Blog

Tips for integrating AI, analytics and automation into the customer experience

Lauren Kindzierski, VP of marketing at HGS, shares that forward-thinking companies are using a combination of bots and brains to deliver next-generation customer experiences.

Photo by istock.com

October 31, 2019

By Lauren Kindzierski, VP of marketing, HGS

Customer service leaders across North America are in full swing preparing for the busiest time in retail. Traditionally, much of this preparation has focused around staffing up to manage peak volume, strategizing to hit KPIs, hardening websites and IT infrastructure and optimizing inventory and fulfillment. Fortunately, as a result of the proliferation of artificial intelligence, automation and predictive analytics, tides are turning in retail and this may be one of the last seasons we have to rely on tried and true ramp-up methods.

Here are some tips to prepare for now, as well as the next generation of retail customer experience.   

Reducing the contact center chaos of the peak season

While the holidays are not the best time for piloting something entirely new, making small changes now can improve peak manageability and prepare your organization for later automation projects.

Straightforward tweaks can fuel your automation business case. In the front office, for example, many customer care professionals, with the help of IT, have already automated (e.g., using IVR, scripts, email reminders and alerts) some customer interactions such as: product and store locators, gift card balances and password reset. In addition to the traditional automation tactics, adopting an SMS/text notification strategy for order status updates, shipping alerts and returns confirmations is entirely feasible within the weeks leading up to Black Friday, ultimately reducing the need for additional agents and contacts in their centers.

In addition to reducing headcount, automation can positively affect revenues. In one successful e-commerce operation, a name-brand manufacturer and online retailer of camping equipment improved conversion rates by 40% by devising and embracing an abandoned shopping cart strategy. When a customer removed an item from his/her online shopping cart, an automatic 10% discount offer popped up. If the customer accepted the 10% discount, a live chat agent appeared to close the transaction.

Even for those retailers reluctant to automate customer-facing processes, there are opportunities to implement robotic process automation (RPA) for back office processes such as agent screen transitions and lookups to respond to incoming requests faster.

Manage social media volume spikes by leveraging AI

One of the biggest challenges in social media management is handling an ever-growing volume of posts with limited resources. Engaging with actionable posts and weeding out non-actionable posts is a time-consuming process. To address this challenge, brands can build an AI model to extract spam, news articles, retweets, international posts and other non-actionable content so that agents don't waste valuable time reviewing posts that don't warrant a response. In addition, AI can be used to help prioritize posts in the order of importance. For example, posts where customers are in-store shopping and need immediate help locating an item would be prioritized over someone asking a typical customer service question, such as inquiring about store hours.

Use predictive analytics to capitalize on customers' next move

Today it is very easy for customers to ‘engage' and ‘buy' using channels like mobile, social media and e-commerce. At the same time, customers have started expecting much more from brands, such as seamless experiences across channels that reflect history, preferences and interests. Customer service leaders need to have access to a unified customer data platform that can provide them with data-driven insights to help understand each customer's profile, journey and history across channels. Then, customer service leaders should leverage a predictive analytical model to become proactive in customer engagement strategies such as: predicting the next step in the customer journey, understanding your high-value customers and their purchasing behavior, predicting customer churn, as well as cross-sell offers that customers will likely say yes to.

How to prepare and execute

Implementing CX automation, analytics and AI is best done in phases. It can take six to 12 months to devise, implement and optimize the model and find the right balance of bots (machines) and brains (humans). Taking the leap typically involves:  

•    Mapping the customer (and agent) journey, in the front and back office.
•    Engaging a data analyst to analyze historical contact driver reports.
•    Objectively determining the top 10 reasons customers contact you.
•    Determining which of those 10 contact types could be automated entirely, partially automated, or prevented.
•    Mapping both the customer and agent journey for each contact type.
•    Finding a partner who has AI, analytics and automation capabilities.

Forward-thinking companies are already using a combination of bots and brains to deliver next-generation customer experiences and optimize operations. Even with this year's holiday season around the corner, it's a good time to start preparing for next year's continuous improvement projects to remain competitive.