Pandemic Darlings The pandemic economy, in original documents
Home Court filings Bofa Ca Unemployment In re: Bank of America California Unemployment Benefits Litigation — S.D. Cal., No. 21-md-02992 Exhibit 119 — In re Bank of America California Unemployment Benefits Litigation (Dkt. 350-120, S.D. Cal. No. 3:21-md-02992)

Court filing

Exhibit 119 — In re Bank of America California Unemployment Benefits Litigation (Dkt. 350-120, S.D. Cal. No. 3:21-md-02992)

Filed October 24, 2024 in In re Bank of America California Unemployment Benefits Litigation; one of 1415 filings from this case.

Record facts

CourtU.S. District Court for the Southern District of California
Filed2024-10-24

U.S. District Court for the Southern District of California · No. 3:21-md-02992-GPC-MSB · Doc. 350-120 · 2024-10-24 · Docket on CourtListener

Full text

EXHIBIT 119 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12049 
Page 1 of 34

The US Contact Center 
Decision-Makers’ Guide 2021
The Interaction Analytics Chapter
Sponsored by
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12050 
Page 2 of 34

In your fast-paced contact center environment, analyzing
every interaction in order to optimize call outcomes, 
customer experience and revenue generation could be a 
challenge. But CallMiner interaction analytics makes it easy. 
Gain an unparalleled view into conversations, leveraging 
AI to help guide agents in realtime DURING their calls, or 
to coach them post-call on the best practices that work at 
YOUR organization.
CallMiner leverages AI and 
machine learning to deliver 
interaction analytics that produce 
enterprise-wide ROI.
Take a FREE CallMiner analytics test drive starting 
with your own phone calls!
Visit us at: http://callminer.com/free-speech-analytics/
•	 Identify techniques/agents that produce the best outcomes
•	 Uncover insights to inform product, marketing, sales
•	 Maintain compliance through 100% monitoring
•	 Enhance employee experience with real-time agent guidance
Leverage interaction analytics to optimize 
performance across the enterprise.
Intelligence from Customer Interactions
Download with our 
Compliments
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12051 
Page 3 of 34

 
 
 
 
 
 
 
 
3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
“The 2021 US Contact Center Decision-Makers’ Guide (13th edition)” 
© ContactBabel 2021 
Please note that all information is believed correct at the time of publication, but ContactBabel does not 
accept responsibility for any action arising from errors or omissions within the report, links to external 
websites or other third-party content. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12052 
Page 4 of 34

 
 
 
 
CallMiner is a recognized leader in the speech analytics software industry, harvesting key 
customer and operational insights from multi-channel customer interactions.  Uniting with our 
customers and partners, our platform drives contact center efficiency, exceptional customer 
and employee experience and significant improvements in top and bottom-line corporate 
performance.   
CallMiner Eureka offers both real-time monitoring and post-call analytics, delivering actionable 
insights to contact center staff, business analysts, and executives. The results include improved 
agent performance, sales, operational efficiency, customer experience, and regulatory 
compliance.  
With over 2 trillion words analyzed annually, CallMiner serves some of the world’s largest call 
centers, delivering highly effective, usable, and scalable customer engagement analytics 
solutions.  
Highlighted by multiple customer achievement awards, including six Speech Technology 
Magazine’s Reader’s Choice Awards, CallMiner has consistently ranked number one in 
customer satisfaction.  
 
Learn more about our customer engagement and speech analytics solutions to 
help your business: 
Web: CallMiner.com  
 
Social Media: LinkedIn, Twitter, Facebook, YouTube, Blog  
 
User Community: EngagementOptimization.com  
 
Email: marketing@callminer.com  
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12053 
Page 5 of 34

 
 
 
 
 
 
 
 
5 
INTERACTION ANALYTICS 
On first glance, customer interaction analytics can be seen as providing similar information to 
management information and reporting systems: taking masses of data and making sense of what they 
mean to the contact center's performance and perhaps even inside the wider business. However, the 
vital thing to understand about analytics is that it gives contact centers the answer to 'Why?', not just 
'What?'. Why are average handle times so different across agents? Why are customers of this product 
upset? Why are people calling the contact center?  
Customer interaction analytics solutions offer huge opportunities to gain business insight, improve 
operational efficiency and develop agent performance. In fact, the list of potential applications for this 
technology is so high that businesses could be forgiven for being confused about how to target and 
quantify the potential business gains. 
Depending on the type of business, the issues being faced and even the type of technology being 
implemented, drivers, inhibitors and return on investment can differ greatly. While an analytics solution 
may be implemented to look at one particular pressing issue, such as automating the QA process, it will 
further develop over time into looking at business intelligence and process optimization. 
Interaction analytics can be used in many different ways to address various business issues. This is an 
advantage – it is hugely flexible – but it can also make its message to the market more complicated. 
However, depending upon how interaction analytics is used, it can assist in: 
• 
agent improvement and quality assurance 
• 
business process optimization 
• 
avoidance of litigation and fines 
• 
customer satisfaction and experience improvements 
• 
increases in revenue and profitability 
• 
improvements in contact center operational performance, and cost reduction.  
Like most contact center applications, analytics can be used to cut costs, but its promise goes far beyond 
this. No other contact center technology provides the business with this level of potential insight that 
goes far beyond the boundaries of the contact center, and can offer genuine and quantifiable ways in 
which sub-optimal business processes can be improved.  
This is not to say that the science of customer contact analytics is yet at its zenith. Significant 
improvements are still being made to the accuracy and speed of the speech engines, the sophistication 
of analytical capabilities, the integration of various data inputs and the usability of report. The 
integration of sophisticated AI and machine learning capabilities within the analytics solutions offers the 
chance to take analytics far beyond what was imagined a few years ago.  
Some of the actionable findings from analytics may seem very simple – the recommendation to change 
a few words in a script, for example – but the overall potential impact upon the cost, revenue, agent 
capability and customer experience that is possible through analytics is perhaps unprecedented. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12054 
Page 6 of 34

 
 
 
 
 
 
 
 
6 
There are various elements to customer contact analytics solutions, including: 
• 
Speech engine: a software program that recognizes speech and converts it into data (either 
phonemes – the sounds that go to make up words – or as a text transcription, although there are 
solutions which directly recognize entire spoken phrases and categorize calls based upon the 
occurrence of those phrases) 
• 
Indexing layer: a software layer that improves and indexes the output from the speech engine in 
order to make it searchable  
• 
Query-and-search user interface: the desktop application where users interact with the analytics 
software, defining their requirements and carrying out searches on the indexed data 
• 
Reporting applications: the presentation layer of analytics, often in graphical format 
• 
Business applications: provided by vendors, these pre-defined modules look at specific issues such 
as adherence to script, debt collections etc., and provide suggestions on what to look for 
• 
Text analytics: this solution combines the transcription of customer calls with other forms of text 
interactions such as email, web chat and social media. It then uses natural language processing 
models along with statistical models to find patterns 
• 
Desktop data analytics: a solution that gathers metadata from agent desktop and CRM applications 
– for example, account ID, product order history and order value – and tags them to call recordings 
or digital records, enabling deeper insight. 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12055 
Page 7 of 34

 
 
 
 
 
 
 
 
7 
Like any technology, customer contact analytics has its own descriptive language, and some of the more 
common words or phrases someone researching this industry would find include: 
• 
Categorization: the activity of grouping conversations according to user-defined topics, such as 
complaints, billing issues, discussions of specific products, etc. Agent capability can be viewed by 
these categories, suggesting specific training needs as well as identifying any required changes 
to processes. Categorization can be done by the business based on their own experiences and 
requirements, through using vendors’ out-of-the-box categorizations for common analytics use 
cases, or by implementing AI and machine learning to find categories within the business’s data 
• 
Discovery: requiring a transcription-based solution, analytics will seek out phrases and words 
that are showing up in noteworthy patterns, showing how they fit together and how they relate 
to each other, discovering trends automatically 
• 
Metadata: non-audio data, which may be taken from CRM, ACD or agent desktop applications, 
which is tied to audio recordings or other interactions, improving the ability to correlate, 
discover patterns and pinpoint specific types of interaction 
• 
Search: if the analytics user knows what they want to find, the search function can return a list 
of calls with these words or phrases within them. Speech-to-text / transcription applications 
return the sentence or whole interaction so that the user can see the context as to how this has 
been used, offering the opportunity to run text analytics on top of this as well 
• 
Closed-loop analytics: where also known as “closed-loop marketing”, this activity involves 
tracking the entire customer lifecycle (i.e. connecting the initial contact all the way to the sale, 
and into ongoing support and post-sale activity), in order to draw actionable insights about how 
elements of the customer lifecycle impact upon sales success and marketing effectiveness. From 
a perspective more closely focused upon the customer experience, “closed-loop” refers to the 
continued, iterative use of automated alerts, follow-up of issues (e.g. through call-back) to 
support root cause analysis, and the identification and resolution of suboptimal processes.  
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12056 
Page 8 of 34

 
 
 
 
 
 
 
 
8 
DRIVERS FOR CUSTOMER INTERACTION ANALYTICS 
Customer interaction analytics offers huge opportunity to gain business insight, improve operational 
efficiency and develop agent performance. In fact, the list of potential applications for this technology is 
so high that businesses could be forgiven for being confused about how to target and quantify the 
potential business gains. Depending on the type of business, the issues being faced and even the type of 
technology being implemented, drivers, inhibitors and return on investment can differ greatly. While an 
analytics solution will be implemented to look at one particular pressing issue, such as compliance or 
automating the QA process, it will further develop over time into looking at business intelligence, 
process optimization, customer experience improvements and revenue increase. 
There are various ways to segment the uses of analytics, and it may therefore be useful to divide them 
into one of two groups: those that are around solving a specific known problem, and those which are of 
a more strategic, long-term nature, although there is some crossover between the two groups.  
Figure 1: Uses of customer contact analytics 
Problem-solving/issue resolution 
Strategic/long-term 
Compliance with regulations 
Gathering competitive intelligence 
Verbal contracts/repudiation 
Feedback on campaign effectiveness and 
pricing information 
Redaction of card information for PCI purposes 
Understanding the customer journey 
Adherence to script 
Understanding why customers are calling 
Identifying agent training requirements 
Improving contact center performance metrics 
Reducing the cost of QA 
Optimizing multichannel/inter-department 
communication 
Identifying and handling problem calls 
Deepening the power and functionality of the 
workforce optimization suite 
Estimating customer satisfaction and first call 
resolution rates 
Identification and dissemination of best 
practice 
Predictive routing 
Identification and handling of dissatisfied 
customers, and those at high risk of churn 
Real-time monitoring and in-call feedback 
Maximizing profitability by managing customer 
incentives 
One-off discovery/analysis via cloud 
‘Tell-me-why’/root cause analysis 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12057 
Page 9 of 34

 
 
 
 
 
 
 
 
9 
USE OF INTERACTION ANALYTICS 
Compared to recording-based functionality which has penetration rates of over 90% in most sectors, 
interaction analytics (especially of the omnichannel variety) is still to reach its full maturity, although the 
general long-term increase in penetration rates and the enthusiasm shown by contact centers to learn 
more about the subject is very positive. 
The positive correlation between size and penetration rate is very noticeable for interaction analytics, 
which may require significant investments. As importantly, having huge volumes of recorded 
interactions and a large customer base to learn from means that business patterns can be identified 
more accurately, and any improvements reap correspondingly higher rewards.  
Large operations are also more likely to have the budget and resource to use analytics to its potential, 
although there is also a significant level of long-term interest in implementing analytics in the small and 
especially the medium contact center sectors. 
Figure 2: Use of interaction analytics, by contact center size 
  
 
 
 
14%
22%
33%
21%
6%
8%
19%
9%
23%
26%
26%
25%
10%
20%
7%
12%
43%
22%
14%
30%
5%
2%
3%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Small
Medium
Large
Average
Use of interaction analytics, by contact center size
Don't know / NA
No plans to implement
Will implement after 12
months
Will implement within 12
months
Use now, looking to
replace/upgrade
Use now, no plans to
replace/upgrade
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12058 
Page 10 of 34

 
 
 
 
 
 
 
 
10 
Against a virtual ubiquity of call recording, the penetration rates of interaction analytics are much lower: 
31% of this year’s respondents use it now, with a further 36% stating that they have plans for 
implementation.  
Respondents from the TMT and outsourcing sectors report the greatest use of analytics this year, with 
those in the public sector least likely to be doing so once again. It is probable that the use of interaction 
analytics is driven more by contact center size in call volumes than through the requirements of specific 
types of business: many of the public sector contact centers are smaller than average, whereas those in 
outsourcing and TMT are amongst the highest. 
Figure 3: Use of interaction analytics, by vertical market 
 
 
 
 
29%
32%
33%
27%
22%
31%
10%
13%
4%
9%
22%
14%
8%
7%
11%
20%
17%
16%
9%
48%
28%
17%
40%
11%
21%
10%
4%
28%
18%
24%
5%
22%
10%
30%
21%
16%
36%
12%
5%
28%
50%
27%
22%
38%
30%
42%
32%
27%
30%
4%
11%
4%
4%
9%
3%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Technology, Media & Telecoms
Outsourcing & Telemarketing
Manufacturing
Retail & Distribution
Transport & Travel
Services
Insurance
Medical
Finance
Public Sector
Average
Use of interaction analytics, by vertical market
Use now, no plans to replace/upgrade
Use now, looking to replace/upgrade
Will implement within 12 months
Will implement after 12 months
No plans to implement
Don't know / NA
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12059 
Page 11 of 34

 
 
 
 
 
 
 
 
11 
As we might expect, the use of post-call speech analytics – the bulk analysis of call recordings – is the 
most widely used type of interaction analytics functionality. 39% of analytics users have also 
implemented functionality which can analyze the agent desktop activity which is linked to these calls. 
Real-time (or near real-time, i.e. within the call) speech analytics is used by 27% of interaction analytics 
users. 34% of respondents that state that they use multichannel analytics.  
The rise in non-voice interaction volumes has meant that there is an increased requirement to 
understand and analyze the customer journey, and there is some interest being shown in optimizing the 
back office and its processes. 
Figure 4: Use of various interaction analytics functionality (from only those respondents who use analytics) 
Interaction analytics type 
 
% respondents using this 
functionality 
Post-call speech analytics 
63% 
Desktop analytics 
39% 
Multichannel analytics (i.e. email, web chat, social media, etc.) 
34% 
Customer journey analytics 
34% 
Real-time speech analytics 
27% 
Back office analytics 
24% 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12060 
Page 12 of 34

Leveraging Emotion to Improve CX & Elevate 
Contact Center Performance
Positive customer experiences are key to organizational success – and understanding emotion is critical to achieving that. Here are strategies and best 
practices to create emotional connections that drive growth.
Brands have officially entered a new era, one in which customer experience (CX) is the be-all end-all. In order to compete effectively, brands need to elevate 
their approach to CX. The journey starts in the contact center.
The key ingredient for delivering a five-star CX in the contact center is understanding and leveraging emotion across customer interactions. Emotion is a key 
driver for developing strong and lasting customer relationships. In fact, emotion has a bigger impact on brand loyalty and customer retention than ease of 
engagement and CX effectiveness1, and 60% of loyal customers use the same type of emotional language they’d use for family and friends when speaking of 
their favorite brands.2
Organizations must take advantage. Forward-looking organizations are investing in their talent and technology –deploying speech analytics and AI in the 
contact center and leveraging sentiment analysis to help measure customer emotions.
Sentiment analysis helps agents develop an emotional connection in the contact center. Sentiment analysis is the process of understanding how customers 
feel about products, promotions, brands or the interactions they have with your organization. Powered by speech analytics technology, sentiment analysis 
enables companies to gain actionable insight on a customer’s emotions, attitudes and opinions, as well as customer service and agent performance.
Unlike traditional feedback methods, such as post-call surveys, sentiment analysis derived from unsolicited feedback captures 100% of customer interactions 
to paint a clearer picture of how your customers really feel and why. By including analysis of the acoustic measures that accompany what customers say, 
organizations can understand the emotional intensifiers that depict the intent, action and satisfaction.
Understanding emotion isn’t just beneficial to customers – it’s also a driver in attracting and retaining hardworking, long-lasting contact center agents. 
Speech analytics delivers real-time, crucial feedback to agents, equipping them to respond appropriately and improve the outcome of the interaction. 
Companies can monitor agent performance by identifying patterns, and agents are able to self-assess and use the data to help improve and develop the 
skills they need (empathy, politeness, efficiency, etc.) to reach better outcomes.
Speech analytics is also ideal for encouraging the agent performance you want. For example, highly scored interactions can be used for agent bright spot 
analysis with examples to emulate – including the most appropriate sentiment expression or response. And when agents are given appropriate feedback with 
the tools needed to succeed and improve, their confidence is boosted, as is their likelihood of retention.
In today’s landscape, only the emotionally intelligent will survive. Brands can get ahead by equipping their contact centers to understand emotion and extract 
and leverage the data to drive a better, more human-centric CX. AI-powered speech analytics gives contact centers the ability to understand customers 
better, provide improved service, develop higher performing agents, and create stronger customer connection. The bottom-line impact on growth, reputation 
and revenue will speak for itself.
1 Forrester, 2019. The US Customer Experience Index, 2019: Some Small Gains, Widespread Stagnation, No Real Leaders. Retrieved from: https://go.forrester.com/blogs/cx-index-
2019-results/ 
2 Deloitte, 2019. Exploring the value of emotion-driven engagement. Retrieved from: https://www.deloittedigital.com/content/dam/deloittedigital/us/documents/offerings/offerings-
20190521-exploring-the-value-of-emotion-driven-engagement-2.pdf 
Intelligence 
from Customer 
Interactions
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12061 
Page 13 of 34

 
 
 
 
 
 
 
 
13 
POST-CALL ANALYTICS 
Initial implementations of speech analytics solutions were focused upon analyzing large numbers of 
recorded calls, often a long time after the actual event. Many of the original users purchased these 
solutions to assist with compliance and as part of a larger quality assurance system, and these benefits 
have not decreased over time. Being able to analyze 100% of calls automatically can provide high quality 
information for the QA process, giving a fair and accurate reflection of the agent’s performance.  
Post-call speech analytics is vital for business intelligence, performance improvement, QA and 
compliance. As the majority of contact centers have call recording in place, the raw material is already 
available. In fact, the amount of recorded voice data available to most businesses can be overwhelming, 
and post-call speech analytics that analyze 100% of recorded calls is proving hugely valuable.  
It should be noted that some recording environments are still mono rather than stereo, meaning that 
there is no distinction between the caller and the agent except through context. This is a clear 
disadvantage for effective post-call speech analytics, as in order to learn from customer feedback and 
experience, clearly a business needs to know whether it is the customer talking about products, 
processes or competitors, rather than the agent. More recording systems are moving to stereo, and this 
will further improve the accuracy and potential benefit of speech analytics, and some vendors have 
restructured their solution to offer software-based speaker separation for analytics. 
Figure 5: Usefulness of post-call analytics 
 
35%
33%
31%
31%
23%
12%
8%
19%
17%
54%
42%
54%
20%
8%
21%
4%
8%
4%
15%
42%
25%
4%
23%
12%
56%
54%
4%
4%
8%
4%
4%
8%
15%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Flagging instances of non-compliance with regulations or script
Automating / speeding up the quality monitoring process
Gaining insight into customers
Identifying training requirements at an agent level
Identifying improvements to business processes
Influencing future scheduling of staff, or routing of calls
Providing information about competitors
Usefulness of post-call analytics
Very useful
Somewhat useful
Not useful
Do not use analytics for this
Don't know
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12062 
Page 14 of 34

 
 
 
 
 
 
 
 
14 
Analytics is seen as being valuable for flagging instances of non-compliance with regulations or script, 
with 65% of respondents that use analytics for this purpose reporting that it is very useful.  
The automated quantification of agent performance and capabilities, feeding into the training and skills 
upgrades required should be one of the most important outputs for interaction analytics, and a little less 
than half of respondents using analytics for this purpose state that it is very useful. A similar proportion 
indicate that analytics is very useful for speeding up the overall quality monitoring process as well 
through automation. 
27% of analytics users state that it is very useful in identifying improvements to business processes. 
Optimizing processes and gaining actionable insight that can be applied to the customer journey will 
become one of the most important uses of analytics, as users’ sophistication increases and solutions’ 
capabilities are explored more fully. 
There is little enthusiasm around the use of analytics for providing information about their competitors, 
with more than half not using it for this purpose at all. This is a very underused area of analytical usage 
at the moment, and one which we would again expect to see growing significantly in future years. 
A growing proportion of respondents report that analytics helps influence scheduling or routing 
strategies, and as more tightly integrated WFO suites are used we would expect this to continue to 
change for the better.  
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12063 
Page 15 of 34

 
 
 
 
 
 
 
 
15 
REAL-TIME ANALYTICS 
Some solution providers suggest that ‘real-time analytics’ should perhaps be more accurately referred to 
as ‘real-time monitoring and action’. Analysis (“a detailed examination of the elements or structure of 
something1”), refers to the discovery and understanding of patterns in data, and is currently something 
that by definition only happens post-call when all data are fully present. Real-time monitoring on the 
other hand, looks for and recognizes predefined words, phrases and sometimes context, within a 
handful of seconds, giving the business the opportunity to act.  
AI can be trained to understand intent and recognize patterns through immersion in vast quantities of 
historical data, so that when a call is taking place it can draw upon this knowledge and provide advice or 
action that has proven successful previously, moving towards the actual provision of real-time analytics.  
AI assists in real-time speech analytics through applying the results of machine learning that have been 
carried out on large quantities of previously recorded conversations, providing: 
• 
agents with the understanding of where their conversational behavior is falling outside of 
acceptable and previously successful norms (such as speaking to quickly or slowly, or in a 
monotonous fashion) 
• 
an assessment of the meaning of non-verbal cues such as intonation, stress patterns, pauses, 
fluctuations in volume, pitch, timing and tone in order to support sentiment analysis 
• 
understanding the actions and information that have been seen to provide successful outcomes 
in previous similar interactions, and relaying this to the agent within the call.  
For some businesses, real-time analysis is an important and growing part of the armory that they have 
to improve their efficiency and effectiveness. There is potentially a great deal of benefit to be gained 
from understanding automatically what is happening on the call, and in being able to act while 
improvements are still possible, rather than being made aware some time after the call of what has 
happened. 
Real-time analysis can be used in many ways: 
• 
monitoring calls for key words and phrases, which can either be acted upon within the 
conversation, or passed to another department (e.g. Marketing, if the customer indicates 
something relevant to other products or services sold by the company) 
• 
alerting the agent or supervisor if pre-specified words or phrases occur 
• 
offering guidance to the agent on the next best action for them to take, bringing in CRM data 
and knowledge bases to suggest answers to the question being asked, or advice on whether to 
change the tone or speed of the conversation 
• 
escalating calls to a supervisor as appropriate 
 
1 http://www.oxforddictionaries.com/definition/english/analysis  
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12064 
Page 16 of 34

 
 
 
 
 
 
 
 
16 
• 
detecting negative sentiment through instances of talk-over, negative language, obscenities, 
increased speaking volume etc., that can be escalated to a supervisor 
• 
triggering back-office processes and opening agent desktop screens depending on call events. 
For example, the statement of a product name or serial number within the conversation can 
open an agent assistant screen that is relevant to that product 
• 
making sure that all required words and phrases have been used, e.g. in the case of compliance 
or forming a phone-based contract 
• 
suggesting cross-selling or upselling opportunities. 
Many solution providers have worked hard to bring to market new or improved solutions to assist with 
real-time monitoring and alerts, and recognition of key words, phrases, instances of talk-over, emotion 
and sentiment detection, pitch, tone, speed and audibility of language and many other important 
variables can be presented on the agent desktop within the call, triggering business-driven alerts and 
processes if required. Speaker separation and redacted audio output (e.g. stopping sensitive data being 
included in text transcriptions) further add to real-time analytics’ capabilities. 
The speed of real-time analysis is crucial to its success: long delays can mean missed, inappropriate or 
sub-optimal sales opportunities being presented; cancellation alerts can show up too late; compliance 
violations over parts of the script missed-out may occur as the call has already ended. However, it is 
important not to get carried away with real-time analysis, as there is a danger that businesses can get 
too enthusiastic and set alert thresholds far too low. This can result in agents being constantly 
bombarded with cross-selling and upselling offers and/or warnings about customer sentiment or their 
own communication style, so that it becomes a distraction rather than a help.  
The effectiveness of real-time analysis may be boosted by post-call analytics taking place as well. For 
example, by assessing the outcomes of calls where specific cross-selling and upselling approaches were 
identified and presented to agents in real time, analysis can show the most successful approaches 
including the use of specific language, customer type, the order of presented offers and many other 
variables (including metadata from agent desktop applications) in order to fine-tune the approach in the 
future. Additionally, getting calls right first-time obviously impacts positively upon first-call resolution 
rates, and through picking up phrases such as "speak to your supervisor", can escalate calls 
automatically or flag them for further QA.  
Real-time analysis offers a big step up from the traditional, manual call monitoring process, and is 
particularly useful for compliance, debt collection, and for forming legally-binding contracts on the 
phone, where specific terms and phrases must be used and any deviation or absence can be flagged to 
the agent's screen within the call. Finance, telecoms and utilities companies – and indeed, any business 
where telephone-based contracts are important – are particularly interested in this.  
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12065 
Page 17 of 34

 
 
 
 
 
 
 
 
17 
Respondents using real-time analytics report that it is particularly valuable for flagging non-compliance 
with scripts or regulations in real-time, and also in identifying and handling dissatisfied customers more 
effectively.  
While real-time analytics’ ability to identify cross-selling and upselling opportunities is a little less highly 
rated, 69% of respondents that use analytics for this purpose state that it is very useful.  
Figure 6: Usefulness of real-time analytics  
 
 
 
65%
60%
45%
20%
10%
10%
5%
10%
10%
10%
25%
10%
10%
10%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Flagging instances of non-
compliance with regulations or
script
Handling unhappy customers
Identifying cross-sell / up-sell
opportunities
Usefulness of real-time analytics 
Don't know
Do not use
analytics for this
Not useful
Somewhat useful
Very useful
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12066 
Page 18 of 34

 
 
 
 
 
 
 
 
18 
TEXT ANALYTICS 
As with speech analytics, text analytics can be applied historically or in real time. It can be applied to 
interactions between customers and agents (as in the case of email, web chat or social media contact), 
or by looking at customer feedback, whether on the business’s own website or on third-party sites. 
Unlike speech analytics, text analytics does not require a speech recognition engine to identify the 
words being used, but the general principles and opportunities are similar. Much of the data analyzed by 
text analysis is unstructured (i.e. is not found in traditional structured databases), such as emails, web 
chats, message boards, RSS feeds, social media etc. The collection and processing of this data may 
involve evaluating the text for emotion and sentiment, and categorizes the key terms, concepts and 
patterns.  
Historical text analysis is useful for business intelligence, whether about how the company and its 
products are perceived, or the effectiveness of the customer contact operation. It is important to note 
that many uses of historical text analysis work best when they are used shortly after the comment is 
made, rather than weeks or months afterward: an issue that is commented upon by many customers 
may need to be acted upon rapidly. For example, confusion about a marketing message, complaints 
about phone queues, or a case of system failure which prevents customers from buying on a website 
need to be identified and handled as quickly as possible. For longer-term issues, such as gathering 
suggestions on new functionality for a product release, such urgency is less important.  
Most large companies will have formal customer satisfaction and feedback programs, and also will 
monitor third-parties such as TripAdvisor or Yelp, which provide structured data in the form of scores, 
and efforts should be made to identify the most important data sources. Text analytics helps to dig 
deeper into the actual unstructured comments left by customers, which are otherwise very difficult and 
time-consuming to categorize and act upon, especially where there are many thousands of comments. 
Industry-specific vocabularies can be used to identify and understand more of the relevant comments, 
and place them into the correct context. Solutions should also be more sophisticated than simply to 
identify key words or phrases: the sentiment of the whole comment should be considered (for example, 
“loud music” in a shop may be exciting to one customer, but irritating to another). Many comments are 
mixed-sentiment, and may also mix a 5-star review with some more critical comments, which the 
analytics solution will have to take into account: the comments are where the real value is found, with 
both positive and negative insights available to be understood. 
Perhaps the most obvious potential contact center use of AI-enabled text analytics is in handling digital 
enquiries, where web chats generally take far longer than phone calls (due to agent multitasking, and 
typing time) and some email response rates can still be measured in days. As the cost of web chat is 
broadly similar to other channels such as email, voice and social media, there is considerable room for 
increasing efficiencies and lowering costs. Real-time text analytics can be used to assist agents when 
answering emails or handling web chats, or to identify customers at risk based on feedback comments 
they have left, initiating an action aimed at alleviating their problem immediately.  
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12067 
Page 19 of 34

 
 
 
 
 
 
 
 
19 
PREDICTIVE ANALYTICS 
Predictive analytics is a branch of analysis that looks at the nature and characteristics of past 
interactions, either with a specific customer or more widely, in order to identify indicators about the 
nature of a current interaction so as to make recommendations in real-time about how to handle the 
customer.  
For example, a business can retrospectively analyze interactions in order to identify where customers 
have defected from the company or not renewed their contract. Typical indicators may include use of 
the words “unhappy” or “dissatisfied“; customers may have a larger-than-usual volume of calls into the 
contact center; use multiple channels in a very short space of time (if they grow impatient with one 
channel, customers may use another); and mention competitors’ names. After analyzing this, and 
applying it to the customer base, a “propensity to defect” score may be placed against each customer, 
identifying those customers most at risk. Specific routing and scripting strategies may be put in place so 
that when the customer next calls, the chances of a high-quality customer experience using a top agent 
are greater and effective retention strategies are applied. 
AI can be applied across the entire customer journey, including sales, marketing and service, helping 
organizations understand customer behavior, intent and anticipating their next action. For example, an 
AI solution may find a pattern amongst previous customers that they are likely to search for specific 
information at a particular point in their presales journey, and proactively provide this information (or 
an incentive) to the customer before they have even asked for it. AI can also help with customer 
onboarding through predicting which customers are likely to require specific assistance.  
Machine learning will allow AI to go beyond simply what they have been programmed to do, seeking out 
new opportunities and delivering service beyond what has simply been asked of them. Through 
understanding multiple historical customer journeys, AIs will be able to predict the next most-likely 
action of a customer in a particular situation, and proactively engage with them so as to avoid an 
unnecessary inbound interaction, providing a higher level of customer experience and reducing cost to 
serve. 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12068 
Page 20 of 34

 
 
 
 
 
 
 
 
20 
SCREEN/DESKTOP ANALYTICS 
Desktop analytics (also known as screen analytics) allow businesses to record an agent’s desktop in 
order to assist with quality assessments at an agent level, and also to identify areas within systems and 
processes that cause delays within customer interactions.  
Additionally, management can search for examples where agents skipped compulsory screens or 
ignored guidelines around how best to close the sale, in order to maximize future compliance with 
regulation and company procedure. 
Average call duration is a metric that has been measured in contact centers since their very first 
inception. However, businesses have had to rely upon anecdotal information in order to decide whether 
excessively lengthy calls are a factor of agent inexperience or inability to answer the customer’s 
question, or if there is a particular step within the procedure when delays are occurring in an otherwise 
competently-handled call (for example, from a lack of training about a particular area, or a badly 
designed screen layout).  
Desktop analytics can provide information about exactly how long each step with an interaction takes, 
providing management with the insight as to which processes could potentially be automated, and how 
much time (and thus, cost) would be saved. Businesses would also gain insight into how agents actually 
research issues that they cannot immediately answer (for example, do they research the company 
website, a knowledge base or the wider Internet, and if so, which method is the most successful?).  
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12069 
Page 21 of 34

 
 
 
 
 
 
 
 
21 
BACK-OFFICE ANALYTICS 
The back office is the part of the organization that processes activities supporting the rest of the 
business, such as order processing and fulfilment, payment and billing, and account creation and 
maintenance. Much of what the back office does is driven by interactions in the contact center which 
trigger the relevant processes, which the back office then have to deliver upon. ContactBabel research 
has found that around 4 in 5 complaints are actually about failures occurring within back-office 
processes rather than within the contact center itself, so analyzing and improving the back office is in 
the interests of the customer-facing departments as well. 
WFO solution providers are developing applications that can be used in the back offices and branches of 
large organizations as well as their contact centers. Far more employees work in these spaces than in 
the contact center, although many back offices lack the same focus upon efficiency and the tools to 
improve it. With the increased focus on the entire customer journey, back office processes are starting 
to fall within the remit of customer experience professionals, who have the remit to alter and optimize 
any area of the organization that impact upon the customer experience, no longer being restricted to 
the physical environment of the contact center. The industry is likely to see back office and contact 
center workforce management systems being closely integrated, or even working as a single centralized 
function that can track and analyze the effect of different departments and processes on others 
throughout the customer journey. 
The back office has somewhat different requirements to the contact center, and will require different 
functionality, including: 
• 
supporting different metrics and deadlines to those of the contact center 
• 
presence management, needed where there are multiple steps within a process that must be 
carried out by different individuals 
• 
deferred workload and backlog management 
• 
workload allocation based on large batches of work arriving at once, rather than be distributed 
throughout the day such as is found within the contact center 
• 
forecasts built on contact center events and volumes 
• 
different service levels and resource requirement calculations: many back office processes take 
considerably longer than a contact center interaction 
• 
adherence to schedule without data from an ACD and capacity modelling (which includes 
employee skills and resource availability) 
• 
the identification of bottleneck processes. 
The use of desktop analytics and screen recording in the back office means that even non-customer-
facing employees to have their performance measured and optimized in the same way as their front 
office colleagues. 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12070 
Page 22 of 34

 
 
 
 
 
 
 
 
22 
CUSTOMER JOURNEY ANALYTICS 
In the long-term, the use of customer contact analytics will improve the customer journey as many 
business process improvements will be enabled by the complete understanding of what is happening 
each step of the way, whether within the customer interaction cycle, or in one of the other processes 
occurring elsewhere within the organization. 
Businesses that understand the reasons that customers are contacting them are able to staff and train 
agents appropriately, provide feedback on company products and services to relevant departments, and 
identify suitable self-service opportunities. They are also able to understand the various levels of 
customer effort required at each stage within the interaction process.  
While it is impossible to quantify ROI upfront, there is a strong argument that “you don’t know what you 
don’t know”. An individual agent may not notice that a new trend is happening until they receive several 
calls about it, but even if they are proactive, they may not receive that type of call again for several 
hours or even days. Analytics and closed-loop feedback identifies trends across the entire operation as 
they happen, instead of waiting on agents to realize something out of the ordinary is happening. 
However, there is no guarantee what will be found, and few businesses will initially implement analytics 
in the hope that optimizing the customer journey and hopefully gaining insight will save costs and 
increase revenue. Many solution providers comment that early adopters of analytics – who often 
started with compliance and agent quality assurance – are now looking at how they understand sales 
effectiveness, marketing campaigns and process improvements. Longer term, understanding and 
optimizing each part of the customer journey will be a key use of analytics.  
Customer journey analytics aims to gather together the various data sources, channels, triggered 
processes and customer touchpoints involved in the customer interaction in order to optimize the 
overall customer journey. By fully understanding the customer experience, businesses can identify and 
rectify inefficiencies, helping to break down the boundaries and siloes between channels and between 
the front office and the back office. 
Customer journey analytics goes beyond the measurement of individual interactions and touchpoints. 
Sophisticated analytics solutions use data inputs from multiple sources, both structured and 
unstructured, in association with journey maps, which are produced by employees in multiple roles 
within the organization who document how various processes currently work and how they could be 
optimized. This is particularly the case in larger businesses which are increasingly looking at the 
effectiveness of back office processes that can impact upon whether the customer has to contact the 
business multiple times.  
Customer effort and engagement is very dependent upon the effectiveness with which channels work 
together, as well as the level of first-contact resolution. Proactively engaging the customer at the 
appropriate time within the customer journey has an opportunity to reduce the effort required for the 
customer to fulfil their interaction completely. As part of a wider omnichannel engagement, businesses 
must seek to understand how and why customers prefer to engage with them, optimizing the flow of 
information throughout any connected processes and channels so that the organization becomes easy 
to do business with. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12071 
Page 23 of 34

 
 
 
 
 
 
 
 
23 
VOICE OF THE CUSTOMER ANALYTICS 
Customer surveys have been an integral part of most businesses since time immemorial. Recently, there 
has been a great increase in the number of organizations implementing “Voice of the Customer” (VoC) 
programs, increasingly based around large-scale analysis of call recordings, as well as using formal 
surveys of customer experience to offer the customer a chance to feed-back, and the business to learn.  
VoC programs strive to capture customer feedback across multiple channels of engagement (IVR, live 
agent, email, etc.), while enabling closed-loop strategies to support customer retention, employee 
development and omnichannel experience optimization. VoC programs typically trigger alerts with role-
based delivery via the use of text and speech analytics, offer statistical modelling services to pinpoint 
root causes, and digitally track progress and results with case management. 
The definition of what a VoC program includes runs the gamut across vendors from simply sending alerts 
based on key words derived from a survey, to more complete solutions that directly contribute to 
contact center optimization and overall CX improvement.  Examples of more complete VoC program 
features include: 
Closed Loop 
• 
Automated Alerts: as surveys are completed, real-time alerting capabilities will immediately 
identify and inform teams of customers in need, while assigning ownership for follow-up 
• 
Callback Manager: an interactive system that enables callback teams to conduct detailed case 
reviews and disposition follow-up activities for eventual root-cause analysis 
• 
Case Management: root-cause exploration tools enable back-end analysis of the customer’s 
initial concern, enabling operational support teams to proactively uncover, track and mitigate 
systemic problems. 
Coaching 
• 
In-The-Moment Coaching Tools: as surveys are completed, real-time alerting capabilities will 
identify when a frontline employee is in need of immediate coaching intervention 
• 
Performance Ranker: the performance ranker helps managers develop weekly and monthly 
coaching plans by outlining strengths and weaknesses for each employee, while identifying 
opportunities for peer-based knowledge sharing 
• 
Behavior Playbooks: playbooks with scorecards help managers coach to specific behaviors by 
outlining how to best demonstrate each behavior, showcasing best-practice examples and 
suggesting sample role-plays.  
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12072 
Page 24 of 34

 
 
 
 
 
 
 
 
24 
Reporting 
• 
Real-time Insight – text analytics zeros in on key issues from multichannel survey feedback 
• 
Role-based Reporting – define type and frequency of report delivery based on responsibility, 
title, geography and more 
• 
Call Recording – drill-down detail can include IVR and live agent call recording for additional 
insight. 
VoC programs are frequently ongoing engagements with result measured by internal CSAT scores, NPS 
benchmarks and efficiency improvements. VoC surveys discover what the company is doing wrong (and 
right), where improvements can take place, how the company is perceived against its competition and 
how it can improve. It is important to view the survey from the customers’ perspective, rather than 
checking boxes that just relate to internal company metrics, which is self-serving. Surveys should also be 
ongoing, to check whether real improvements are being made after the issues have been identified. 
It is vitally important before beginning to survey customers, that a business: 
• 
Clearly determines the purpose and aims of the survey 
• 
Considers adopting a variety of question types. Scored questions enable a business to produce 
statistically significant and representative data. Free comments allow the gain of real insight into 
customers’ perception of service 
• 
Selects an experienced company to set up and host the survey. Businesses will benefit from 
their expertise and knowledge and avoid potentially costly errors 
• 
Ensures that the survey can be carried out throughout the day, including peak times and 
through different channels, to gain a truer picture of the customer experience 
• 
Makes sure that the results of the survey can be collated and analyzed in a wide variety of ways. 
It is pointless to amass information if it cannot be evaluated and the results disseminated 
usefully 
• 
Has procedures in place to act upon the information that it finds. The survey may have 
uncovered some broken processes in the service which need attention. It will also inevitably 
throw up disgruntled customers whose specific concerns need addressing. In this instance, the 
survey platform should provide some mechanism for alerting and following-up to ensure that 
dissatisfied customers are escalated to the appropriate staff 
• 
Adopts a unified approach across the business to assessing and monitoring customer 
satisfaction. If a business continues to reward agents based on traditional call performance 
metrics, it is merely paying lip service to good service. If agents are rewarded based on customer 
satisfaction ratings, it will increase agent engagement and retention at the same time as 
improving the service it offers to customers. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12073 
Page 25 of 34

 
 
 
 
 
 
 
 
25 
Alongside these direct customer surveys, VoC analytics solutions can also gather insight from recorded 
digital and voice channels. Aggregation of customer surveys and analytical results can identify the root 
cause of any issues identified, and provide actionable insight for changing processes and/or agent 
handling techniques. VoC should be seen as a continuous process, rather than a one-off project, and 
ongoing analysis allows the business to operate a closed-loop system, whereby identified issues can be 
actioned and continuously checked to make sure that the problem does not reoccur. 
Figure 7: Effectiveness of methods for gathering customer insight (where used) 
 
 
 
 
38%
30%
28%
25%
23%
20%
43%
65%
61%
57%
55%
68%
19%
5%
12%
18%
22%
12%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Speech analytics (i.e. of recorded calls)
Customer experience research calls and emails
Formal process for gathering agent comments
Projects studying the customer journey
IVR or SMS (i.e. automated, near real-time surveys)
Meetings with supervisors who pass on agent insights
Effectiveness of methods for gathering customer insight
Very effective
Somewhat effective
Ineffective
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12074 
Page 26 of 34

 
 
 
 
 
 
 
 
26 
The previous chart looks at contact center professionals’ opinions of the effectiveness of each method 
of gathering customer insight.  
Automated analytics solutions get reasonable approval ratings, although IVR/SMS surveys get a mixed 
response. However, these methods are used by less than half of respondents as the following table 
shows.  
Figure 8: Use of methods of gathering customer insights  
Method 
Proportion of respondents using 
this method 
 
Meetings with supervisors who pass on agent insights 
95% 
Formal process for gathering agent comments 
80% 
Customer experience research calls and emails 
71% 
Projects studying the customer journey 
52% 
IVR or SMS (i.e. automated, near real-time surveys) 
49% 
Speech analytics (i.e. of recorded calls) 
48% 
 
 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12075 
Page 27 of 34

 
 
 
 
 
 
 
 
27 
MEASURING THE ROI OF ANALYTICS 
As part of the research for this report, thousands of contact center professionals were asked for their 
views on interaction analytics, particularly about what would hold them back from implementing it. By 
far the most important issue raised was how to build a strong enough return-on-investment (ROI) case 
to get the required corporate buy-in. 
Return on investment for customer interaction analytics can come from numerous sources, depending 
upon how the solution is used. Generally, it will come from the avoidance of a specific cost, (including 
the reduction of a risk in the case of compliance), or the increase in revenue.  
The return on investment of customer interaction analytics used for compliance can at first glance be 
difficult to prove, but it is the avoidance or reduction in litigation and regulatory fines which can be 
placed against the cost of the solution. Large banks will have funds put away running into the tens of 
millions of pounds each year against the possibility of paying out, and any significant reduction in fines 
would pay for a speech analytics solution very quickly. In the UK, the banking industry had put aside 
several billion pounds to pay compensation for the mis-selling of PPI (payment protection insurance), 
and having the ability to prove that no regulations had been broken would have been of great use.  
Most vendors have tools which can be used to estimate return on investment, often based on what they 
have seen in similar operations elsewhere, and they are keen to share them with potential customers. 
Estimates of the time taken for the solution to pay for itself usually vary between 6 and 18 months.  
Variables to be considered for ROI measurements include: 
Cost reduction: 
• 
Reduction in headcount from automation of call monitoring and compliance checking 
• 
Understanding and minimizing the parts of the call which do not add value  
• 
Avoidance of fines and damages for non-compliance 
• 
Reduction in cost of unnecessary callbacks after improving first-call resolution rates through 
root cause analysis  
• 
Avoidance of live calls that can be handled by better IVR or website self-service 
• 
Reduced cost of QA and QM  
• 
Understand customer intent, e.g. an insurance company received a lot of calls after customers 
had bought policies from their website. Analysis was able to show that customers were ringing 
for reassurance that the policy had been started, meaning the company could immediately send 
an email to new customers with their policy details on it, avoiding the majority of these calls 
• 
Lower cost per call through shortened handle times and fewer transfers  
• 
Lower new staff attrition rates and recruitment costs through early identification of specific 
training requirements 
• 
Identifying non-optimized business processes (e.g. a confusing website or a high number of 
callers ringing about delivery) and fix these, avoiding calls and improving revenue. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12076 
Page 28 of 34

 
 
 
 
 
 
 
 
28 
Revenue increase: 
• 
Increase in sales conversion rates and values based on dissemination of best practice across 
agents, monitored by script compliance 
• 
Increase in promise-to-pay ratios (debt collection)  
• 
Optimized marketing messages through instant customer evaluation 
• 
Reduced customer churn through dynamic screen-pop and real-time analytics 
• 
Quicker response to new competitor and pricing information  
• 
Increase sales revenue by automating manual, non-revenue generating activity by identifying 
and improving self-service options 
• 
Route specific customer types to the best available agents to optimize empathy by matching 
communication styles 
• 
Some businesses assign a revenue value to an improvement in customer satisfaction ratings or 
Net Promoter Score® 
• 
Understand and correlate call outcomes, using metadata and call analysis to see what works and 
what doesn’t. 
Also, the improved quality of agents, better complaints handling and improved business processes 
outside the contact center should be considered. 
It is important for the CFO to see the customer data and brand loyalty as assets, and to consider the 
effect that complaints and general dissatisfaction have upon those assets. Analytics helps businesses to 
understand why these assets (i.e. the customer base) may be shrinking over time, and to put actions in 
place to turn that around. In order to get sign off on an analytics project, these benefits must be 
monetized. 
Against these potential positives, costs to consider include: 
• 
License fees or cost per call analyzed 
• 
IT costs to implement (internal and external) 
• 
Upgrade to call recording environment if required 
• 
Bandwidth if hosted offsite: the recording of calls is usually done on a customer's site, so if the 
speech analytics solution is to be hosted, it will involve of lot of bandwidth, which will be an 
additional cost, especially when considering any redundancy 
• 
Maintenance and support agreements, which may be 15-20% annually of the original licensing 
cost 
• 
Additional users – headcount cost – decide who will own and use it, do you need a speech 
analyst, etc. 
• 
Extra hardware e.g. servers  
• 
Ongoing and additional training costs if not included  
• 
Extra work generated by findings 
• 
May need extra software to extract data from the call recording production environment. 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12077 
Page 29 of 34

 
 
 
 
 
 
 
 
29 
Any business case needs to be built with support from the potential end-users, understanding the 
specific key performance indicators that are important to them, rather than focusing on IT specific 
issues. Whatever the variables and factors that businesses choose to build the ROI and business case, it 
is important to gather benchmark data before the solution is deployed, so as to be able to quantify any 
change accurately. If possible, use a ‘control and experiment’ approach : for example, one sales team 
carries on as they were, while the other may have their scripts changed or receive tailored training 
based on analytical insights. It is also important to get business users involved early in the process, 
giving them a key part in defining the right business case and the desired ROI. 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12078 
Page 30 of 34

 
 
 
 
 
 
 
 
30 
DEVELOPING THE USE OF ANALYTICS 
Once the implementation has been made, businesses then need to make sure the solution delivers what 
was promised, and hopefully this initial success will provide a platform for the analytics solution to be 
directed elsewhere. 
Vendors strongly recommend that businesses put baseline measurements in place before any 
implementation takes place, such as how many calls are tagged with a particular issue. The vendor and 
customer implementation team monitor and suggest changes to processes and approaches based on 
findings of the initial analysis, and measurement post-implementation will quantify the cost savings or 
alteration to other key metrics. 
If the initial use of analytics is successful, the business can seize the opportunity to use this enthusiasm 
and positivity to roll analytics into other areas. Analytics can deliver insight which is of use to other parts 
of the business as well as the contact center, and is an opportunity to demonstrate to the rest of the 
business that there is a wealth of information that can be mined to support the decisions that other 
departments have to make. Pointing to examples where customers are changing supplier due to 
superior products from a competitor, or where another business’s marketing campaign is creating a high 
turnover in your customer base will grab the attention of senior decision-makers elsewhere in the 
enterprise. 
To be successful, analytics must be integrated into the existing systems, processes and structure. 
Embedding it within the overall culture of the wider business is perhaps the surest way of ensuring 
success. At a contact center level, connecting analytics output with the quality management process 
means that the operation can find a place for analytics within their world, which will encourage them to 
consider it for business intelligence purposes later on. Businesses may also wish to consider solutions 
where analytics output is shown automatically across the organization, sharing dynamic reports and 
graphics on a regular or exceptional basis to business owners elsewhere in the enterprise. 
Although every user’s requirements from analytics will be different in some way, it may be useful to 
consider looking for some of the following key words and phrases: 
• 
names of competitors 
• 
obscenity or profanity 
• 
names of your specific products or services 
• 
references to management (e.g. “supervisor” or “manager”) as this may indicate the customer is 
dissatisfied with the agent 
• 
active opinion (e.g. “it would be good if”, “I would like”, “I want”) 
• 
key commercial words (e.g. “buy”, “purchase”, “interested in”) 
• 
phrases which indicate compliance, such as those found in the terms and conditions 
• 
customer dissatisfaction (e.g. “I’m not happy”, “I want to close my account”) 
• 
references to the agent’s performance (e.g. “you’ve been really helpful”, “rude”). 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12079 
Page 31 of 34

 
 
 
 
 
 
 
 
31 
Two examples of interesting, value-add opportunities that analytics provides are root cause analysis, 
and discovery. 
 
'Tell-me-why' or root-cause analysis 
Tell-me-why is a starting point for analysis. A business which knows it has a problem with its web self-
service function can find out more about the problem through automated analysis of calls, rather than 
through asking agents directly or listening to recordings. Inputting 'website', 'web' or similar, searches 
the index of words or phrases and returns likely calls. Speech-to-text-based systems can search for other 
words in the conversation that occur frequently (without the need for users to predefine these searches 
in advance), and group them together into categories, rated by relevance, importance of words etc. (e.g. 
if 'website' and 'password' occur together far more frequently the usual, this is probably an area to 
explore further). The use of speaker separation – whether through having dual channels or using 
software-based algorithms – means that the system can differentiate the customer from the agent, 
giving a greater accuracy of results.  
 
Discovery 
'Discovery' is a term often used within the customer contact analytics industry, and refers to a deep, 
automated analysis of trends, patterns and results which are identified by the speech analytics solution 
rather than the knowledge or insight of the human operators. Discovery will help users to find calls that 
are similar to each other, perhaps through similar groupings of words or phrases, and explore these links 
to discover the issues driving them. Many solutions offer automated discovery and this is an area that 
will always be improving and becoming more subtle and effective, having huge potential benefits for 
businesses.  
The ability to see trends – to know that the instances of the words 'website' and 'password' have 
increased by 2,000% this week compared to the norms of the past 6 months - quickly identifies likely 
pain points for the customer and potential broken processes. The continual tracking and analysis of 
similar information or categories over time also allows a business to see whether the remedial action 
that they put into place has actually worked.  
Of course, any analysis where the direct beneficiary is not the contact center must be properly aligned 
to the organization’s objectives and strategy, encouraging changes to be made to areas that have 
already been earmarked as needing improvement. Otherwise, if the focus is not aligned with strategic 
goals, information merely becomes ‘nice to know’, rather than actionable. 
Customer interaction analytics has the ability to tear down the virtual wall between the contact center 
and other areas of the business, meaning that the business intelligence extracted can be shared and 
valued by parts of the organization that otherwise have little to do with the contact center. With the 
historical and ongoing difficulty in getting the business to value the customer contact operation fully, 
this can only be a good thing politically.  
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12080 
Page 32 of 34

 
 
 
 
 
 
 
 
32 
Some real-life examples of where analytics has delivered improvements include: 
• 
an insurer improved first call resolution by over 6 percentage points by understanding and 
correcting how agents respond to specific types of denied claims issues  
• 
Identifying the types of low-to-medium complexity calls that could be handled less expensively 
but still effectively via self-service channels.  The result can be either reduced headcount or 
extended service hours 
• 
improved sales conversions by 41% and collections revenue by 20% by identifying the skills that 
differentiated top performing agents from bottom performing agents, and then focusing training 
and coaching programs on those key skills  
• 
analyzing and fixing back office processes that were generating unnecessary repeat calls and 
driving poor customer satisfaction 
• 
highlighting the five key customer queries and developing FAQs for agents, which significantly 
reduced average handle time on these calls 
• 
reducing call volume by 2% by identifying and fixing issues with the password reset process  
• 
identifying opportunities in verbatim customer feedback to address specific customer segment 
needs, increasing sales by 30% the following year 
• 
categorizing all customer calls by reason for the call and any subtopics, measuring agent 
performance (handle time, customer satisfaction rating, and issue resolution) by call type. 
Identified the type of calls that had excessively high handle time due to sub optimal customer 
identity verification, and improved coaching and training decreased handle time by an average 
of 36 seconds, saving $5 million per year  
• 
determining that 57% of calls could be handled through a self-service web portal, but the 
customers were not aware that they could do this online 
• 
quality program was transformed by providing targeted data on the major reasons for customer 
dissatisfaction 
• 
discovering that only 2% of calls taken at night were critical, reducing headcount on the night 
shift 
• 
reducing QA headcount from 40 agents to fewer than 10 by implementing automated scoring on 
100% of calls. 
 
For more information about interaction analytics, please download ContactBabel's free "Inner Circle 
Guide to Customer Interaction Analytics". 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12081 
Page 33 of 34

 
 
 
 
 
 
 
 
33 
ABOUT CONTACTBABEL 
ContactBabel is the contact center industry expert. If you have a question about how the industry works, 
or where it’s heading, the chances are we have the answer.  
The coverage provided by our massive and ongoing primary research projects is matched by our 
experience analyzing the contact center industry. We understand how technology, people and process 
best fit together, and how they will work collectively in the future.  
We help the biggest and most successful vendors develop their contact center strategies and talk to the 
right prospects. We have shown the UK government how the global contact center industry will develop 
and change. We help contact centers compare themselves to their closest competitors so they can 
understand what they are doing well and what needs to improve.  
If you have a question about your company’s place in the contact center industry, perhaps we can help 
you.  
Email: info@contactbabel.com  
Website: www.contactbabel.com    
Telephone: +44 (0)191 271 5269 
 
 
To download the full “2021 US Contact Center Decision-Makers’ Guide”, 
free of charge, please visit www.contactbabel.com 
 
 
 
Case 3:21-md-02992-GPC-MSB     Document 350-120     Filed 10/24/24     PageID.12082 
Page 34 of 34

File and source

File
gov.uscourts.casd.709615.350.120.pdf
Size
4,875,055 bytes
SHA-256
421d12090db952a253804f89137867c16083ce57e189f62f954da8d8446cb7a1
Our copy
gov.uscourts.casd.709615.350.120.pdf
Original
PACER (login required)
Back to top