Frame AI Blog

When Things Go Well…

Think of some of your favorite companies. What keeps you coming back?

Big shout out to Chewy. I recently changed my shipping address but then saw an email confirmation indicating my auto-ship order for my dog Chief’s food was en route to my old address. A quick chat with Ross later, and another bag was on its way to me at no extra charge or hassle. 

When Things Go Poorly…

Now think about companies you’ve turned your back on. 

I experienced a major service failure with an event software company ( I won’t name names). Days before a big event, calendar invites hadn’t been automatically emailed to RSVPs as expected. I knew this was going to hurt our live attendance rate. 

The stakes for customer experience are only getting higher. According to Zendesk, 61% of customers would now defect to a competitor after just one bad experience. That’s a 22% jump from the previous year. Make it two negative experiences, and 76% of customers are ready to say goodbye.

For many companies, Support interactions represent a precious few opportunities to develop a human connection with customers - and it’s critical to set the right tone with what you say and do.

We recently shared 5 data-backed tips for what to say in Support interactions. Here, we highlight 5 everyday actions Support agents take that can make a world of difference. We analyzed millions of Support cases to examine their impact on everything from customer satisfaction to case re-open rates. Here are 5 things that customer whisperers *do. *

For many businesses, Support interactions represent the only human interaction they have with their customers.

According to Zendesk, 78% of senior company leaders agree that support agents play a vital role in customer retention. Needless to say, it’s critical to make the right impression in these interactions and put your best foot forward (vs. foot-in-mouth.)

Here, we’ve made some observations about what to say that will help you crush it no matter what kinds of cases get thrown your way based on the millions of Support cases we analyze.

On paper, you did everything right. So why did that case get….(insert any and every negative outcome you weren’t expecting)….escalated, re-opened, low customer satisfaction and effort scores, etc…

Even if you’re following a Support playbook to the letter, the devil is often in the details when it comes to following “best practices.” Here are five common areas where Support team playbooks miss the mark when trying to improve customer experience. 

CX teams measure team effort to understand and manage their overall cost of serving customers and to cultivate happy and engaged teams that play a starring role in great customer experiences. Team effort can also make a huge difference in popular metrics like employee promoter scores (“EPS”), quality assurance (“QA”), and operational metrics like case resolution times and escalation rates. When customer-facing employees’ jobs are harder than they need to be, CX teams often see high turnover rates and low EPS scores, lower QA scores, more escalations, and delays in case resolution times — the list goes on. And each of these unfortunate side effects increases your cost of service and creates stubborn obstacles that stand in the way of improving customer experience.

This is the final post in a three-part series about assembling your customer feedback stack so that you can take action faster and more effectively than ever before. In case you missed it, catch up on part I here and part II here .

Too often, when organizations think about getting started with a traditional CX program, things generally go one of two ways:

Highlights:

• Fastly is an edge cloud platform provider focused on empowering developers to build superior online experiences by moving data and applications closer to users, making websites and applications higher performing and more secure. Fastly stands behind the best of the web with a diverse and innovative set of global customers. Fastly uses Frame AI to reduce customer effort by over 20%, gain insight into customer sentiment across 100% of exchanges, and prioritize urgent account activities.
• All customer communication is represented in one unified CX view, resulting in better, faster business decisions and increased ROI from CX improvements.

Before implementing Frame AI, edge cloud platform provider Fastly relied on a time-consuming manual process to surface the insight they needed. Most of their data came from customer support interactions, which were split across Zendesk and customer Slack channels. “We needed a view that would unify insights across both channels and our Salesforce CRM. Without that, we had limited visibility,” says Jill Kahn, Senior Director of Customer Support Engineering.

Supporting customers within individual support channels resulted in siloed metrics and reporting. Zendesk offered operational reporting like high-level volume trends. But every insight into what the tickets were about, and what Fastly could do to improve their surrounding processes, came from manual tagging and analysis.

Surveys Alone Were Not the Answer

Fastly knew that it could improve the customer experience by investing in CX, but the team was still left with one big question: What could we proactively do to drive a better Customer Experience?

Fastly implemented surveys to dig deeper into its customers’ needs, but doing so added as many questions as it did answers. “Every second of performance and usability is critical to our customers’ success, so it’s imperative that we know what needs our attention immediately - we simply cannot wait for survey results to roll in, and to spend the time digging around to see what’s most urgent. Frame AI saves us an enormous amount of time and execution risk in prioritizing our attention around what’s happening at all times,” says Jill Kahn.

From an AI Skeptic to a Frame AI Believer

Like many companies, Fastly saw great potential in applying AI, but the team wanted to be practical about adding new tools and shiny software. How would AI adapt to their existing set-up and processes?

Frame AI’s pragmatic approach – both with cleaning and aligning B2B customer data, and also providing actionable suggestions where the team already worked – gave them a way to move forward with confidence.

“When we adopted Frame AI, we were looking to extend processes and metrics that were already in place. It was essential that a new solution work well with what we already had,” added Khan . “Frame AI’s customizable approach helped us set clear goals and expectations to be successful.”

Ultimately, Fastly chose to implement Frame AI because it provides valuable customer insights, while maintaining a high standard of efficiency, contributing to a 25% increase in tickets resolved per agent per day.

How Fastly Uses Frame AI to Improve Customer Experience

Fastly’s first step was unifying data from Zendesk, Slack, Salesforce, CSAT and NPS data in Frame AI, to get a unified view of their end-to-end customer experience. By monitoring all communication channels at once, Frame AI helps Fastly bring rapid attention to urgent issues.

“Frame AI produces a daily checklist for our executive team that directs our attention to open cases needing oversight while also highlighting themes across multiple accounts and single accounts needing extra attention,” explained Kahn, “This overview impacts key metrics, including resolution rates and satisfaction, that ultimately drive customer happiness, retention, and growth.”

Before implementing Frame AI, Fastly measured customer wait times by manually combining wait times during a customer call. Over the first six months of implementation, Jill’s team ** reduced customer effort by decreasing wait times by more than 20%** with the help of Frame AI’s insights.

Customer Insights from 100% of Conversations, in Real-Time

Where surveys were only providing Fastly with insight into a small percentage of its customers’ sentiments, Frame AI has expanded Fastly’s Voice of Customer analysis to include 100% of customer exchanges.

Fastly’s customer-focused executive team uses Frame AI’s quick link views to understand sentiment across different channels. This insight has allowed them to parse out the performance of specific channels and fine-tune their support capabilities.

Act Quickly on Opportunities

Frame AI helps Fastly identify and address issues early on, improving the overall time to resolution.Kahn’s team receives alerts for predicted emergencies based on AI-driven logic that Frame AI has set up across multiple data sources. This means customer teams can step in as soon as there is a known issue, instead of waiting for multiple exchanges on a potential emergency.

Each alert generated by Frame AI includes context that enables prioritization and rapid response, such as account size, location and product configuration. “We can instantly identify accounts and themes that need attention, improving the overall customer experience,” said Kahn.

Find the “Why” in Voice of Customer

Altogether, Frame AI has made reporting on Voice of Customer trends far more actionable and specific, with far less manual effort. Fastly uses the intelligence that Frame’s proprietary AI provides in three ways:

• Feeding data back to the executive team to advocate for customers at the strategic level
• Streamlining how tickets are handled from a process level, so that customers reach resolution faster
• Directly empowering Support Engineers to identify and rapidly solve urgent and unusual issues

“Frame AI’s ability to unify customer voice over our helpdesk, CRM, team chat, and survey data has helped us surface new insights and maintain a unified view of individual accounts, operations, and segments,” added Kim Ogletree, SVP of Client Services. “We’ve invested heavily in instrumenting those operations, but Frame AI makes it so much easier to find the ‘why’ that is driving certain behaviors and metrics, allowing us to act quickly.”