Do you know “How Facebook Can Predict Your Politics, Your Love Live, and Even Your Sister’s Name?” Nearly six of 10 American adults are Facebook users, providing updates, liking things, and making connections.
Consequently, Facebook is a rich data depository. Analysts have made some interesting discoveries with Facebook’s collected information.
Siblings are 57% more likely to share a first initial than what random chance allows. Congressional candidates getting the most “likes” won races approximately three-fourths of the time.
And, data scientists note cities conducive for those seeking mates have an “inverse correlation between ‘single rate’ and ‘relationship formation’ rate.”
Webopedia tells us “Predictive analytics is the practice of extracting information from existing data sets in order to determine patterns and predict future outcomes and trends.” But it “does not tell you what will happen in the future.”
Rather, analytics provide a forecast “with an acceptable level of reliability”—an important tool for businesses seeking success.
This week: predictive analytics and its role in the financial services industry.
‘Predict what’s next and then have the flexibility to evolve.’—Marc Benioff, CEO of salesforce.com
First, know three basics per “A Predictive Analytics Primer” from Harvard Business Review.
1. Data. Collect data through various interactions. “Lack of good data is the most common barrier to organizations seeking to employ predictive analytics.”
2. Statistics. Regression analysis and knowing how “a set of independent variables… are statistically correlated with the purchase of a product” facilitates understanding how each variable impacts purchase behavior.
3. Underlying assumptions. Monitor whether assumptions remain true. The big one: “That the future will continue to be like the past.” Ask how has consumer behavior changed? How will that be impactful?
Data analytics requires planning. “Community banks need a smart data governance plan that defines what information their systems are capturing, how frequently it’s captured and how it’s protected,” according to “Big Predictions for Big Data for Community Banks.”
Plans should cover business need and identify value to various bank operations.
One consideration is the connection of data streams. “This may be complicated for banks with older systems, but this step is a factor to consider during system upgrades.”
‘One can’t predict the weather more than a few days in advance.’—Stephen Hawking
How is the financial services industry using analytics?
“Financial Firms Embrace Predictive Analytics” says Big Data News. “Firms are increasing investments in predictive analytics amid an explosion of data sources” and four “growing use cases” exist:
1. Optimization of capital risk to “integrate the performance of capital” held in varying locations.
2. Regulatory compliance as analytics finds disruptions created by errors, oversight, or other issues.
3. Operational intelligence is obtained as analytics monitor happenings across all operations.
4. Bettering customer satisfaction and loyalty as companies engage based on consumer behavior.
Analytics help foster efficiencies. A recent Forbes article notes, “Financial services companies have reduced their time to decision-making by 13% with analytics.” This compares to 10% from companies not using predictive information.
An example: Single-screen dashboards presented suggested outcomes in real time, and “helped transaction bankers improve their decision making in the areas of risk, fraud mitigation, liquidity and collateral management.”
Read “Brave ‘Now’ World: Predictive Analytics Disruption in Financial Services” at itbusiness.ca to learn about the new field of “machine-to-machine process intelligence.” Here, analysis moves forward, “based on real-time results” and stems from a flexible baseline model.
This process allows analysis of “Transactional experience moments.” For example, a credit cardholder habitually buys baby products but does not hold a mortgage. “A predictive trigger could recommend that a call from a mortgage advisor to make an offer [will]… pre-empt the… decision to seek a mortgage elsewhere.”
New, unstructured, dynamic data gleaned from the internet and other venues contributes to telling of different consumer needs stories.
“Banks have plenty of data from ATMs, websites, phone calls, emails, and mobile transactions but have done little to leverage it,” says analytics services provider firm 7. As institutions mine data, they will “better anticipate what customers are trying to do and personalize their experience.”
Indeed, “Banks that excel in these areas and engage with their customers by becoming the key custodian of all of their value transactions and empowering them… will be the real winners in the customer acquisition and retention battle.”
Another application for analytics is in hiring decisions. “Wells Fargo uses predictive analytics to hire employees better able to meet its performance requirements and fit into its corporate culture,” according to a 2012 project reported at bai.org.
Using analytics, the bank found teller retention rate up 15% and retention rate for personal bankers up 12%. Predictive analytics helped Wells Fargo identify the “best success indicators” to aid in recruiting efforts.
They discovered the best tellers “tend to have experiences in financial services, retail telecommunications and hospitality, and they also showed strong academic performance in high school and above.”
‘The best way to predict the future is to create it.’—Peter Drucker
Despite the potential of predictive analytics to lead to successful outcomes, not all companies enjoy the anticipated return on investment. Gallup provides some insight with “5 Reasons Why Your Company’s Analytics Program is Failing.”
Pitfalls include: Lack of determining problems to be solved, using incorrect metrics to gather data, a lack of infrastructure and data, the wrong employees on the job, and a corporate culture that does not enable “data-driven decision-making” or is not “capable of making the insights, behaviors and necessary changes a priority.”
A “best guess” prediction is a suggestion of likelihood derived from data analysis; an outline for possible outcomes. By itself, it does not guarantee success.
Success is found in making decisions, continually monitoring course of events, adapting when necessary, and following through. You must be open to welcoming, influencing, and creating change.
Per James Carville, “I’d rather not predict. I’d rather affect.”
LORA BRAY is an information research analyst for CUNA’s economics and statistics department. Follow her on Twitter via @Bray_Lora and visit the CUNA blog, “The Research Roundup: Economic Perspectives.”