We have a huge opportunity to capitalize on Big Data by applying analytical models to historical customer interaction data to predict consumers’ needs.
We are living in the Information Age where enormous amounts of data is being created, published and stored every second. According to Eric Schmidt, Google’s former CEO, “There were 5 exabytes of information created by the entire world between the dawn of civilization and 2003. Now that same amount is created every two days.”
Big Data is made up of both structured and unstructured data. Structured data refers to information with a high degree of organization and is easily searchable and straightforward. Spreadsheets are a good example of structured data because information is properly arranged in a relational database and can be filtered rather easily. Unstructured data refers to information that does not fit well into relational databases. The majority of an enterprise’s data is made up of unstructured data, specifically in the form of text, voice, photo and video.
Now think about how customer service teams communicate with consumers and how these interactions are typically stored. Usually these conversations are in the form of unstructured data including: phone calls, social-media posts, chat logs, emails, video chats, online reviews and photos. Trying to analyze these interactions to report the voice of the customer back to the rest of the organization has become extremely challenging, especially now that brands have 10 times the amount of conversations to sort through due to the rise of social media.
“Smart” customer care interactions use predictive analytics technology and Big Data to ultimately deliver a better customer experience in real time.
“Smart” Customer Experiences
Customer expectations have changed dramatically in the past couple years thanks to smartphones, which have made our lives easier, more connected and real-time. We have become a culture of multitaskers who always look for the fastest and easiest way. For this reason, smartphones are rewiring society, making us a little more impatient. Consumers are starting to lose tolerance with customer service when it comes to interactions that are not “smart.” “Smart” customer care interactions use predictive analytics technology and Big Data to ultimately deliver a better customer experience in real time.
Here are some ways to use Big Data and analytics to deliver a “smart” customer experience:
1. Provide anticipatory customer service. Micah Solomon, the author of ‘High-Tech, High Touch Customer Service,’ describes the concept of anticipatory customer service as companies that predict customer needs and proactively address them. Anticipating a customer’s needs gives customer service an opportunity to provide a wow experience by fixing the problem before it amplifies. This is where predictive analytics technology plays a big role. Predictive analytics is a technology that uses predictive modeling to find the probability or likelihood that a future event will take place, such as placing an order or recommending a friend. In order to predict consumer behavior, you must have lots of historical customer data, which all contact centers store.
Now, think of all of the ways contact centers can leverage this powerful technology to provide anticipatory customer service. For example, by analyzing past order history, predictive models may uncover a specific loyal customer who calls in every Tuesday at 4:30 p.m. to place his same order. Wouldn’t it be a wow moment for that customer if instead customer service called him on Tuesday at 4:30 p.m. and asked if he wanted to place his usual order? This type of proactive service would deepen relationships because customers would feel like brands really knew and cared about them.
No longer does customer service need to be reactionary. Reactionary service is not going to hold today’s less loyal and ready-to-jump customer. Customers should never be told on the phone to go see the website or to email their questions. Instead, by leveraging predictive analytics, customer service reps can anticipate the customer’s preferred communication channel and be ready to handle it.
2. Leverage conversion-based routing. By now the contact center industry has probably heard of and even implemented skills-based routing. Leaders in the industry have taken that technique to the next level by leveraging Big Data and predictive analytics to implement conversion-based routing.
When it comes to outbound telemarketing programs, success is usually based on a number of factors including: lead sources, agent skills, time of day, phone number dialed, script and offer. The challenge with most outbound programs is that the success of campaign is based on lead cost. Another challenge is every phone agent excels at converting different types of leads. For example, Agent Suzy may be good at converting leads that are generated online, dealing with product X and geographically reside in Texas. Whereas, Agent Bobby may be good at converting leads that are generated from List A, dealing with product B and geographically live in Michigan.
To measure success, managers need to analyze conversion rates by source, day, time, rep, and product in real-time. Unfortunately, running data through a dialer makes it difficult to get timely reports with actionable intelligence. At the end of the day, many contact centers find it hard to prioritize leads in order of highest likelihood to convert, assign them to the right sales agent and ensure they get followed up on.
To conquer this problem, leading companies are building predictive models to analyze their historical conversions to identify trends. Then with the help of predictive analytics, contacts centers are able to implement conversion based routing techniques to ensure leads are being prioritized and routed to the best agent, at the right time of day, using the best channel to reach a consumer and finally presenting the right offer. The end result is a decrease in lead acquisition cost, increase in contacts made, decrease in dial attempts and, most importantly, an increase in conversions.
Reactionary service is not going to hold today’s less loyal and ready-to-jump customer.
3. Use predictive analytics to engage in the right social conversation. When companies starts providing social-media customer service, one of the biggest challenges will be learning how to efficiently sift through the social conversions to determine which posts are actionable and which are irrelevant or spam. With over 250 million blogs in the social-media landscape, this is no easy task. However, it’s absolutely critical that brands prioritize and assign the relevant, actionable posts at the top, due to the limited resources available to monitor and engage.
The solution? Once your social-media customer service program has been running for a period of at least three months, take all of your tagged and categorized posts and build a predictive model. The model will then help sort through all the new incoming posts, scoring and prioritizing them in order of relevancy. The great part of predictive analytics is that the model continues to learn from itself, getting smarter each month as more data is analyzed. Companies can also use predictive analytics for social-media customer service programs to predict which social posts need quality assurance teams to review them, which posts have legal issues, what the ideal response should be, and which posts have a lot of brand risk due to author influence.
4. Use predictive engagement on your website and optimized mobile site. Predictive engagement technology is one way that companies can use Big Data to personalize the online shopping experience. This technology uses clickstream data and/or historical transaction data from websites to predict the best content to display to that visitor. If a customer already has an account on an e-commerce website, then chances are the company knows what that customer has purchased in the past. Using predictive engagement technology, the company can now display images, offers and product recommendations that cater to that customer.
What this customer sees on your website will be different from what other customers see. If a first-time consumer stumbles upon your website, there won’t be any historical data. Instead, the technology will leverage clickstream data to predict what this customer may be interested in purchasing. Predictive engagement is not new technology, but it’s very effective and yet under-utilized by most companies. Amazon.com pioneered the use of clickstream and transaction data for personalization years ago by recommending products based on your browsing behavior and purchase history.
Consumers are starting to expect a level of personalization at every step of their customer experience journey. The thought process of customers is “if I am giving up the right to my privacy by supplying you with my personal data, then the least you can do is use it to personalize my shopping and customer service experience.” With all of the Big Data that companies collect and store on their customers, more consumers are demanding brands leverage that data to enhance their customer experience and “really get to know them.”
5. Offer a smartphone “white glove” concierge service. Many companies are starting to develop smartphone and tablet apps specifically for servicing customers. These apps are a great way to offer our on-the-go society help without having to call customer service. Typically, these service apps still allow customers the ability to contact customer service with the click of a button if they choose. The future of mobile customer service will leverage the power of analytics to determine a customer’s loyalty status and then connect them to the appropriate customer service rep when they click the 800 number from their app.
More consumers are demanding brands leverage data to enhance their customer experience.
In addition, analytics will help determine where in the app the customer got stuck. Now the customer service team member can assist the customer from the last point in the process without having the customers repeat what they’ve already done or were trying to do. Lastly, predictive analytics can scrub your customer database to help identify the tech-savvy smartphone users. Once uncovered, service teams could push a bill to them via the service mobile app, so they have the option of paying it from their mobile device, versus having to go online or use an interactive voice response system to pay their bill.
Consumers know that enterprises should be able to do these smart interactions today and have low tolerance if these options aren’t being deployed. Leverage the power of Big Data to capitalize on big opportunities for customer care engagements.