the use case of your analytics project. Normally, the business unit that benefits most from a data analytics initiative provides the budget for the project. In which case, department heads of the relevant business unit are considered sponsors. Some examples include Sales Managers, Operations Managers, and for organisation-wide projects, Finance Managers.
Sponsors often ask questions about how the project will:
Moving to the cloud simplifies staffing requirements. Roles and responsibilities that are not part of a company’s core business can be delegated to a cloud service provider (CSP). For example, server administrators or storage administrators can be transferred out of your organisation, without losing the competency.
This leads to a reduction in costs on many fronts. By transferring out server-related roles, training requirements are drastically reduced. The management of a smaller, more focused IT team is simplified.
Lastly, SAS on the cloud empowers business units to share data securely without relying on IT for data requests. This frees up the IT team to attend to more mission-critical concerns.
With a cloud-based analytics platform, your organisation does not need to tie up limited capital in hardware nor the facilities to house them. Cloud service providers normally charge based on actual usage, typically billed monthly. This frees your organisation to put your resources behind other initiatives.
Existing on-premise hardware and equipment should not prevent your organisation from joining the cloud. In our earlier article in this series, we discussed how on-premise hardware and equipment can be repurposed, either being reallocated to a transactional system or sold in a secondary market. Previous investments in IT infrastructure for an on-premise data analytics platform need not go to waste.
Hardware capacity requirements are not constant. They fluctuate during peak (running of batch processes) and off-peak times. As discussed in the Enablers section, stakeholders make decisions whether to prioritise performance or costs. By moving your SAS workload to the cloud, you can have the best of both worlds. Unlike an on-premise server, you are not bound by your commitment to certain server specifications. Instead, infrastructure can be designed in a way that schedules increased resources during batch processes and revert to minimal ones during off-peak times.
Figure 4. Capacity-Cost Performance
With on-premise servers, it is common to have reduced development and testing environments. Unfortunately, development and testing in such environments can lead to untested scenarios that may emerge when SAS jobs are promoted to the production environment. Because the cloud is flexible, development and testing servers can have the same capacity as the production server, while keeping the costs low as the test environment can be powered down when testing is not required.
When the servers have been tuned properly to make jobs run efficiently, processing can be done faster. In a cloud implementation where an organisation is billed for its usage, this can translate into savings.
Faster processing doesn’t only help your company save on cloud expenses, practitioners that work with more responsive systems are observed to be more efficient. Without waiting times to accomplish their tasks, can accomplish more for less.
When insights are uncovered faster, the Visionaries can make decisions faster. In a competitive landscape, speed is everything. Companies that can go to market faster normally end up edging out the competition.
As such, a migration activity can be justified to Sponsors through cost-savings along with the potential to drive income (through innovation).
The decision to move to the cloud is never made in a vacuum. Regardless of the role you play in your organisation, Practitioner, Enabler, Visionary, or Sponsor, we hope that this article has allowed you to appreciate the concerns of the other stakeholders in your company.
By helping them address the challenges that they face, it may be easier for you to gain their support in your cloudification endeavour. If you’d like to explore migrating your SAS workload to the cloud, leave your details so we can help you win your team over.
Explore migrating your SAS
workload to the cloud
Leave your details below so we can help find the best solution for your organisation.
At MySASteam, we believe that every SMB deserves the Power to Know. Please leave your details below so we can find the best solution for your organisation. We’ll call you.