5 minute read
Introduction
Contact centers are the critical front lines for many businesses—a crucial channel available to assist and support customers, especially when immediate answers are needed. They are one of the most common ways people interact with brands sharing issues, feedback, and frustrations. But they are also extremely difficult to operate with inefficiencies, high turnover, and variability in agent effectiveness. According to Deloitte’s 2019 Global Contact Center Survey:1
90% of companies say the traditional voice channel is used to handle the most complex transactions.
61% believe the complexity of those interactions will increase as self-service options handle simpler inquiries in the future.
Conversations between agents and customers often contain a gold mine of insights—about your products and services as well as the quality of the experiences your agents provide. Extracting these nuggets of information and uncovering specifics on where you should focus attention from thousands, or even millions, of hours of calls is a massive challenge.
More calls, more problems
Transcribing and analyzing all of these calls across many hours has proven to be nearly impossible, at least in a timely enough manner to make any insights actionable. Organizations do their best to understand these customer experiences manually, resorting to sampling a small percentage of random calls.
But that’s not enough. Although Sampling enables fast feedback, it provides a vastly incomplete picture and risks missing critical blind spots. Brands typically focus on metrics like abandonment, average time on hold, time per call interaction, or first-call resolution percentage. While these are useful data points, they miss the context and conversational details between agents and customers that could yield actionable insights.
of 700 global contact center executives surveyed by Cisco this year cited difficulty analyzing data as one of their top challenges. Beneath the surface of every call is a wealth of experience insights.
Cisco, data survey 2020²
The need for AI-powered speech analytics
Brands need to rethink their approach of mining for customer experience insights by examining every single call. With more agents working remotely and anxious customers looking for immediate answers, contact centers need a scalable, efficient lens to sift through every interaction.
To solve this problem, brands must combine text analysis of call transcripts with voice analysis of every acoustic call recording. From text-based transcripts, sentiment analysis of words across customer calls can be uncovered and surfaced as frequent topics and themes. Voice recordings can codify silence in a call, or enumerate overtalk indicating a customer’s excitement or disappointment.
With both sets of data codified into a data set, artificial intelligence can easily sift through every interaction, and surface insights that were previously hidden. These kinds of analytics are now more important than ever as organizations try to keep pace with changing consumer behavior and fluctuating demand.
Role 1: Contact Center Operations, Heather
The role:
As a leader of the contact center, Heather is in charge of the overall performance and budget of the center. She is focused on the call experience and how efficient and effective agents are.
Key responsibilities:
- Performance of agents, teams, and entire center
- Call quality and customer satisfaction
- People and process improvement
- Costs, budget, and maximizing efficiency
Primary focus:
- Operational efficiency, as measured by contact center KPIs (first-call resolution, wait time, call duration/cost, CSAT)
- Call quality and agent interactions with customers, including how the agent answers the call, how they navigate the caller to a resolution, and how they end the call
How AI-powered speech analytics helps Heather meet organizational objectives:
- Improves accuracy of KPIs by correlating call analytics with the subset of calls that have direct customer feedback from post-transaction surveys and other measurements of call resolution.
- Surfaces top issues and drivers of long or unresolved calls with aggregate-level KPIs and trending topics, themes, and sentiment.
- Improves first-call resolution by identifying areas for improved scripts, response protocols, and more focused training.
- Identifies agents who need additional support and training, as well as top-performing agents who can assist in coaching and mentoring.
When brands are unlocked with AI-powered speech analytics, the experience signals from each call drive actionable insights and unique value for different roles across the enterprise.
of companies report seeing an increase in CSAT over the course of a year when issues are resolved faster and on the first call.
Role 2: Customer Service, Jason
The role:
As a leader in customer service, Jason provides direction to the customer service team across all channels. He is focused on ensuring customers get what they need in the fastest, most efficient way possible.
Key responsibilities:
- Customer satisfaction across all channels
- Identifying areas for training and process improvement
- Increasing speed of response while reducing cost to serve
- Finding ways to increase use of self-service channels
Primary focus:
- Service effectiveness across all channels including contact center, digital, physical locations, and all customer touchpoints
- Increasing self-service channels to reduce cost and time to provide customer service
How AI-powered speech analytics helps Jason meet organizational objectives:
- Surfaces deep and detailed insights into why customers choose the contact center channel, providing opportunities to consider structural changes and process improvements.
- Identifies trending issues in contact center interactions that can be compared to other channels and best practices that can be shared across channels to improve service response and customer satisfaction.
Uncovers self-service opportunities (automated chat topics, help pages and videos on website, etc.) by analyzing recurring questions and topics across the contact center.
greater ability to predict customer satisfaction when companies understand the entire experience versus looking at individual touchpoints.
McKinsey and Co.⁴
Role 3: Customer Experience Insights, Sarah
The role:
As a leader in customer experience insights, Sarah leads the ongoing understanding and improvement of the holistic customer experience across all touchpoints. She is focused on identifying areas to innovate and drive organizational change.
Key responsibilities:
- Analyze how key contact center customer interactions fit into and affect overall customer journeys
- Improve customer experience and drive outcomes that reduce churn and increase loyalty
- Work with managers of contact center and other channels to identify areas for innovation
Primary focus:
- Finding key areas for achieving business objectives through customer experience improvements and operational changes
- Optimizing overall customer journeys to reduce churn and increase customer lifetime value
How AI-powered speech analytics helps Sarah meet organizational objectives:
- Accelerates decisions to improve contact center experiences through more detailed insights, overlaying call analytics with survey feedback from a subset of interactions.
- Reveals potential for innovation when call analytics are integrated with other experience data into a single view of all customer journeys across contact center, physical and digital channels.
- Identifies opportunities where churn can be reduced and when customers are receptive to additional products or offers as part of a contact center interaction.
Up to a 95% increase in profit can be realized when retention rates increase 5%.
Harvard Business School and Bain & Co., research⁵
5 considerations when shopping speech technology
If you are looking at new ways to better understand your contact center interactions at scale and improve the end-to-end customer experience, there is no better solution than AI-powered speech analytics. Here are five things to look for when considering technology offerings:
- Quality transcription means quality data. Quality should not be defined by word error rate alone. Maintaining sentence context is paramount as is insertion of punctuation and grammar to allow you to get to the why behind topics. Having speaker separation is also important to allow for call analysis at call and aggregate levels. Transcription quality should be great out of the box and your partner should provide you multi-language support along with pre-defined topic lexicons.
- The power to understand and predict. Artificial intelligence can help you quickly understand data in ways far deeper than just what topics are trending. In addition to dialogue transcription and text analytics, solutions should also provide acoustic analytics that detect silent time, overtalk, or vocal emotion. This enables you to see what topics are driving the most negative or positive customer impact and helps agents learn to solve customer pain in ways that increase satisfaction.
- Automate analysis and empower your teams. Look at call analytics from the lens of quality assurance: If you can quickly pinpoint call topics and specific agents causing customer confusion or dissatisfaction you can create training programs or process improvements. But you should also consider how easily these insights can be shared with each employee. By giving frontline agents their own dashboard, you can empower them to self-coach and improve areas of poor performance, especially crucial in remote working environments.
- Future proof your business. Think about your needs today, but look ahead and stay agile in your approach. Avoid the per-seat model that doesn’t provide flexibility as you evolve your workforce mix of shifts and agents. Also, make sure your technology partner has an agile and significant roadmap that can grow with you.
- Value realization. Work with your partner up front to make sure the solution is easy to use, intuitive, and quick to deploy. Gone are the days of waiting 9 to 12 months to get an on-premise solution set up then another 90 days to fine tune before getting accurate results. Ask your partner what its current user engagement numbers look like and be sure your license gives every agent access to custom dashboards. Also confirm that you have the ability to import call transcriptions into other solutions, which will add value to your existing tech stack and create a seamless workflow for visibility of action.
Gold mine of insights to spur action
While speech technology has been available in some form for nearly two decades, it never fully delivered on its promise due to the complexity and cost to implement. Speech tools to date have also had to balance speed and scale—as in the number of calls transcribed at a time or over a day—with accuracy. The more calls they were able to transcribe quickly, the less accurate the transcriptions and analysis became.
With the launch of Medallia Speech, the signals in every call can now be unlocked and made actionable. This scalable, accurate, and rapid solution enables organizations to quickly get the right insights to the right people across the enterprise.
- Create better customer experiences based on transcript-level text analysis that uncovers call reason, themes and topics, customer effort, and actionable suggestions.
- Improve agent coaching and training with AI-driven insights into acoustics like silence time, overtalk, and agent/client emotion.
- Ensure accurate reporting for contact center and customer experience leaders by processing every voice interaction with automatic scaling and best-in-class transcription accuracy.
- Reduce churn and discover self-service opportunities by combining voice interactions with feedback and data analytics across all your channels for a complete, rich view of each customer’s journey with your business.
With AI-powered speech analytics, organizations can drive efficiencies and performance within the call center, but also layer those signals into Medallia Experience Cloud for a unified view of the end-to-end customer experience across every channel.
Conclusion
The modern contact center requires not only an agile mindset, but powerful and flexible solutions that can scale with an organization. Without a scalable platform, humans can only listen to and analyze so many calls and small samples may not capture the most pressing issues fast enough.
AI-powered speech analytics provide organizations the ability to understand financial impact, increase efficiency, train agents better, and improve the overall customer experience.
- For agents and contact center managers, fast and accurate analysis of every call helps surface areas to improve call quality and operational efficiency.
- For customer service leaders, detailed insights into the center can be compared across channels to identify areas to drive faster resolution through self service.
- For customer experience insights leaders, a unified view of the end-to-end customer journey helps identify ways to drive different outcomes that increase customer lifetime value.
By unlocking valuable insights from every call, organizations can drive efficiency and performance within the contact center and across the entire business.
