《IBM商业价值研究院:虚拟代理技术的价值(2023)(英文版)(20页).pdf》由会员分享,可在线阅读,更多相关《IBM商业价值研究院:虚拟代理技术的价值(2023)(英文版)(20页).pdf(20页珍藏版)》请在三个皮匠报告上搜索。
1、 Benchmark Insights The value of virtual agent technology Improve customer service and boost financial results with AI-enabled systems How IBM Cognitive Care can help World-renowned brands have transformed their contact centers with IBM Cognitive Carea customer engagement center powered by data,AI,a
2、nd hybrid multi-cloud technology that dynamically supports both customers and live agents.IBM Cognitive Care can help your enterprise achieve not only a technology and process transformation,but also a lasting cultural shift in your organization.For more information,visit Joe Petrone,Gillian Orrell,
3、and Carolyn Heller Baird Key takeways Significant financial benefits 99%of study respondents report reduction in cost per contact as a result of using virtual agent technology(VAT).A recent Forrester Consulting study estimated that a large organization that implements VAT can achieve$5.50 cost savin
4、gs per contained conversation.1 Our analysis reveals an average containment rate of 64%with a 38 percentage-point swing from highest to lowest.94%of VAT Leaders have already achieved or exceeded their business case.Satisfaction scores increase Some organizations fear VAT could cause significant decr
5、eases in customer and employee satisfaction.However,we found average improvements of 8 and 7 percentage points in customer and agent satisfaction,respectively.Average human agent handle time dropped by 12%.VAT Leaders achieve even more:average improvements of 12 and 9 percentage points in customer a
6、nd agent satisfaction and a 15%reduction in handle time.Leaders implemented early,integrated VAT with backend systems,and trained it on high volumes of contacts.Rising to the challenge When COVID-19 struck,organizations with mature VAT capabilities rapidly expanded operations to support business con
7、tinuity.The majority of these organizations saw an increase in customer satisfaction.Far beyond experimentation Virtual agent technology(VAT)is having its moment.If youve recently called a companys customer care number or conducted an online chat about a service,product,or payment,there is a good ch
8、ance you were interacting with virtual agent technology.VATs popularity is surging because of the technologys seeming ability to do everything:better serve customers,enhance employee experience,reduce costs,and increase revenue.The COVID-19 pandemic has rocketed VAT adoption into hyperdrive.For emer
9、gency situations,when speed and simplicity are mission critical,out-of-the-box implementations can go live in a matter of hours.However,companies that thoughtfully design,develop,and deploy VAT have the potential to unlock even more value for their organizations and their users,be they customers,emp
10、loyees,or human agents.Until now,actual data about VAT performance has been limited to individual case studies,making it difficult for organizations to derive meaningful comparisons.As we embarked on this research,we knew organizations needed answers to four key questions:How efficient is AI-enabled
11、 VAT?What are the key attributes and practices that drive value from VAT?What is the impact of AI-enabled VAT on customers,end users,and human agents?What impact does VAT have on organizations financial performance?1 99%of respondents report an increase in customer satisfaction as a result of using
12、virtual agent technology.20%is the average contribution to human agent satisfaction attributed to virtual agent technology.96%exceeded,achieved,or expect to achieve their anticipated return on investment for their virtual agent technology implementation.Our analysis revealed that VAT is having a pos
13、itive impact on nearly every measure,including customer satisfaction,employee satisfaction,and revenue,and a remarkably high proportion of respondents(96%)have exceeded,achieved,or expect to achieve their anticipated ROI.Armed with detailed metrics and benchmarks,decision makers can now assess their
14、 current and proposed VAT initiatives with far greater specificity and relevancy.They can use this information to determine the right enhancements for their customer care or service desk solutions,no matter where they may be on their digital journey.Our study approach The IBM Institute for Business
15、Value(IBV),in cooperation with Oxford Economics,surveyed 1,005 respondents,across 12 industries and 33 countries,to gather performance data from organizations employing VAT daily,in live situations,for periods from six months to just over four years.2 We gathered data for three types of VAT implemen
16、tation:1.VAT directly interacting with external customers 2.VATdirectlyinteracting withinternalcustomerssuch as employees,contractors,and suppliers VAT supporting service desk agents,commonly known as“agent assist technology.”(See the Study methodology section on page 15 for more information.)2 Meas
17、uring VAT performance VAT performance is measured by a combination of new metrics,unique to the way the technology works(see Figure 1),and traditional contact center metrics impacted by the technology.1.Intent recognition Intent is the customers purpose when making contact and during the contact,for
18、 example,questioning a charge or asking about a product.Intent is distinct from contact because customers may have more than one intent per contact.VAT is trained to recognize and respond to each intent separately,making it vastly more sophisticated than a traditional keyword search.Customers might
19、say,“How do I settle my account?”when their intentfrom VATs perspectiveis“Pay my bill.”Training VAT to recognize true customer intent behind the many ways customers express their intent is critical for successful implementation and maintenance.The skill required to program VAT to ask good clarificat
20、ion questions should not be underestimated(see“Insight:Conversational user experience(UX)design”on page 4).Average intent recognition for all survey respondents across all three types of VAT implementations is 68%.For situations in which VAT directly interacts with external customers,the average is
21、70%.And for intelligent agent assist implementations,the average intent recognition is 65%.This could be a reflection of organizations making larger investments in VAT systems aimed at external customers.Figure 1 The new contact center performance measures Inbound contacts Intent recognition In-scop
22、e segment Containment 1 2 3 Does the VAT understand the customers needs?Is the VAT trained to meet the customers needs?64%Median 70%Benchmark 64%Median 70%Benchmark Human agent supported by agent-assist VAT Intent recognition 65%Median 75%Benchmark 1 In-scope segment 62%Median 70%Benchmark 2Human ag
23、ent routed VAT directly interacting with customer or end user Does the VAT meet all the customer needs?68%Median 75%Benchmark Source:IBM Institute for Business Value.3 2.In-scope segment Once virtual agent technology recognizes customer intent,it should also know if it has been trained to support th
24、at intent.Across all survey respondents,the average proportion of inbound contacts with intents that fall within the scope of the VAT is 63%.The difference between the lowest and highest in-scope segment is 36 percentage points.3.Containment Not all companies measure containment the same way.Some de
25、fine containment as the percentage of all contacts fully handled by the VAT.This can result in low containment percentages when the VAT does not yetor will never handle certain types of contact.In this study,containment is the portion of total contacts the VAT has been trained to handle that it reso
26、lves without any escalation toor involvement froma human agent.Measuring containment at contact level is pragmatic,though it can under-report the achievement of VAT when somebut not allof a customers intents are resolved by VAT.The use of sophisticated and automated analytics to measure containment
27、fairly is important for future development.Average containment across relevant respondents is 64%The difference between the lowest and highest reported containment is 38 percentage points.This is a wide range,and the financial implications can be significant.Every contact contained by VAT means no h
28、uman agent time.Based on IBV analysis of human handle time across all contacts,an average of four minutes of human agent handle time is saved when a contact is contained by VAT.Performance for containment is strongly linked with machine learning and systems integration factors(see“North American gro
29、cery chain:Outsourcing to AI”on page 6).For example,66%of respondents in the top quartile for containment report that their VAT can execute complex actions on behalf of customers,compared with only 23%in the bottom quartile.Insight:What we mean by virtual agent technology Virtual agent technology re
30、fers to the deployment of automation that uses a combination of artificial intelligence(AI)technologies or capabilities like machine learning,natural language processing,natural language generation,sentiment analysis,language translation,speech-to-text,robotic process automation(RPA),and optical cha
31、racter recognition(OCR)to automate dialogue,and in some cases backend process steps,to serve end users.This may or may not take the form of a chatbot.Some chatbots qualify as VAT because they incorporate machine learning as well as natural language processing.Integrated voice response(IVR)systems on
32、ly qualify as VAT when they use AI to power a conversational experience and the user is not limited to uttering a set of specific key words.Insight:Conversational user experience(UX)design Conversational user experience(UX)is a nascent area within design that focuses on creating systems that can eng
33、age in natural conversation.User interaction works through the words alone,whether voice or text.This requires formal knowledge,adapted from social science,of how human conversation works.While some chatbots and voice assistants are limited to answering questions or following commands,virtual agent
34、technology using natural conversation can engage in a wider range of activities,from hellos to goodbyes.The goal is mutual understanding,and it includes many natural mechanics for displaying and repairing misunderstandings.On top of this foundation,designers can create experiences that reflect the o
35、rganizations brand qualities.3 4 Influencing VAT performance Many factors have the potential to influence VAT performance,from levels of strategic intent and investment to the skills of the people maintaining it and the choice of specific technologies implemented,such as the standard of parsers used
36、 for natural language processing.A large pharmacy store chain in the US implemented sentiment analysis for contacts that are escalated by the VAT to human agents so agents can quickly prioritize customers at risk of becoming dissatisfied.This enables agents to directly influence their most important
37、 metric:customer satisfaction.Some factors influence individual metrics and others have the potential to influence broader performance across multiple metrics(see Figure 2).Machine learning and systems integration factors have the greatest impact on levels of relative performance and are interrelate
38、d.Machine learning factors:The longer the VAT is in use,the more content it has been trained to handle,and the more it does,the better it should perform.Systems integration factors:The greater the level of integration between the VAT and other systems and the more complex actions the VAT is enabled
39、to take,the better the results.Organizational factors:Specific circumstances such as the emphasis placed on tracking results and industry-related aspects such as complexity of customer contacts and digital uptake can influence performance.5 Figure 2 Factors influencing VAT performance Machine learni
40、ng factors Systems integration factors Organizational factors Duration in use Number of separate intents in scope%of contacts being handled Volume of contacts being handled Degree of integration between Type(s)of VAT implemented VAT and other systems Complexity of action the VAT Level of investment
41、in VAT is enabled to take Performance focus Industry North American grocery chain:Outsourcing to AI Competition among grocery chains is fierce,and with thin margins,customer loyalty is critical.A large North American grocery chain recognized that a smart customer contact center could enable a strong
42、er,expanded loyalty program.Hiring,training,and retaining call center agents is challenging.The chain needed a cost-effective,predictable,scalable way to handle increasing call volumes and complexity with minimal capital investment.The company turned to IBM to design,build,and operate an enterprise
43、cognitive call center solution that leverages VAT with voice integration.The solution features call flows designed to optimize customer experience,while lowering operational costs and scaling for the future.Integrated with backend systems,the VAT system is enabled to authenticate callers and take ac
44、tion.Each call is logged in the companys customer relationship management system,which can track the solutions performance like it would for a human agent.The chain only pays for calls successfully handled and saves 40%on each contained call.Its like outsourcing the call center to VAT.Leader attribu
45、tes Our analysis identified a small group(105 organizations)of VAT pioneers we call the“Leaders.”Each Leader demonstrates three critical attributes that are the result of deliberate business decisions:Early adopter of VAT Large proportion of inbound contacts are within the scope of their VAT High de
46、gree of integration exists between their VAT and backend systems.Comprising 10%of our sample,Leaders span all industries and regions surveyed,as well as all three types of VAT implementation,in almost equal proportions.Leaders also report outperforming their competition on revenue and profitability
47、for the last three years.Leaders outperform other respondents on core metrics used to measure the efficiency of VAT(see Figure 3).Figure 3 The efficiency of Leaders VAT far exceeds what others report 6 75%62%74%63%Containment*First contact resolution*Containment definition for this study:The portion
48、 of total contacts that the virtual agent technology has been trained to handle that it resolves without any escalation to or involvement from a human agent.80%66%Intent recognition Leaders Others“Your first release will be your worst.”Virtual agent technology developer,financial services industry,U
49、K Continuous improvement Continuous improvement is a critical component of VAT performance.Reviewing and retraining VAT is conducted by skilled humans and through automation.While all respondents report using automation,more of the respondents outside the Leader group use manual methods(41%)than Lea
50、ders(33%).But its the way Leaders select virtual agent conversations for review that truly differentiates them.Leaders use a wider range of techniques and more sophisticated methods(see Figure 4).Leaders dont just passively measure;they act.Nearly all Leaders have expanded the scope of their VAT in
51、response to their reviews,compared with 59%of others.Whether enhancements entail additional VAT training or upgrades to the VAT interface,Leaders are taking action to continually improve performance(see Figure 5).Figure 4 How virtual agent conversations are selected for review Customer intent 86%68%
52、Sentiment or tone analysis results 74%55%Contacts where SLAs missed 74%51%Neurolinguistic program confidence level 64%14%Specific call categories,e.g.,abandoned calls or by agent 58%51%High volume,low containment 58%45%Algorithm Random 55%36%30%49%Leaders Others Figure 5 Results of virtual agent qua
53、lity reviews in the last 30 days Scope of virtual agent expanded 91%59%Additional virtual agent training 82%64%Customer interface changed 82%64%Scope of virtual agent reduced 2%13%Leaders Others 7 Hong Kong bank:Continual VAT enhancement pays off In the quest to improve customer utility and ease of
54、engagement,a leading bank in Hong Kong launched a first-of-its-kind AI virtual assistant to address retail customers inquires for select services.Over three years,the VAT was extended to support 13 services,including account services,payment platform,bill payment services,and credit card,with contex
55、tual conversation in Chinese(Cantonese and Simplified Chinese),English,and even a mix of languages.Today,the companys VAT can talk with a customer to handle personal banking requests,such as a money transfer,and is integrated with the banks e-banking services.The VAT system has won more than 15 awar
56、ds,and customer use of the channel has grown by a factor of 6.This success led the bank to implement VAT for its business banking digital engagement platform.Here,too,results are impressive:User surveys report 75%satisfaction,and use of the VAT channel has grown by a factor of 10.VAT impact Virtual
57、agent technology directly impacts customers,human agents,and organization financial results.Impact on customers There are a variety of reasons why organizations adopt VAT,but improving customer experience is,by far,the most cited one(see“Hong Kong bank:Continual VAT enhancement pays off”).For 46%of
58、Leaders and 33%of other respondents,it is the single most important driver(see Figure 6).Figure 6 Reasons for implementing VAT 8 46%22%Leaders Most important reason Others 33%15%10%19%23%8%13%11%Leaders All reasons Others Customer experience Improve human agent experience 88%69%68%48%41%72%45%42%61%
59、64%Improve brand differentiation Reduce cost Increase revenue 99%of all respondents report positive absolute change-to-date in customer satisfaction for all inbound contacts as a result of using VAT.In other words,regardless of whether improving customer experience was the driver for their implement
60、ation,practically all report an increase in customer satisfaction scores as a result of using VAT.The vast majority of Leaders credit VAT for having a“significantly positive”impact on their customer satisfaction scores(see Figure 7).Figure 7 How the use of VAT has impacted customer satisfaction scor
61、es 70%63%Customer satisfaction last 30 days 12 8 Percentage-point change in customer satisfaction from using VAT Leaders Others Impact of VAT on customer satisfaction 2%1%20%12%25%73%67%Leaders Others Somewhat negative impact Somewhat positive impact Neutral Significantly positive impact Vodafone:VA
62、T success As part of its customer experience strategy and digital transformation,Vodafone Group became one of the worlds first telecommunications companies to adopt agent technology for customer service.Today,TOBi,Vodafones AI-based digital assistant,is available throughout the UK,Europe,Africa,and
63、the Middle East.TOBi provides a central,modular platform to which channels can connect,enabling individual markets to prioritize the channels that suit the preferences and needs of their customers.Channels deployed include SMS,the My Vodafone app,local Vodafone homepages,Facebook,WhatsApp,and Amazon
64、 Alexa.As a result,the number of Vodafone customers messaging versus calling the call center has almost doubled.In Italy,two-thirds of customer contacts are now contained through TOBi.9 Impact on human agents Agents who feel valued and empowered with the proper tools and support are more likely to d
65、eliver a better experience to customers.And the cost of replacing a dissatisfied agent who leaves the organization can be as much as 33%of the exiting employees salary.4 Leaders,especially,are keen to improve the human agent experience;68%cite it among their reasons for adopting VAT,and our data sug
66、gests it is having a positive impact.By calculating all respondents reported change-to-date as a percentage of their current human agent satisfaction,we determined VATs average contribution to current human agent satisfaction to be an impressive 20%.VAT also impacts human agent handle time.Average c
67、hange-to-date as a result of using VAT is a 12%reduction,with Leaders reporting an average reduction of 15%.The only area where study results point toward a minor adverse impact is human agent attrition.Absolute change-to-date was an average 1%increase in annual agent voluntary turnover.In the conte
68、xt of typical levels of contact center turnover,this is not large,but the direction is interesting.This finding could suggest that VAT is a factor in the decision to leave for some agents,while it contributes to improved satisfaction for those who stay.Impact on financial performance When you reduce
69、 the time it takes human agents to resolve contacts,you reduce cost to serve.Whether youve empowered agents with VAT to reduce their manual effort or youve provided customers or employees with VAT for self-service or to conduct the initial intercept,these implementations can have a significant finan
70、cial impact.For example,the Forrester Consulting study estimated that a large organization could achieve an average cost saving of$5.50 per contained conversation using IBMs Watson Assistant.A conversational AI platform,Watson Assistant allows users to interact with business systems using natural hu
71、man language.Forrester Consulting estimates that over three years,with a conservative 25%containment rate,the cost savings would be worth more than$13 million.5 Respondents in our study report both top-and bottom-line benefits from using VAT.Every respondent indicates that VAT has contributed to an
72、increase in organization revenue,the average increase being 3%.Total revenue uplift for respondents reporting results for VAT directly interacting with external customers sums to$5 billion.6 In addition,99%of respondents report that VAT has reduced their cost per contact,with 25%citing a reduction o
73、f 20%or greater.Leaders,particularly,have enjoyed VATs benefits,both the increased satisfaction and decreased costs(see Figure 8).Figure 8 Leaders average reported impact of using VAT 10 Customer First contact satisfaction resolution*Revenue increase increase*NPS points Human agent handle time FTEs
74、handling inbound contacts-18%-15%-18%14 12 3%5 pts Cost per contact*Percentage-point increaseReturn on investment 54%of all respondents report having already achieved or exceeded their expected return on investment(ROI)for their VAT implementation.94%of Leaders are in this position compared to 49%of
75、 others(see Figure 9).It took,on average,two years to achieve expected ROI.For those who exceeded ROI expectations,the average time-in-use was only six months longer at two-and-a-half years.64%of all respondents use a threshold rate of return for their VAT decision,while close to three quarters of L
76、eaders do.The average threshold rate of return for Leaders is 25%,nearly double the average of other respondents.Figure 9 Achievement of ROI for VAT Leaders Others 5%46%42%7%64%30%6%Have not achieved ROI and uncertain if it will be achieved Have not achieved ROI but expect to do so Have achieved exp
77、ected ROI Have exceeded expected ROI Optimizing VAT investments Organizations are transforming their infrastructures,platforms,and cloud strategies alongside their VAT investments.VAT maturity for customer care relies heavily on an organizations ability to link and leverage the ever-increasing volum
78、e of customer data across internal and external sources.Cloud is a key enabler to achieve this across channels and ecosystems.More than 200 survey respondents provide insights into their organizations use of cloud computing across their service desk systems.All report using cloud for customer relati
79、onship management(CRM)and chat messaging systems,as well as their VAT.Considerably more Leaders report having fully integrated cloud strategies for their service desk systems(see Figure 10).Figure 10 Strategy to move service desk systems to the cloud 11 13%68%19%56%44%Fully integrated cloud strategy
80、 for service desk systems,with specific actions plans and KPIs Cloud strategy for service desk systems with specific action plans and KPIs Cloud strategy for service desk systems,but it lacks specific action plans and KPIs Leaders Others “It was COVID that sparked VAT as an opportunity.It drove us t
81、o innovate and take some risks to quickly implement something wed already planned.The faster you launch something,test it,and iterate on it,the more value you can get out of it.”HR digital transformation leader,telecommunications,Australia The pandemic effect Todays business environment precipitated
82、 by the COVID-19 crisis is an ideal time to leverage the benefits of VAT.A majority of respondents saw a jump in contact center inbound volumes due to COVID-19.Having invested more in VAT,Leaders were prepared to reap the benefits.Far more Leaders say they“significantly”increased the number of inten
83、ts in the scope of their VAT(29%versus just 8%for others).Accordingly,three out of four Leaders increased the number of full-time employees(FTEs)reviewing and training their VAT,while as many as 59%of others did nothing to increase FTEs for these tasks.Leaders focus on VAT expansion and enhancement
84、during the pandemic is paying off.Almost half report a“significant”increase for intent recognition,compared to just 23%of others.More Leaders also cite increases in containment and first contact resolution.Although just over half of other respondents experienced an increase in customer satisfaction
85、for contacts supported by VAT,this is true for 70%of Leaders.More Leaders also saw significant bumps in agent satisfaction.Tellingly,their pandemic experience has made VAT enthusiasts out of the majority of respondents,with most expecting to increase VAT investments as a result(see Figure 11).Figure
86、 11 Nearly all Leaders and more than half of other respondents plan to increase their investment in VAT as a result of COVID-19 12 44%47%COVID-19 impact on organizations investment in VAT in next six months 5%49%Significantly increase investment in VAT Somewhat increase investment in VAT 36%9%9%No i
87、mpact on investment in VAT Somewhat decrease investment in VAT 1%Significantly decrease investment in VAT Leaders Others Action guide The value of virtual agent technology Establish the business value you want to deliver and how you measure success whether you are mature,mid-journey,or making a firs
88、t decision about using VAT.The Forrester Consulting study estimated that a large organization can achieve payback within 6 months.7 Respondents to the IBM Institute for Business Value survey8 report achieving benefits for individual metrics up to:50%cost savings 18 percentage-point increase in custo
89、mer satisfaction 20 percentage-point increase in first contact resolution 20 percentage-point increase in human agent satisfaction 15 point increase in NPS 9%increase in revenue attributed to use of VAT.If you dont have virtual agent technology today:For a situation in which you need to“stop the ble
90、eding,”you can get started by deploying VAT as an MVP within a few days.Move to automate thousands of conversations to scale capacity for business continuity and enhanced customer service.Focus on targets that dont have additional downstream implications or dependencies on other technologies you may
91、 be implementing.Develop a work-from-anywhere capability by leveraging a CCaaS partnership.If youre on the journey:Through agile cocreation,establish an intent resolution plan to expand the scope of your VAT;train it to recognize and handle the vast majority of inbound contacts.Ensure your plan is a
92、ligned to your customer experience journey and develop 100%consistency in customer interactions across channels and additional lines of business.Enhance personalization and your ability to predict and proactively resolve customer queries by modernizing your data and cloud solutions.Conduct a value a
93、ssessment periodically to stay current as technologies are sunset and refined solutions are introduced.If you want to stay the course:Dedicate the time and investment needed to conduct the proper integration with your backend systems,particularly as you develop more complex use cases and integrate y
94、our VAT to achieve content intelligence.Create a VAT center of excellence to purposefully manage growth,coordinate measurement,and optimize the use of automation for continuous improvement activities.Refine the user experience leveraging AI and data to further personalize interactions with your cust
95、omers to develop high-value interactions that enhance your digital brand.Build a longitudinal contact view into your measurement systems to target a substantial reduction in repeat contacts.13 Regardless of where you are with regard to VAT,there are additional factors and steps to consider.Incorpora
96、ting these actions throughout the process will help ensure long-term success.Establish governance for continual improvement:Adopt a broad strategic vision for your VAT experience managed with a governance program to oversee VAT implementations across the enterprise.Determine consistent processes for
97、 monitoring the platform,channels,and change management.Create a consistent set of metrics,KPIs,and reporting methods that support new business objectives,digital brand,and customer and employee expectations.Build the services required to support digital change in the plan with the core competencies
98、 that support new digital roles.Expand throughout the enterprise:Consider the value of scaling your virtual agent technology across the enterprise in all lines of business for your customers and employees.Enable your virtual employees like you enable your human employees.Dedicate the time and invest
99、ment needed to integrate with 100%of your backend systems that are required to support end-to-end engagements with customers and employees in every channel.Expand the scope of your VAT by training it to recognize the vast majority of your contacts where customers require support.Optimize the use of
100、automation for continuous improvement activities.14 Study methodology We conducted three types of analysis to investigate the significance of machine learning,systems integration,and organizational factors on VAT performance:We created peer group sets by factor and examined results for these sets,lo
101、oking for consistent trends.We analyzed the average result for each metric for each set,looking for both a statistically significant differential across the entire range and a logical progression of results across the range.Using the factors identified as most significant based on the peer group ana
102、lysis,we conducted regression analysis for twelve key metrics,including intent recognition,containment,customer and human agent satisfaction,and cost per contact.We grouped respondents into quartiles based on their results for three specific metrics,including containment,and reviewed them based on t
103、he machine learning,systems integration,and organizational factors.Unless otherwise stated,values shown are median average values.About the authors Joe Petrone Joe.P Joseph Petrone is a Partner and the Global Cognitive Care Offering Leader for IBM Services.Joe has deep knowledge and more than 20 yea
104、rs experience in high-tech contact centers and service desks,assisting clients with building,managing,and driving efficiencies through enterprise-scale distributed workforces.Joe has focused his last five years helping customers digitally transform their operations infrastructure and processes to im
105、prove cost and service and to align support through channel rationalization.Gillian Orrell Gillian Orrell develops the IBM Institute for Business Values Performance Data&Benchmarking content for customer experience,contact centers,and sustainability and leads delivery of benchmarking services to cli
106、ents in Europe and beyond.For 25 years,she has been helping clients measure and improve their performance across a wide range of priorities and processes.Carolyn Heller Baird Carolyn Heller Baird is the Global Research Leader for Customer Experience and Design with the IBM Institute for Business Val
107、ue.In addition to her research on topics impacting customer experience,Carolyn is responsible for the Chief Marketing Officer studies for the IBV C-suite series.Her work spans nearly 20 years as an experience strategy consultant.The IBV would like to thank the following individuals for their contrib
108、utions to the development of this report:Francesco Brenna,Jessica Chiu,Gaurav Chugh,Michael Conway,Christine Dee,Markus Graulich,Manish Goyal,Avdyl Haxhaj,Ross Judd,Raquel Katigbak,Rahul Kumar,Joseph Ma,Luca Marchi,Bob Moore,Vij Prasad,Prad Paskaran,Alexis Rodriguez Lorenzo,Paul Sherlock,Sebastian W
109、eir 15 Notes and sources 1“The Total Economic Impact Of IBM Watson AssistantA Forrester Total Economic Impact Study Commissioned by IBM.”Forrester Consulting.March 2020.Note:This study estimates benefits for a composite organization based on four companies Forrester Consulting interviewed.The compos
110、ite organization has attributes including$10 billion revenue,40,000 employees,1 million customer conversations monthly,and implementation of three types of VAT over three years.https:/ IBM_Watson_Assistant-March_2020_v3.pdf 2 Study respondents were senior executives,directors,or managers of contact
111、center or customer service operations,channels,digital technology,or IT with hands-on knowledge of their VAT implementations.3 Moore,Robert J.,and Raphael Arar.Conversational UX Design:A Practitioners Guide to the Natural Conversation Framework.AMC Books.2019.4 Hall,John.“The Cost of Turnover Can Ki
112、ll Your Business And Make Things Less Fun.”Forbes.May 9,2019.https:/ 5“The Total Economic Impact Of IBM Watson AssistantA Forrester Total Economic Impact Study Commissioned by IBM.”Forrester Consulting.March 2020.Note:This study estimates benefits for a composite organization based on four companies
113、 Forrester Consulting interviewed.The composite organization has attributes including$10 billion revenue,40,000 employees,1 million customer conversations monthly,and implementation of three types of VAT over three years.https:/ IBM_Watson_Assistant-March_2020_v3.pdf 6 For each of these respondents,
114、we multiplied organization revenue by the percentage by which the use of VAT has impacted the organizations annual revenue.The sum of these 343 calculations is$5 billion.7“The Total Economic Impact Of IBM Watson AssistantA Forrester Total Economic Impact Study Commissioned by IBM.”Forrester Consulti
115、ng.March 2020.Note:This study estimates benefits for a composite organization based on four companies Forrester Consulting interviewed.The composite organization has attributes including$10 billion revenue,40,000 employees,1 million customer conversations monthly,and implementation of three types of
116、 VAT over three years.https:/ IBM_Watson_Assistant-March_2020_v3.pdf 8 Average benefits across all respondents in this study for individual metrics are as follows:14%cost savings 8 percentage-point increase in customer satisfaction 10 percentage-point improvement in first contact resolution 7 percen
117、tage-point improvement in human agent satisfaction 4 point increase in NPS 3%increase in revenue attributed to use of VAT.16 The right partner for a changing world At IBM,we collaborate with our clients,bringing together business insight,advanced research,and technology to give them a distinct advan
118、tage in todays rapidly changing environment.IBM Institute for Business Value The IBM Institute for Business Value,part of IBM Services,develops fact-based,strategic insights for senior business executives on critical public and private sector issues.For more information To learn more about this stud
119、y or the IBM Institute for Business Value,please contact us at .Follow IBMIBV on Twitter,and,for a full catalog of our research or to subscribe to our monthly newsletter,visit: Benchmark Insights Benchmark Insights feature insights for executives on important business and related technology topics.T
120、hey are based on analysis of performance data and other benchmarking measures.For more information,contact the IBM Institute for Business Value at .Copyright IBM Corporation 2020 IBM Corporation New Orchard Road Armonk,NY 10504 Produced in the United States of America October 2020 IBM,the IBM logo,a
121、nd are trademarks of International Business Machines Corp.,registered in many jurisdictions worldwide.Other product and service names might be trademarks of IBM or other companies.A current list of IBM trademarks is available on the web at“Copyright and trademark information”at document is current a
122、s of the initial date of publication and may be changed by IBM at any time.Not all offerings are available in every country in which IBM operates.THE INFORMATION IN THIS DOCUMENT IS PROVIDED“AS IS”WITHOUT ANY WARRANTY,EXPRESS OR IMPLIED,INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY,FITNESS FOR
123、 A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT.IBM products are warranted according to the terms and conditions of the agreements under which they are provided.This report is intended for general guidance only.It is not intended to be a substitute for detailed research or th
124、e exercise of professional judgment.IBM shall not be responsible for any loss whatsoever sustained by any organization or person who relies on this publication.The data used in this report may be derived from third-party sources and IBM does not independently verify,validate or audit such data.The results from the use of such data are provided on an“as is”basis and IBM makes no representations or warranties,express or implied.10035310USEN-01 17