Predictive analytics use historical data to deliver insights and results that predict future events, trends, and outcomes. Business leaders apply these predictions to their enterprise strategies to try to gain a competitive advantage.
IBM Watson is the most well-known use of predictive analytics. Watson is used for a variety of purposes, including helping businesses predict customers’ behaviors and spot cybersecurity risks.
TechRepublic’s cheat sheet about predictive analytics is a primer on this popular big data practice. We will update this guide periodically with the latest information and tips about predictive analytics.
SEE: All of TechRepublic’s cheat sheets and smart person’s guides
Executive summary
- What is predictive analytics? Predictive analytics uses historical data that are relevant to your business, and then applies a combination of statistical algorithms and/or machine learning techniques to determine the likelihood of future outcomes or trends.
- Why does predictive analytics matter? Predictive analytics give organizations a competitive advantage. Business leaders gain knowledge that helps them make decisions about which processes and products need improving, where to invest in staffing, maintenance, and innovation, how to reduce operation costs, and more.
- Who does predictive analytics affect? Predictive analytics technology is used around the world in public and private industries by healthcare providers, city planners, sales and marketing professionals, systems admins, and others.
- When is predictive analytics happening? Companies in virtually every public and private industry sector use predictive analytics.
- How can I start using predictive analytics? Predictive analytics solutions are available in on-premise and cloud-based systems.
SEE: Free PDF download–How to build a successful data scientist career (TechRepublic)
What is predictive analytics?
Predictive analytics uses historical data from structured, unstructured, and semi-structured sources that are relevant to a specific business, and then applies a combination of statistical algorithms and/or machine learning methods to ascertain the likelihood of future outcomes and events. A great example is a credit report.
When a financial institution performs a credit check on a person who is applying for a loan, the predictive analysis in the credit check looks at the individual’s current assets and debts, his/her employment, and his/her history of paying off loans and indebtedness. Based on these elements and other factors, the analytics produce a score that indicates to the loan underwriter the likelihood that the person applying for the loan is a good credit risk.
Additional resources:
- IBM Watson: The smart person’s guide (TechRepublic)
- Predictive analytics and machine learning: A dynamic duo (ZDNet)
- Data to analytics to AI: From descriptive to predictive analytics (ZDNet)
- Quick glossary: Big data (Tech Pro Research)
- 10 things you shouldn’t expect big data to do (TechRepublic)
- Understanding the differences between AI, machine learning, and deep learning (TechRepublic)
- Smart parking, smart lighting, fleet management at heart of Nokia’s IoT platform update (TechRepublic)
Why does predictive analytics matter?
Predictive analytics offer organizations a competitive advantage and make it easier to plan for the future. It could even save lives, for example, in the case of pandemics.
Before predictive analytics technologies were widely available, many organizations devised strategies by manually reviewing forecasts and then projecting their necessary revenues, budgets, and resources. Predictive analytics makes this process easier and more automated, thanks to various tools and analytics software.
Some of the ways that predictive analytics are being used include:
- predicting the likelihood of certain diseases and/or medical conditions affecting specific demographic populations so preventive healthcare measures can be taken;
- predicting the likelihood of parts and equipment failures so preventive maintenance can be administered to avoid system failures;
- predicting which financial portfolio mixes present the most opportunity and/or the most risk;
- predicting the likelihood of a disruption in a company’s supply chain;
- predicting customers’ preferences and buying patterns;
- predicting traffic flows and infrastructure needs for city planning; and
- predicting critical safety risks on railroads.
It’s also important to remember that predictive analytics can misfire.
Additional resources:
- Big data: Can it predict the spread of Zika? Cloudera thinks so (TechRepublic)
- How big data is going to help feed nine billion people by 2050 (TechRepublic)
- Transforming the agriculture industry using IoT and predictive analytics (TechRepublic)
- How big data analytics help hotels gain customers’ loyalty (TechRepublic)
- HR analytics: An effective yet underused employee retention and recruiting tool (TechRepublic)
- Algorithms can be racist: Why CXOs should understand the assumptions behind predictive analytics (TechRepublic)
- Planet analytics: big data, sustainability, and environmental impact (ZDNet)
Who does predictive analytics affect?
In the not so distant past, only very large organizations with internal data science talent and a lot of money to throw at storage and computing resources could afford predictive analytics; now there are many cloud-based providers of predictive analytics that address a broad spectrum of industry verticals and applications, so even very small organizations can afford predictive analytics by paying a monthly subscription fee.
Additional resources:
- What GM has learned from 20 years of collecting data from cars with OnStar (TechRepublic)
- Big data, business analytics to hit $203 billion by 2020, says IDC report (TechRepublic)
- Predictive analytics or gut instinct: Which works better? (TechRepublic)
- Watch out: Don’t go down the wrong big data path (TechRepublic)
- Job description: Data Scientist (Tech Pro Research)
When is predictive analytics happening?
The global market for predictive analytics is projected to grow to $3.6 billion USD by 2020. Companies in all industry verticals are using predictive analytics to project future outcomes of business strategies and operations, although not every organization is proceeding at the same rate of adoption.
As more cloud-based predictive analytics solutions become available, the playing field is beginning to level between major enterprise players with robust budgets and resources and smaller companies that want predictive analytics but must find tools at affordable price points.
Additional resources:
- AI missed repeat in 2017 Kentucky Derby, but outperformed the experts (TechRepublic)
- ServiceNow launches machine learning, AI automation engine (ZDNet)
- Analytics acquisition deals continue: it’s an all-you-can-eat buffet (ZDNet)
- Download: Big data in 2017: AI, machine learning, cloud, IoT, and more (TechRepublic)
- How Hershey used IoT to save $500K for every 1% of improved efficiency in making Twizzlers (TechRepublic)
- How Anaconda’s data science platform will help IBM speed up enterprise machine learning adoption (TechRepublic)
- Microsoft’s R Server 9: more predictive analytics, in more places (ZDNet)
- IBM Watson bets $1 billion on healthcare with Merge acquisition (TechRepublic)
- Bias in machine learning, and how to stop it (TechRepublic)
How can our business use predictive analytics?
Predictive analytics should be in every company’s technology portfolio. Major vendors, including SAP, IBM, Information Builders, Oracle, SAS, and Microsoft, offer on-premise and cloud-based versions of their systems; this gives companies flexibility and choice when deploying predictive analytics.
The companies most likely to use on-premise systems are in the drug and pharmaceutical industry, research institutes and universities, life science companies in areas like genomics, and other companies that require high levels of analytics, compute power, and predictive intelligence that are central to their businesses.
For smaller and midsize companies with limited IT spend, or for companies where predictive analytics is critical to the business but not a core component, predictive analytics solutions are available on a per usage or a per subscription basis from cloud-based providers. Most of the cloud-based vendors offer “try and buy” opportunities so companies can test the software first before entering into a contract.
Companies in the early stages of using predictive analytics might want to look into cloud solutions that are offered as Software as a Service (SaaS) and that combine predictive analytics targeted to the needs of a specific sector (e.g., healthcare) with consulting and expertise in that industry. This can help the company launch its predictive analytics with analytics reports and best practices that have already been established for that industry.
A key to whether predictive analytics provides useful insights to companies is the business leaders must know how to harness the technology for strategic advantages. This means identifying the right kinds of data that are able to answer well-construed questions and/or data algorithms so the results of these queries can predict future trends and business scenarios.
Additional resources:
- 6 questions every business must ask about big data architecture (TechRepublic)
- New IBM cloud services could get your company data to the cloud 10x faster (TechRepublic)
- 10 books to get you started on big data: TechRepublic’s picks (TechRepublic)
- Executing a Predictive Analytics Proof of Concept (110consulting)
- 3 ways to maximize your big data team’s cognitive computing investment (TechRepublic)
- How a PhD in data science can fix the talent gap and improve the discipline (TechRepublic)