Analytics experts from Fortune 500 companies, along with other top data practitioners, will converge on Las Vegas this month—and that's not the only stop on the Predictive Analytics Worldtour, now in its 10th year. Washington DC, London and Berlin will also host PAW conferences, which have attracted nearly 15,000 attendees representing more than 10,000 companies in the past decade.
The wide availability of big data, combined with advanced computing technologies, is expected to push the Predictive analytics market up to $12.41 billion by 2022. Clearly, predictive analytics is a big deal. But for the non-experts among us, crunching data to foretell the future sounds a little like crystal-ball gazing. Does it really work? Let’s take a closer look at how it’s down and why it’s so important.
What is predictive analytics?
Predictive analytics relies on the enormous stores of data from both internal and external sources, along with a range of technologies including data mining, statistical modelling, and machine learning algorithms. And many of us experience it every day—in simple predictive applications like Waze that use data on traffic volume, construction zones or traffic accidents to suggest faster, alternate routes and on social media platforms and retail websites that serve up articles and ads based on past interests.
Gartner defines predictive analytics as information technology that embraces four key attributes:
1. It emphasizes prediction instead of description, classification or clustering).
2. It expedites analyses to hours or days, versus months that may be spent on manual analyses.
3. It focuses on uncovering insights that are highly relevant to the business.
4. It makes big data more accessible to business users.
Predictive analytics is one stage in the evolution of artificial intelligence (AI). Where more traditional business intelligence applications focus on the past—what has happened and why—predictive analytics answers the question: What is going to happen?
It’s an important distinction. As author and organizational theorist Geoffrey Moore says, ““Without big data analytics, companies are blind and deaf, wandering out onto the web like deer on a freeway.” When an organization can’t see where its headed, chances are, it won’t end up where it wants.
By leveraging historic and current data for analysis, predictive analytics empowers organizations to anticipate what’s to come, allowing proactive, data-driven decision making to mitigate risk and maximize opportunities for growth.
Use cases for predictive analytics
The desire to know what’s going to happen in the future isn’t new. Historical artefacts suggest that humans have practiced fortune-telling for millennia. Ancient people relied on the movements of the planets and starts (astrology), divination using dust, sand or salt (abacomancy) and other practices to predict the future. Our fascination with prediction even spills into popular fiction—Hogwarts offers courses in arithmancy, ancient runes and divination.
What is new is the volume, variety and quality of data available to enable accurate forecasts. So, how are businesses using predictive analytics?
• Banking and financial services organizations use it to detect fraud, accelerate application screening and inform buy-sell decisions.
• Retail and entertainment brands use it in their PR, marketing and sales efforts to anticipate trends, optimize customer experiences, inspire product development and generate long-term value for the business.
• Supply chain and risk management professionals use it to forecast inventory requirements, automate onboarding risk assessments and identify potential reputational, regulatory, financial or strategic risks on the horizon.
But there’s one hurdle that most companies must overcome to achieve the desired results from predictive analytics: the data. Forrester Research Analyst Michele Goetz notes, “Most organizations [83%] simply don’t recognize this as a problem. When asked about challenges expected with AI, having well-curated collections of data for training AI was at the bottom of the list.”
Unfortunately, the lack of relevant, clean data is one of the biggest barriers to successful predictive analytics endeavours. Siloed data is common, and the data in each silo—from sales, marketing, customer interactions, etc.—may be in completely different formats. Even when internal data is readily available, companies benefit from complementary datasets, such as news or social commentary, to fill in the blanks and achieve actionable business intelligence.
Organizations at the forefront of applying predictive analytics will realize substantial competitive advantages. Efficiency will climb, and with it, the opportunity to use human resources where they have the most impact—in person-to-person interactions and in making critical decisions that require emotional intelligence. Experts believe that deep learning and predictive AI analytics will be as transformative on society as internet and cellular technology was over the past decade.
Are you ready to make magic with predictive analytics?
