Your business already gathers a large amount of data—why not put it to work for you? Predictive analytics mines your data to provide a robust, comprehensive view of your customer experience. Imagine identifying positive trends you can leverage to increase profitability, or uncovering emerging issues so you can reduce risk with early preventive action. With such meaningful insights, you have the power to forecast business outcomes before you take action.
To maximize return on your investments, it’s important that your business operates efficiently. A key part of that is using your resources and assets in the most effective ways to achieve key business goals. Predictive analytics can drive efficiencies by helping you identify your company’s strengths, weaknesses, opportunities, and risks, so you know where you need to act and what you need to do.
Predict before you act
Predictive analytics can identify issues before they escalate, enabling your business to take preemptive action and reduce risk. This approach is particularly effective for supply chain management and its extended network of suppliers and manufacturers. Any delay in production or delivery can impact all parts of the supply chain.
For example, distributors can plan for how increased fuel costs will impact delivery schedule and frequency. What are the impacts if you make frequent small deliveries vs. less frequent large shipments? Predictive analytics can forecast the possible outcomes of a process change, so you can plan an efficient strategy rather than finding out through trial and error.
Predictive analytics can also help with:
- Accurate inventory management – Ensure warehouses have enough supply to meet customer demand and limit overstocking the wrong product or inventory.
- Predictive maintenance – Identify potential future equipment failures to prevent unplanned work stoppages, service disruption, and customer dissatisfaction. For instance, utility companies can assess if a water line or transformer is on the verge of breaking and make necessary repairs before any equipment failures.
- Quality assurance – Failure patterns can identify issues in product quality, enabling manufacturers to minimize customer complaints and reduce defective product returns.
- Demand forecasting – Predict customer demand and ensure that your production facilities can supply enough goods.
Effectively manage risk
Predictive analytics also play a key role in managing risk. Leveraging data science is proving highly useful in financial services and insurance industries. For example, banks can identify good (or bad) candidates for mortgages and loans to reduce potential losses.
Analytics are also being used to strengthen fraud detection and prevention. Text analytics can identify trends or language that indicate the likelihood of fraud and identify suspicious claims that need further investigation. Tracking patterns of user activity, such as the location from which a customer typically accesses an account or how quickly the customer enters login information, also helps determine the legitimacy of a fraud claim.
For example, Infinity Property and Casualty Corporation used real-time claims scoring to determine whether claims were legitimate, reducing the time it took to send suspicious claims to its Special Investigations unit from 14 days to less than 24 hours.
Hire the right people
Developing a more productive workforce is another key way to run a more efficient business – and analytics can help you target the right people. Companies can use predictive analytics to identify the skills and professional attributes most likely to lead to high job performance.
For your existing employees, analytics can help you better understand their needs and how to boost efficiencies. Do your teams have the tools they need to succeed? What are common employee grievances? Do you have the workforce you need to achieve key objectives or do you need more people? Predictive analytics can help you forecast workforce requirements, determine how to best fill open positions, and identify the factors that lead to greater employee satisfaction and productivity.
Data without limits
Predictive analytics can open up a wealth of valuable insights to drive a more efficient business. You can use data science to implement new processes, forecast demand, mitigate equipment failure, and protect your company and your customers against fraud. You’ll have the opportunity to make smarter decisions backed by data, and position yourself as a cutting-edge business leader.
How can your business leverage predictive analytics?
Like this post? Take it with you – download the PDF >
With anytime online access and mobile capabilities, banking is becoming more convenient than ever. We can manage our accounts from anywhere we are, transfer money via text message, and make a deposit with just a snapshot of a check. As convenience is evolving to meet consumer demands, financial institutions face greater challenges to compete in a crowded marketplace. Banks need to continuously examine their performance to understand what works well and where they need to improve to ensure they retain customers and attract new business.
Predictive analytics is a powerful tool that uses all the available data an organization can gather to answer key business questions, such as:
- How accurate and reliable are current forecasts and targets based on historical data?
- What factors impact cost and margins?
- How can we detect and prevent fraud?
- How do we reduce churn and retain the most valuable customers?
Traditionally, data capture and analysis has relied on market research surveys. To gain more meaningful insights about a company’s customer experience, predictive analytics enables businesses to leverage a broad range of customer data, including:
- Interactions – email and chat transcripts, call center notes, web click-streams, and in-person dialogues
- Attitudes – opinions, preferences, needs, and desires gathered through survey results and social media
- Descriptions – attributes, characteristics, self-declared information, and demographics
- Behaviors — orders, transactions, payment history, and usage history
As an example, a top 5 bank used predictive analytics to track customer pain points and identify more than 200 emerging issues by analyzing unstructured data from customer emails, banker notes, survey responses, call center transcripts, and other text sources. The bank then developed a customer experience strategy to make improvements to their online and mobile banking services, email and print communications, and other customer touch points. The results included increased customer retention, increased profits, and competitive advantage as a leader in customer experience.
Predictive analytics can also provide banks and other businesses with a clear view into cross-sell and up-sell opportunities, better risk and complaint management, customer satisfaction trends, even ways to increase operational efficiencies. Decision makers can spend less time searching for information and taking action based on best guesses. Making data-driven business decisions helps financial institutions deliver a more satisfying customer experience that helps increase retention and profitability. These insights can also give businesses an edge ahead of the competition—a crucial advantage in the rapidly changing banking industry.
Like this post? Take it with you – download the PDF >
With unprecedented access to numerous competing brands, consumers are more empowered than ever. It’s no longer enough to sell a good product or service; your customers want personalized solutions. But how do you match the right offers to the right customers for more profitable results? With predictive analytics, you can leverage multiple data sources to identify customer pain points, anticipate needs, and create offerings tailored to specific customer segments. You’ll be able to forecast potential outcomes and build marketing campaigns that generate better ROI.
Predictive analytics unlocks the power of your data
Gone are the days of marketing strategy based solely on best guesses or relying on past performance as a predictor of future results. Imagine how much more efficient your sales team could be if they know which products to promote to specific segments, allowing your business to invest marketing resources toward the right audiences.
With predictive analytics, your business can gain valuable insights to increase marketing effectiveness:
- Acquire customers — Who are your most attractive potential customers?
- Grow customers — Which marketing channels do your customers prefer? How do you increase revenue through these channels? What are the best ways to engage your customers?
- Retain customers — Why do your customers leave? How do you keep your high-value customers happy? How can you prevent them from leaving?
Predictive analytics leverages all of your available structured and unstructured data to glean insights that help you understand what kinds of products your customers want and what’s likely to sell well. With a comprehensive data set that might include transactional history, email and chat transcripts, call center notes, survey feedback, and social media comments, survey feedback, and order history, you’ll have a more robust understanding of your customers.
As an example, online technology retailer Sofmap used predictive analytics to examine customer purchase data, and determined their website design didn’t lend itself to simple buying decisions for customers. Armed with this information, Sofmap implemented a user-friendly recommendation engine to improve their customer shopping experience. Their solution worked—page views increased by 67% each month and the company’s profits tripled within a year.
Drive profitable results by analyzing customer behaviors
While effective marketing helps you deliver the right message to the right audience, you also need to know which messages will motivate buying behavior. To get a general sense for what messaging will resonate with customers, companies have traditionally used market research to segment customers by a variety of factors, including demographics, geography, and lifestyle or values.
However, traditional market research is limited in that customers say what they might do in the future, but that may have little bearing on what they actually do. Predictive analysis, on the other hand, can provide insights into customer motivations and needs based on their behavior in the past.
As an example, First Tennessee Bank demonstrated how nuanced audience segmentation can increase ROI on marketing spend. The bank examined how real profit and loss data reflected customer needs, and used the data to develop targeted, high-value marketing offers for various customer segments. The results? A 3.1% increase in marketing response.
The bank saved on operational costs as well. Because they were able to identify a segment of customers most likely to respond to their offer, they reduced mailing costs 20%, and printing costs went down 17%. Rather than blasting the same piece to everyone on the bank’s mailing list, First Tennessee was able to market a specific offer exclusively to the most attractive segment to achieve the best results.
Knowing your customers’ needs and motivations is a key component of any marketing strategy. Using predictive analytics reveals valuable information about your customers that can drive a better return on your investments through targeted messaging and offers. Competition for customers is fierce no matter what the industry, and the winners will be those who develop a data-driven understanding of their customers.
Like this post? Download and take it with you >
Last week, the Consumer Financial Protection Bureau (CFPB) turned 3 years old. When the CFPB launched in July 2011, it wasn’t clear what power this regulatory progeny of the Dodd Frank Act would wield. Today, there is no doubt about the agency’s authority. With 9 products in its ever-increasing purview, the CFPB now hosts over 275,000 complaints in its public database (400,000 received in total), and has enforced 22 regulatory actions costing top financial institutions $4.6 billion dollars to date. As a result, financial services providers are hearing a serious wake-up call to monitor and improve their practices.
As part of its birthday celebration, the CFPB published a list of figures summarizing its achievements to date. For banks and other financial services companies, one of the figures left out may be the most eyebrow-raising – the meteoric rise in the dollar amount of regulatory fines.
From July 2011-2014, the CFPB has ordered $4.6 billion in monetary relief, affecting 15 million consumers.
Source: CFPB Fact Sheet
The agency’s first target was Capital One, ordering payment of $165 million in fines and restitutions for issues such as deceptive marketing practices. Over time, the average amount of a CFPB enforcement action has increased from $160 million in 2012 to almost $280 million in 2014 – a 75% increase in only two years! Even in 2013, when the CFPB levied its biggest penalty to date – a whopping $2.6 billion to Ocwen, a leading mortgage servicer – the average fine was less than it’s growing to be in 2014. The message is clear – it’s getting more and more expensive to ignore customer feedback, and more imperative to find and fix customer experience pain points.
How not to get invited to the party? Leverage customer data
The good news for banks, credit unions, and other financial services providers is there is a clear way to avoid paying a similar costly penalty. The answer is in the data. Complaints do not exist in a vacuum. They exhibit a lifecycle, and if you trace the journey of a complaint from inception as a customer concern, there are typically multiple touch points before the customer resorted to involving the CFPB. These touch points represent multiple opportunities for a company to fix the customer experience issue and prevent an escalated complaint that could spark regulatory action.
The companies doing the best job at predictive prevention of complaints are using decision tree algorithms to find patterns in their customer data. Want to identify targets of highest propensity for escalation to the CFPB? Your business can explore which products tend to generate problems, which types of issues seem most sensitive, which types of customers might be more negatively impacted, etc. Also look at other angles such as how customer tenure affects their likelihood to complain, and how channel use plays a part in customer pain points.
The answers will be different across products and companies because the experiences differ. The solution, however, is the same:
- Journey map complaints starting with their inception as customer issues.
- Identify patterns in transactional and behavioral data that predict complaints.
- Deploy a decision tree model on current customers to assign a propensity to complain.
- Flag customers with the highest propensity to complain, and intervene before their issues escalate to the CFPB.
In the era of data science, we’re no longer analyzing data to answer the question, “What happened?” Today, we are asking “What will happen, and how can we impact it?” If banks and credit unions want to avoid paying hundreds of millions to billions in fines, it is time to embrace predictive prevention.
In our work helping companies with customer experience strategy, data science, and communications, we at Beyond the Arc are really more behind-the-scenes types. Yet sometimes our work comes front and center in recognition and accolades. We’re always flattered when our name appears on “best of” lists, and we wanted to share some recent highlights.
||A panel of CEOs, CIOs, VCs, and industry analysts named Beyond the Arc to CIO Review’s list of 20 Most Promising Data Analytics Consulting Companies, which identified consultants that take on “real analytics challenges.”
||Beyond the Arc included as a member of the “Big Data 100″ by SourcingLine, and recognized as market leaders on their “best of” lists:
Companies were chosen based on “company experience, breadth of knowledge, client references, certifications, industry recognition, and marketing presence.”
||In June 2014, Fortune Magazine mentioned Beyond the Arc in an article about “big data companies to watch” for our experience in predictive analytics.
||Forrester® Research, Inc. included us as an example of a VoC specialist in its February, 2014 report, “Voice of the Customer Vendor Landscape, 2014.” Beyond the Arc was the only consulting-focused vendor included in the report, which also said we are known for helping companies “set up or improve VoC programs through consulting services.”
Check out the Beyond the Arc newsroom for our latest featured articles and contributions to industry analyst reports.
Measuring customer satisfaction is a great first step toward understanding your customers. But it’s rarely enough to drive real business growth or learn how to differentiate your brand. Many businesses rely on the Net Promoter Score (NPS), a simple metric to gauge customer satisfaction based on whether people would recommend the company. NPS may be fine on a high level, as an easy-to-explain, sweeping generalization about your progress (or lack of it), but the score tells only a small part of the story. An effective customer experience campaign needs rich insights.
Why “good” isn’t good enough
A high Net Promoter Score can make your company look great, but it doesn’t provide insight into what is popular and why, which are keys to understanding how to leverage that popularity to build business and capitalize on opportunities. Perhaps more importantly, while a low score is cause for concern, it’s not enough to simply identify that customers are unlikely to recommend your brand. What you need is real insight into what and why customer pain points exist, so you can take targeted action for improvements.
And what about the future? NPS is a good tool for taking the temperature of current customer satisfaction, but it’s no help in predicting how well your customer experience efforts will do in the future. Similarly, the NPS metric doesn’t identify emerging issues that indicate where dissatisfaction is brewing.
How Net Promoter Score works
Using a simple survey, your company can tap into customer sentiment by asking them to rate from 0-10 how likely they are to recommend your company to others.
NPS ranking is based on the premise that your customers fall into three camps: Promoters (loyal brand advocates), Passives (indifferent, could easily go to competitors), and Detractors (dissatisfied, no repeat business).
The percentage of Detractors is subtracted from the percentage of Promoters, and the result equals the Net Promoter Score.
Let’s look at an example: Suppose you survey 30 customers, and 15 of them rank their likelihood to recommend between 9 and 10. That means 50% are Promoters. If 6 people gave a ranking of 0 to 6 on thescale, they are Detractors. 6/30 = 20% Detractors. To determine your score, you subtract the Detractors (20%) from the Promoters (50%), which equals 30%. In Net Promoter Score terms, your score is 30. Typically, a score of 50 and higher is considered excellent.
In a 2014 report, SynGro, a Voice of the Customer software company stated that, “Net Promoter Score is the most widely used Customer Experience metric. 54% of companies surveyed in a recent international CX research study use NPS as a primary measure.” (“Net Promoter Score: Driving Profit with NPS”, SynGro 2014). Yet relying on NPS may mean companies are missing important context to really understand their customer experience.
Whether your NPS score is high, low, or in-between, you get very limited information. That can be especially problematic if your Detractor rating is high. Negative sentiment may be overflowing into social media, influencing other consumers to avoid your brand —and you need to find out why and how to correct that.
Deeper insights give you the power to act quickly
For small companies with simple products and limited customer touch points, surveys and Net Promoter Score metrics may provide enough data to gauge customer satisfaction with the brand. But for companies with numerous products and services, and multiple points of customer interaction, a single number score simply does not provide enough information to take meaningful action. Customer satisfaction programs that treat people as static data points are bound to miss important insights, as Bank of America found out a few years ago.
By tracking Twitter and Facebook comments about Bank of America, Beyond the Arc discovered over 20 service breaks that were only identified through social media text analytics. In our Bank of America case study, we noted that a wave of customer dissatisfaction erupted based on misinformation, which the bank could have avoided if they had identified the issues sooner. A low Net Promoter Score would signal a problem, but social media analysis and Voice of the Customer analytics help target specific issues and provide insight on what customers expect and need to not only resolve the problems but improve their perception of the brand.
An effective Voice of the Customer (VOC) program helps you gain a comprehensive picture of satisfaction levels across the entire customer journey. Far beyond simple surveys, VOC analytics enable you to leverage a broad range of structured and unstructured data sources, such as transactional data and survey ranking, combined with commentary from call centers, email, in-store feedback, and social media.
As Forrester noted in their report on Net Promoter Scores, “NPS is not a fast-moving metric,” and doesn’t answer the question of what’s currently trending. Customer satisfaction may be high today, but tomorrow’s losses could be right under your nose, right now. VOC analytics and predictive analytics give you the power to see what’s coming –and react early and even prevent problems that could impact customer satisfaction.
Net Promoter Score is a good beginning metric that is easy to understand and adapt. However, companies should not rely on it as their only tool for evaluating customer experience. NPS tells only part of the story, and should be factored in along with a Voice of the Customer program that focuses on understanding why customers are happy or not. For best practices on how to get started or make your VOC program more robust, see our blog article, “Building a successful Voice of the Customer program”.
Like this post? Download and take it with you!
Event: Predictive Analytics World, Chicago 2014
Presentation: “Leveraging Predictive Analytics and Alternative Data Sources to Improve the Customer Experience”
Case Studies: CFPB, Capital One, Citibank, Bank of America
Speaker: Steven Ramirez, CEO, Beyond the Arc
Date: Tuesday, June 17, 2014, 3:05 – 3:25 p.m. Eastern Time
Register online >
Using customer feedback to drive decisions is nothing new for businesses, but the locations and forms of conversations have shifted drastically in recent years. Companies now have greater transparency into customer relationships at the same time that individual consumers have a more visible platform to share their positive and negative experiences with individuals worldwide.
In this session at Predictive Analytics World, Steven Ramirez shares how new capabilities, such as using predictive analytics to gain insights from social media and the Consumer Financial Protection Bureau (CFPB) complaint database, can empower banks to identify emerging issues and necessary service changes to prevent loss of business.
Join us on June 17 in Chicago — Register online today >