09 November 2017

How and Why CMOs Need to Use Data Science to Improve the Customer Experience

Customer experience (CX) is how brands have to differentiate themselves today. It’s what ultimately matters not only for attracting new customers, but also for customer retention. After all, it is far more difficult and expensive to find new customers than retain the ones you already have. While there are several aspects of CX that come down to user experience and customer service skills, today CX is also largely driven by data science.

A UBM Tech study found that 94% of companies surveyed drastically increased their customer retention efforts since 2010. However, customer loyalty is a metric that needs to be gauged in order to get an idea of just how effective these retention efforts have been.

Only 38% of the respondents felt that their current loyalty generation efforts were effective and this is largely due to the absence of predictive analytics in customer relationship management (CRM.) This causes marketing departments to use unstructured data such as social media in order to gauge customer loyalty and likelihood of returning to make a purchase.

Gaining a Singular View of Customer Data

mobile-experienceGiven the numerous moving parts in a large organization’s customer service infrastructure, a singular view of the customer had been incredibly difficult to achieve over the years until recently.

A singular view entails seeing not just customer information and purchase history, but also previous interactions with the company – whether that be with sales, marketing, support, accounting or service – all across the enterprise.

Real-time data collection like speech analytics can increase customer satisfaction by efficiently routing calls based on speech patterns, such as matching customers to operators who speak the same dialect or have been trained for certain situations where a certain tone of voice is detected.

CRM profiles that can be seen and updated in real time also contain detailed records of the customer’s reason for calling which also can lead to more efficient service down the line, as well as isolate areas that need improvement.

It’s not just customer service agents who need to see this data in order to know what to expect during interactions with the customer on any given channel. Key decision-makers need these singular views of the customer in order to create more accurate and informed data models for predictive analytics.

For example, if customers frequently keep interacting with the customer service team but not completing their interactions, it’s reasonable to predict what behaviors are going to cultivate repeat purchases or result in higher churn rates.

The More Data, The Better

shopping-channelsThere are many factors that go into customer loyalty but the more data there is to approach a reasonable metric (such as incorporating lifetime value and frequency of purchases), the more accurate predictive analytics are likely to be.

For organizations that aren’t investing in the technological solutions – or the integration of their existing solutions –  that would result in accurate real-time and predictive analytics, they have to rely on more speculative data like social media activity to determine if customers were leaving due to dissatisfaction with customer service.

Given that unstructured measures like social media and email are important means of contact for brands, as well as potentially public-facing, they are important but don’t produce data that is as accurate at predicting customer behaviors, desires, and expectations.

Ultimately, the goal of customer service is to solve the customer’s problem in as few interactions as possible, and allow the customer to self-serve as they see fit. Intelligent call routing can help with this along with speech analytics and having that singular, holistic view of the customer that will inform customer service agents as to what their reason for contacting them is likely to be.

Companies Are Moving to Predictive Analytics Because They Work

know-your-customer-stickynoteThe future is never 100% certain but past behaviors, interactions, and purchases can tell the marketing department a lot about what needs to be fixed with customer service processes and what will get customers to stay.

It’s said that the top 20% of customers are going to bring in 150% of profits and real-time analytics and CRM integration can immediately tip off who’s in that top 20%. Subsequently, companies that have taken active measures to use predictive models report 73% more sales than under the old methods that use dated, siloed technology or unstructured data. Revenue from retained customers increased by 60% and overall ROI was 376%.

Clearly, this means that data science is the present and future of strong CX. As predictive analytics become more sophisticated, it will enable data scientists to build more accurate models and further increase customer loyalty and revenue from the top 20% of customers.

Does Your Organization Have the Right Customer View?

Getting the holistic, 360 degree view of the customer requires building the right customer roadmaps, and integrating your customer knowledge systems, including service, CRM, ticketing and marketing, among others. PTP has helped many top organizations, like PayPal, unify their systems to provide that holistic customer view.

Contact us to see if we can help you achieve your CX goals.

Authored bY

Mark Pendolino

Mark Pendolino is the Director of Marketing at PTP, overseeing the creation of customer experience content focused on helping organizations discover best practices for evolving the customer journey. Prior to PTP, Mark managed teams for companies such as Microsoft, Smartsheet, Fujitsu, and Parsons Brinckerhoff. Mark holds a master’s in Communication in Digital Media from the University of Washington, and a bachelor’s in Technical Communications from Metropolitan State University of Denver. In his downtime, Mark likes to thrash a bit on the drumkit and pretend he’s a rock star.


Subscribe today to have our stories delivered directly to your inbox.

Invalid Email