How can performance marketers tap into AI to use customer feedback surges to stay ahead of the customer curve?
Chris Martinez and Kevin Yang, co-CEOs, Idiomatic, explain how to bridge the gap between how customers speak and how organisations tend to describe and label their issues.
Today’s brands are still trying to drink from a fire hose when it comes to customer feedback. From support inquiries, social media, app/product reviews, chats, surveys, forums, and more, the surge of information from various digital sources can’t be effectively managed with manual analysis, human tagging, NLP, and survey-centric customer feedback initiatives. Keeping up with the feedback and engaging customers across so many channels isn’t just an operational and customer satisfaction challenge. The value of the insight to power better business decisions and reshape how companies address customer experience is lost if these data points don’t roll up into trends and unify real-time insights both across the board and via specific channels.
Top digital brands like Instacart, Thrive Market, FabFitFun, Pinterest and more are using artificial intelligence (AI) platforms to translate millions of customer feedback data points from various digital sources into easily understandable insights. AI is helping them stay ahead of the customer curve by improving satisfaction, cutting support costs/complaints/issues and getting more from their customer data.
Let’s take a look at how they are doing that, the results they are experiencing and how these lessons can be leveraged by other brands hoping to do the same:
Pinterest: Optimizing Help Center Resources to Leverage Employee Time More Strategically
Pinterest’s customer feedback primarily comes from support center calls and from feedback on help center articles. When Pinterest’s monthly active users grew 37% year-over-year, it needed a way to cut down on contact center volume and improve the customer experience so that it could focus on servicing the content that was most valuable to users, rather than reviewing generic positive/negative scores on articles and manual support processes.
To remedy this, Pinterest worked with Idiomatic to create a custom categorization for customer feedback on help center articles and trained machine learning to accurately predict these categorizations. Reviewing ticket data at this level of granularity provided valuable insights to Pinterest on how users talked about certain features, and informed new topics or subtopics in the help center. The more specific help center topics improved searchability, organization and suggestions based on how customers were talking. Additionally, a dashboard was created to highlight articles with low ratings. The User Education team was able to update these articles that caused confusion, or didn’t address questions users had so that it answered the top-rated questions. After the articles were updated, user sentiment could be re-reviewed to see if the improvements made a difference.
This cut down on customer confusion and support contact volume, leading to a 31% decrease in contact volume and a 32% decrease in complaints. Creating and adapting help center content to address common questions sent to support alleviates agent time and customer frustration – a win-win.
FabFitFun: Reviewing Real-Time Insights on the “Why” Behind Customer Satisfaction Scores
With two million subscribers, FitFabFun needed a data-driven way to cross-functionally understand the voice of the customer. Getting surface insights from manually tagging data was a time-consuming task that only allowed for the analysis of small samples. Tapping Idiomatic’s AI platform, FabFitFun was able to analyze and categorize text survey responses and support contacts in real-time. This saved time and allowed for faster decisions to be made.
As a result, FitFabFun was able to:
Proactively address nuanced customer pain points, leading to a 28% decrease in contact volume
Address specific reasons for customer churn, which helped improve retention and decreased complaints by 49%
Use product feedback categorization to get product insights without the need for manual analysis and guided merchandising decisions, which led to a 250% increase in product satisfaction
Instacart: Tapping Ticket Routing to Uncover Nuanced Customer Pain Points and Combat Contact Center Surges
Nothing is worse than connecting with a customer service agent and realizing you have to be transferred to someone else to help with your questions. Now picture you have four million monthly customer support contacts and users self-selecting their ticket category, which can often lack the precision needed to involve specialized agents. Instacart integrated Idiomatic with Zendesk to categorize support contacts in real-time.
By assigning customized AI categorizations to tickets in Zendesk, Instacart streamlined support workflows with ticket routing, agent specialization, and spike notifications and was able to uncover nuanced customer pain points in the process. By routing tickets to specialized agents, the company was also able to reduce support time and save $445k in annual support costs by proactively addressing specific customer pain points and driving down contact volume. Instacart also used Slack to monitor real-time spike notifications to address spiking issues (which happen about 20+ times a month) quicker.
According to Bain&Co, companies that excel at customer experience delivery have higher revenues than the rest of their market. Customer experience matters, and these are just a few examples of how AI can be leveraged to turn surges of information into valuable insights that can help brands stay ahead of the customer curve. Brands will increasingly turn to AI platforms to stay ahead of the competition, empower internal teams, reduce resource needs, strengthen their products and, ultimately, ensure happier customers.
By Chris Martinez and Kevin Yang