New CX Advisory Board member Sean Cramer, the former head of VoC at Amazon, recalls a recent project that empowered the tech giant to close the feedback loop for its millions of customers.
After gathering feedback through various inputs, Amazon’s VoC team is responsible for the treatment of the data. Its mission is to convert insights into actionable takeaways that can upgrade customer journeys across the world. This analysis does not only consist of locating widespread trends in the customer base, but deep dives to identify individual user preferences.
-
Closing the loop with customer feedback
Closed-loop feedback systems are integral to preventing unsatisfied customers from churning. This presents an opportunity for an organization to demonstrate, in a personal way, that the feedback customers are taking time to provide is being acted upon – and the company cares about the outcome.
Reflecting on the drive behind the move to close the feedback loop with Amazon users, Cramer said: “Traditionally feedback that comes through the CX journey falls into what many call the ‘feedback black hole’. Customers provide insight but businesses fail to follow up.”
This notion proved true in the findings uncovered in the 2019 Global State of Customer Experience. Only 14 percent of respondents to last year’s survey said they always close the loop with feedback.
Rather than taking problem-focused approach, Amazon wanted to show its customers how it was taking the initiative to resolve issues.
However, with the e-commerce giant receiving millions of pieces of feedback, it quickly became obvious that a manual process would not suffice. The team needed an automated process that notified customers of improvements being made to neutralize pain points.
Despite many vendors claiming to be a good fit for the project, Amazon was unable to locate a provider that could collate the multiple feedback sources and drive meaningful change. To get the project off the ground, Amazon drew on internal expertise and skillsets to craft a bespoke solution.
-
Theming VoC data
The first step was to structure feedback data into themes such as pricing or selection. These themes would inform machine-learning measures and algorithms that would trigger follow-up communications.
Measures had to be taken to prevent automated messages from seeming artificial.
Cramer says: “One of the things that kept me up at night with this project was the task of ensuring the responses didn’t seem canned. As customers ourselves, we have all received canned messages from companies saying: ‘(insert name) we have got a deal just for you’.”
The follow-up emails in this particular project needed to convey, at scale, that Amazon genuinely cared about the feedback and was actively looking to listen and improve.
[inlinead-1]
“We needed to achieve the right level of personalization without manually hand-typing millions of emails,” says Cramer.
-
Customer feedback data tagging
The themes used to categorize feedback data were tagged so that follow-up messages could be customized to the specific situation. The tagging system needed to be accurate; otherwise, the automation system could malfunction.
“The last thing you want to do is notify a customer that you have made changes with something but you, in fact, have not yet made any changes,” Cramer explains. “Trust takes a lot of time to build but can be destroyed within seconds. So the point of all of this is to ensure that you continue to build that trust and don’t do anything to damage it.”
Cautious tagging allowed Amazon to personalize messages accurately. On many occasions, these messages encouraged the customer to view the changes made. Metrics were used to capture customer sentiment including NPS. Customers that were engaged via this project became more likely to recommend Amazon because the firm had vocalized its active effort to improve their personal experience.
Cramer advises: “If you do not have a way to close the feedback loop with a customer then you need to rethink how you are collecting feedback in that channel so that you can.”