When customers flooded the help site, a change was necessary so that marketers could keep up with demand.
“Our goal has always been to meet Samsung customers where they are, and offer services that are most convenient for them,” said Scott Messina, Director of Search and Design Strategy at Samsung.
This meant updating their support site experience during the pandemic, when many customers were working from home and using their cell phones more than usual. Traffic to the website exploded.
By taking this step, not only did Samsung avoid alienating users who needed help — they also managed to elevate CX and achieved a 19% increase in customer engagement on the support site.
Responding to the soaring volume of traffic
“While our primary focus is usually SEO, ensuring customers googling for answers find our content first, we knew we needed a customer experience to match once they hit our site,” said Messina.
In the first weeks of pandemic lockdowns, customers were finding Samsung’s support site. The problem for Samsung was what to do with customers when they arrived in such high volume. Samsung was using Elasticsearch, an open and free search and analytics platform, to power their help destination. Managing a site of approximately 2,500 help articles became too labor-intensive for Messina’s small team. They also found that the search tool didn’t always provide relevant results.
In response to the increased demand, they implemented Yext Support Answers, a tool introduced earlier this year by AI-powered search company Yext, which specializes in other webpage search applications.
According to Messina, Samsung also moved to a new chat platform, rolled out several big enhancements to their chatbot, and boosted offline customer service with “We Come to You”, which sends a technician in a van to the home for phone repairs.
Improving the help experience
The Samsung content team had created a vast library of articles, videos, FAQs and other assets. But they needed to be able to deliver these with an effective search tool. Additionally, search results themselves have to be intuitive and helpful. For some queries, the Yext solution uses Extractive QA, a natural language processing algorithm, to deliver a simple answer without having to send the user to another page.
If this improves the search experience and makes it more “Google-like,” then customers are less likely to feel like they have to go to Google first. They can go directly to Samsung for help. And this, in turn, leads to more interactions and data on Samsung’s site, which the company can use to further improve its experience. And it sets up the content team for success.
“Our ability to react to customer needs has certainly improved,” said Messina. “A big part of our content management process is collecting all the signals customers give us, making sense of them, and then addressing content gaps. So, in addition to data sources like the call center and surveys on the site, this richer search data and our increased ability to make sense of it has hopefully decreased the amount of time and energy we spend fixing issues and filling content gaps.”
Tracking performance metrics
Eleven weeks after the launch of the Yext Support Answers tool, Samsung saw issues resolved increase by 15%. Customers also wanted to tell Samsung about this improved help experience. Surveys were completed eight times more frequently following the launch.
For broader metrics, CSAT was up 33%, and NPS improved 45%.
“All of these programs have aligned and use similar goals, the main one of which has been improving our Net Promoter Score,” said Messina. “Our thinking is, if we encourage more and more people to recommend Samsung to their friends and family, the rest of the traditional performance metrics will take care of themselves.”
Solving for complexity with AI
In the online help game, a small percentage of customers need special attention. With a well-known brand like Samsung, they need to remain on top of all trending service issues.
“We usually have a pretty typical 80/20 split, where 80% of the service volume makes up only 20% of the unique issues we see,” Messina explained. “There are some typical ones, like customers struggling to get their product to turn on, or looking to get their cracked screen replaced, and it’s rare that a new, unique issue crops up in enough volume to catch us off guard. That said, whenever we have a new or trending issue, search data is often the first place we look.”
A robust AI engine makes sure that no new issues fall through the cracks. And this becomes even more important when a global event like the pandemic changes everybody’s lives and creates a lot of new problems.
“Customer support teams were inundated with new and complex questions related to the pandemic and were unable to keep up with the sheer volume of tickets, holding consumers up from getting the answers they needed,” said Joe Jorczak, Yext’s Head of Industry for Service and Support. “As customer support requests grow even more complex in nature, businesses need to equip themselves with the tools to efficiently and effectively handle them and deliver the experience that consumers expect today.”
He added, “One of the most valuable parts of our AI-powered Support Answers solution is that, unlike on Google, brands can actually see the search queries that are coming through their website, as well as the answers that are being surfaced for them. For brands like Samsung that want to make their customers’ online journeys as seamless as possible, that level of business intelligence — being able to identify trending issues or opportunities to deliver better answers to popular questions — is a game-changer. It’s the key to informing their service strategy and improving customer satisfaction.”
The customer journey if often routed through Google and there’s not much brands can do about that. But by improving the parts of the experience that they own, they can help make their customers happier — while learning more from data they don’t have to share with a walled garden.
The post How Samsung transformed their customer experience with a new support site search appeared first on MarTech.MarTech