《Language IO：2024企業創新研究報告：通過AI和數字化轉型戰略贏得客戶信任的原因和路徑（英文版）（18頁）.pdf》由會員分享，可在線閱讀，更多相關《Language IO：2024企業創新研究報告：通過AI和數字化轉型戰略贏得客戶信任的原因和路徑（英文版）（18頁）.pdf（18頁珍藏版）》請在三個皮匠報告上搜索。

1、The language of innovationWhy and how to gain your customers trust with your AI and digital transformation strategyGetting ahead or left behind?Terms like ChatGPT and generative AI have made it into the public lexicon.From increasing yields and reducing risks in the mining industry to reducing food

2、wastage in agriculture,AI is at the forefront of all things innovation in every space.Its no wonder,then,that its hard to separate innovation from AI theyre so tightly intertwined.Every leader wants to innovate,every brand promises its customers innovation,and most companies have jumped on the AI ba

3、ndwagon.This enthusiasm for AI is not unwarranted:A McKinsey research revealed that AI has the potential to generate up to$2.6 trillion to$4.4 trillion in global corporate profits annually.In this guide,well unpack the top priorities of digital transformation.Well also talk about why your AI integra

4、tion initiatives should start with optimizing customer experience primarily customer support.And this guide will not be complete without information on how you can make translation a seamless part of your innovation were experts on this subject,after all.“AI has the potential to generate up to$2.6 t

5、rillion to$4.4 trillion in global corporate profjts annually.”Chapter 1Accelerating transformation with AI,togetherFor successful digital transformation,you need to bring the whole team along,from the top leadership to the most junior employees.And remember:No matter what decisions you make and what

6、 implementations you roll out,your customers will be impacted.What stands in the wayA Salesforce survey shows that AI has gone mainstream:78%of IT and business leaders say that they know what role AI can play in their organization.And according to another report,76%of executives say their companies

7、have already tested AI applications over the last year.However,despite the hype and the early adoption of AI by several companies,the results have been surprisingly underwhelming.According to IBMs research,about 40%of organizations are still exploring or experimenting with AI but have not yet succes

8、sfully deployed their models.Among the top barriers hindering the rollout of AI initiatives:Limited AI skills and expertise33%Too much data complexity25%Ethical concerns23%Difficulty when it comes to integration and scaling22%High price21%Lack of tools21%The top two inhibitors for generative AIData

9、privacy57%Trust and transparency43%Chapter 1 Accelerating transformation with AI,togetherSome ways to boost transformationHere are a few considerations and solutions to help you tackle these common challenges.Let go of dated systemsLetting go of legacy systems and outdated processes is just as impor

10、tant as adopting new technology.Operating and maintaining legacy technology that no longer makes sense for your organization is taking up most of your costs and resources without bringing any real value.Overhauling dated programming languages,databases and architectures will give you more flexibilit

11、y and boost innovation.Engage everyoneDigital transformation is a team sport.To foster a culture of innovation,you need to engage every employee the new employees and the old-timers,the team players and the lone wolves,the ones who embrace risks and the ones who strictly adhere to rules.Without a te

12、am thats supportive of your ambitions,its hard to drive the project forward.Find the right partnersYou dont always have to start from scratch and hire new resources to get to your goals.Find innovative vendors who wont hit you with the standard fare,but instead give you solutions tailored to your or

13、ganization that align directly with your strategy and budget.Finding such partners will help you reduce time to business value.The cornerstones of successful digital transformationDigital transformation is an enterprise-wide initiative.Every initiative needs support and trust from C-level executives

14、 to the frontline employees.Real innovation starts and ends with customer experience.The quality of experience your customer is getting at every touchpoint is the end goal of your digital transformation strategy.Finding the right trustworthy vendors can be a game-changer when executing your vision.S

15、eeking the best external partners can significantly strengthen your responsible innovation and ethical AI projects and help you get from inception to impact faster.Chapter 1 Accelerating transformation with AI,togetherSome ways to boost transformationPay special attention to integration and interope

16、rabilityWhen your organization adopts multiple technologies,systems and platforms that need to work together seamlessly,its crucial to find ways to integrate them all meaningfully.This is crucial to enable data flow,cross-functional collaboration and workflow automation between all the systems so ev

17、ery employee is equipped with all the data and tools they need and there are no gaps across the organization.Make sure to look for solutions that come with flexible APIs or out-of-the-box integrations.This would help you reduce training and onboarding time and even boost your teams morale.Customer e

18、xperience:the biggest motivatorPeter Drucker,the inventor of modern management,had this to say:“The purpose of business is to create and keep a customer.”Sounds simple enough,right?Create customers who need your product and then keep the customers coming back.Whether your goal is maximizing efficien

19、cy or improving employee productivity or saving costs,it always has a direct effect on the quality of customer experience.And that,in turn,impacts your business value.If youre not building a great relationship with your customer or securing your customers trust,or your customer isnt getting the best

20、 experience,theres no point in innovating.Chapter 1 Accelerating transformation with AI,togetherTrust makes transformation workThe heartbeat of this guide so far has been trust from both employees and customers.Research shows that companies that are trusted outperform their competitors by 400%.Lets

21、explore ways to retain and gain the trust of customers and employees through(and during)digital transformation.EmployeesThe goals of digital transformation projects are usually reducing costs and improving operational efficiency and these hinge on increasing employee productivity.But disengaged empl

22、oyees are not productive:Employees who are actively disengaged at work cost the world$8.8 trillion in lost productivity.Lets tie this to the context of AI.A recent survey revealed that AI is not a concrete thing for most employees.Around 75%of employees say they would be more accepting(and even exci

23、ted about)AI if their organization was more transparent about the direct ways AI could improve their workflows and how it plans to use the technology in the future.But the reality is that more than half(54%)of the employees surveyed said they have no idea who is using AI within their company.So what

24、 use is it if your employees trust you?Research shows that employees who trust their companies are 260%more motivated to work,have 41%lower rates of absenteeism,and are 50%less likely to look for another job.Chapter 1 Accelerating transformation with AI,togetherEmployees who are willing to embrace A

25、I at work say it would increaseEngagement and job satisfaction63%Willingness to go above and beyond55%Overall happiness54%Desire to stay with their company long term49%Trust in their company/leadership48%Trust makes transformation workRetaining employee trust during digital transformation involve?Cl

26、early communicating plan?Tracking progress across every aspect of your complex digital initiativ?Constantly driving alignment between teams(especially business and technology?Creating clear cross-functional collaboration and issue resolution workflow?Seizing quick wins instead of pushing large proje

27、cts to improve morale and instill confidenc?Build,streamline and optimize processes to enable them to effectively meet their work need?Tailoring your messaging and training programs with empathy for different groups of employees and helping them upskillCustomersConsumers these days are more informed

28、 and more tuned into developments in the business space.Theyre comfortable with sharing their data if its used to target personalized suggestions to them and offer support tailored to their needs.Consumers are more likely to choose companies with transparent and ethical AI practices.Not only that,co

29、mpanies need to be able to explain how their AI reached a decision.Its no wonder that 93%of business executives prioritize building and securing trust and agree that it directly improves the bottomline.Consumers spend more at companies they trust 46%purchased more,and 28%paid a premium.And theyre no

30、t afraid to switch:4 in 10 customers no longer purchase from a company due to lack of trust.Another report shows that your transparency surrounding AI and your bottomline are directly linked.72%of customers say its important to know a companys AI policies before making a purchase.Factors important t

31、o buying decisionsPrice94%Descriptiom93%Quality92%Convenience92%Ethical and trusted reputation87%Amount of personal data required87%Speed if delivery87%Survey respondents consider trustworthiness and data protections to be nearly as important as price and delivery time.Chapter 1 Accelerating transfo

32、rmation with AI,togetherCustomers seek transparency as trust erodesCustomer experiences should be better considering all the data companies collect85%79%80%AvergageIm increasingly protective of my personal data84%78%79%AvergageId more likely trust a company with my personal data if its use were clea

33、rly explained80%68%71%AvergageMost companies dont use my personal data in a way that benefits me58%48%51%AvergageIm comfortable with brands collecting first-party data64%39%44%AvergageIm comfortable with brands collecting my third-party data53%23%30%AvergageBusiness buyersConsumersChapter 2Customer

34、support as the starting point of innovationThe direct relationship between customer service and revenue is not new information.Customers are not loyal to brands;theyre loyal to value and good service.In fact,poor customer service could cost so much as$3.7 trillion in sales annually.On the other hand

35、,data from Salesforce suggests that not only are customers more likely to purchase again from brands that offer good customer service,theyre also ready to forgive the brands mistakes.Given all this data,its clear that customer service is the ultimate value driver and the best starting point for inno

36、vation in an organization.There are several opportunities to bring AI-driven digital transformation into optimizing this space and maximizing impact.Chapter 2 Customer support as the starting point of innovationGreat service drives repeat purchases,advocacy,and even forgivenessCustomers forgave a co

37、mpanys mistakes after receiving excellent customer service77%74%75%AvergageCustomers recommend a company based on excellent customer service85%71%75%AvergageCustomers made purchase decisions based on customer service quality82%68%71%AvergageBusiness buyersConsumersMeeting increased customer expectat

38、ionsTodays customer has already been influenced by AI,and their expectations are higher than ever.An Intercom survey said the top three areas in which customer support teams are seeing an increase in expectations are speed of response(63%),resolution times(57%)and knowledge and expertise(49%).To hel

39、p your support team rise to these rapidly evolving expectations,you can implement some of these recommendations(and AI upgrades).Enabling self-service and increasing agent bandwidthResearch shows that employee effort in most customer service organizations goes toward repetitive tasks that impact bot

40、h productivity and agent satisfaction.Nearly 78%of CX leaders say that the biggest inhibitor of agent productivity is low-impact work that could easily be offloaded to technology.Other challenges include accessing relevant training and guidance and supporting customers across different channels all

41、of which can be solved with the right technology.Chapter 1 Why optimize chat support?The top areas where support teams have seen the biggest increase in customer support63%Speed of response57%Speed of resolution49%Knowledge and expertise49%Availability43%Politeness and empathyCustomer service agents

42、 are bogged down by high-effort,low-value problems.Repetitive tasks could be automated with AI,and a well-executed self-service knowledge base system integrated with a chatbot could reduce calls,email and live chat support requests,giving agents more time to work on high-impact projects.Customers wa

43、nt self-service as well:A recent survey revealed that younger generations of customers have a“self-service or no service”mindset:38%of Gen Z and millennial customers are likely to give up on an issue if they cant resolve it in self-service.And if they did not have self-service and seamless transitio

44、ns to live support,they say they?would use the service less(55%?wouldnt buy from that company again in the future(52%?would say negative things about the company(44%)AI tipTo seamlessly collect,analyze and share customer data across the organization,implement support AI solutions that are equipped w

45、ith several capabilities,such as?Adding AI tools like chatbots to your knowledge base systems and self-service portals can give your customers access to instant answers and real-time AI-powered assistance.With intuitive search capabilities and AI-driven functionalities,AI knowledge bases provide 24/

46、7 customer service,FAQs,troubleshooting guides and tutorials on demand.?Theyre also self-improving:With constant input from customers,theyre able to tailor topics and personalize content to users,identify content gaps,pinpoint outdated topics and give you recommendations for new content?AI-powered s

47、elf-service systems help new agents to train,onboard and ramp up faster.With comprehensive,up-to-date resources in a centralized,easy-to-navigate hub,AI knowledge bases enable agents to quickly access training and important customer information.This also helps keep brand voice consistent across agen

48、ts?AI algorithms can also suggest relevant materials to support agents,recommend articles within tickets,and help with writing responses to customers,increasing agent productivity and support quality.Chapter 2 Customer support as the starting point of innovationGetting and routing customer feedbackA

49、ccording to a recent survey,while 69%of CX professionals said they do collect customer feedback,there has been a decline in the success of this feedback reaching the right departments.In 2022,43%said that it“mostly”reaches the right department;in 2023,this figure has dropped to 27%.This shows the im

50、portance of not just gathering customer feedback but also turning that data into actionable intelligence for the right team.AI can help by collecting data that will continually evolve and adapt,helping customer service agents implement a more effective and personalized strategy that improves custome

51、r experience.AI tipTo seamlessly collect,analyze and share customer data across the organization,implement support AI solutions that are equipped with several capabilities,such as?Predictive analytics to anticipating customer behavior and needs and addressing them proactivel?Sentiment analysis to un

52、derstand customer feedback without them having to spell it out and improving experienc?Hyperpersonalization to recognize your customers and rewarding them for their loyalty with recommendations and offers tailored just for themChapter 2 Customer support as the starting point of innovationProviding f

53、ast response and resolutionA recent Accenture report shows that 47%of customers feel less valued when they face difficulty reaching or talking to unsupportive customer service agents.Thats a significant portion of your customer base that youre alienating without the right communication.An agent mann

54、ing live chat can only help one customer at a time.But a chatbot can establish personalized contact with and help multiple customers,looping in a human only when necessary.This helps you scale customer support without necessarily hiring more agents.This also reduces the inbound volume of tickets and

55、 increases operational efficiency.These are only some of the use cases that you can harness AIs vast potential for and some of the things to consider as you implement innovative solutions.In the next chapter,well speak to AI-powered translation for customer service and close out with what you should

56、 look for when it comes to security,data protection and responsible innovation.AI tipWell-designed AI-powered chatbots can?Automatically resolve low-touch,high-frequency issues by giving customers relevant knowledge base articles while they wait for an agen?Learn from each customer interaction and s

57、elf-improving so it can deliver more relevant and customized conten?Help your agents give timely support to your customers in their preferred languages,with the right AI-powered translation platform integrated no language skills requiredChapter 2 Customer support as the starting point of innovationC

58、hapter 3Eliminating language barriers with innovationProviding intelligent CX in your customers preferred language is another effective way to use AI and also improve efficiency.Whether youre expanding to new geographic locations or adding more languages to your customer support,AI-powered,real-time

59、 translations can help you support your global audience on your live chat,email,chatbot,FAQ and even social media customer service.First,a few important defjnitionsWhen discussing real-time translation,you might have heard the terms LLM and NMT.What are these,and how do these technologies power tran

60、slation?What are large language models(LLMs)?LLMs are systems and algorithms that have been trained on massive amounts of data to process,predict and generate natural language.For example,autocomplete and predictive text on your messaging apps are language models.These immense deep learning models a

61、re capable of self-learning and unsupervised training.From basic grammar and syntax to voice,tone and intent,these models learn to understand a variety of information.An example of an LLM-based model is Googles PaLM 2.Given the fluidity and complexity of human language,you can imagine the size of th

62、e datasets used to train these systems.But the most technologically advanced,incredibly accurate solutions today are powered by NMT.What is neural machine translation(NMT)?Neural machine translation is the next evolution of machine translation.Put simply,machine translation is a machine(like your co

63、mputer)translating text from one language to another.NMT takes this several steps further:It leverages artificial neural networks,such as transformers and recurrent neural networks(RNNs),to continuously improve the accuracy of real-time translations.NMT models are built and improved through an itera

64、tive training process that involves exposure to millions of sentence-specific translations between language pairs,gradually adjusting its output to identify and minimize translation errors over time.Chapter 3 Eliminating language barriers with innovationThe difgerences between LLM and NMTThe biggest

65、 difference lies in the purpose.LLM is more generic,trained on all kinds of data.However,NMT is more focused and specifically tailored.More than merely learning and providing a loose translation of specific words,NMT stands apart from less effective translation technologies with its focus on two key

66、 principles:context awareness and efficiency.NMTs are designed specifically to understand sentences in the unique context they are being used,with the corrective process being focused on picking up the nuances associated with translating word pairings from one language to another,as well as how to a

67、ccurately translate culturally specific idiomatic expressions.By contrast,less capable translation models often produce confusing or contextually and grammatically incoherent results,which can easily lead to the kind of miscommunication that any effective customer support system should be actively s

68、eeking to avoid.NMT models are built for accuracy and the speed and efficiency needed to produce coherent and meaningful translations in real time.When you couple context awareness with the ability to instantly produce translations,it becomes immediately clear why NMTs have become the ideal real-tim

69、e translation solutions for all kinds of customer support channels,particularly in any industry where speed and context are crucial for providing optimal customer service and experiences.Chapter 3 Eliminating language barriers with innovationWhat to expect from an advanced real-time translation solu

70、tionWhen it comes to integrating a real-time translation solution across the entirety of your customer service channels,the most important things to keep in mind are overall reliability and consistency of outcomes.Data security and storageHere are some security standards that we prioritize at Langua

71、ge I/O and we recommend that you do too.Zero data retention Establish security controls and zero data retention policies to ensure that your conversational transcripts are fully erased and never stored.If your selected vendor integrates with a variety of machine translation engines,make sure that th

72、ey only integrate with those that are zero-trace.This means that the third-party engines used do not store translated content in their database,log files,or anywhere else.Auditing When selecting a partner for your digital transformation project,make sure that they have a comprehensive audit in place

73、 and they maintain compliance with current and emerging legal and ethical standards.Data masking Ensure that any personally identifying information is masked before its processed.Its possible to create quality outputs without exposing confidential data to external vendors or AI models.Translating co

74、ntent often requires the use of third-party machine translation engines.Not only is it important that your technology provider uses only those MT engines that adhere to strict security measures,but its also critical that any personally identifiable information is encrypted before being sent to those

75、 engines.Chapter 3 Eliminating language barriers with innovationContext awareness and accuracyEven tech giants slip when they expand to new regions with inaccurate translations.Poor translations mean poor customer service experiences,which can erode customer trust in your brand.The ideal translation

76、 technology integrates with multiple best-in-class engines and dynamically selects the most viable engine for each translation,based on language pair and use case.The ability to switch between engines is critical,as not all engines are equally proficient in all languages.Not only that,you also want

77、your translation technology to understand the nuances depending on different contexts.Speed and integrationsMachine translations should be instant,without any latency issues.Your technology provider should be able to deliver translations across every digital channel through which you communicate wit

78、h customers.Whether youre looking to enable self-service,answer emails,provide instant chatbot support,be available on social media for your customers or all of the above,your translation technology should be equipped to make it happen.This platform should also integrate directly into your CRM,eithe

79、r with an integration or an API.How Language I/O does itLanguage I/O offers an NMT-based solution that doesnt rely on a single model to drive real-time support translations.This is because the accuracy of one NMT system will not always align with that of another when translating the same sentence fr

80、om the same language.In fact,in almost any case one model will be preferable to another in terms of both accuracy and efficiency in the unique context of the task presented.This is why Language I/O aggregates various of the worlds leading NMT engines and dynamically applies the most accurate model f

81、or the language pair involved in the translation.The selection of each specific NMT model for each task is powered by automation,which considers the individual depth of knowledge associated with each model as it relates to the specific language pair.Once the best possible NMT engine has been selecte

82、d,Language I/O further optimizes results by applying our own self-improving glossary of brand-and industry-specific terminology,ensuring all contextual and linguistic nuances have been considered.This enables highly accurate and context-aware translations even when there are jargon,misspellings,acro

83、nyms,brand terms and user-generated content that dont necessarily adhere to grammatical rules.ConclusionA fjnal note on responsible innovationWhile digital transformation is an exciting journey,its crucial to remember the responsibility that comes with the power of a tool like AI.Securing your emplo

84、yees and customers trust should always be a priority when taking on a project of this magnitude.Innovation is meaningful only when it comes with transparency.Be transparent about your plans and goals to your employees,and make sure to disclose how and why youre collecting customer information to mai

85、ntain their trust as well.Especially when it comes to translations in the customer service realm,its important to be culturally sensitive and mindful of customer privacy.Whatever technology you implement might have access to a lot of personal data,or the AI models your technology uses could create o

86、r reinforce unfair biases.Ensure that your provider is actively working against data breaches and bias so that the product is ultimately not causing harm.Read this nextWant to learn more about evaluating and selecting the right multilingual support technology provider for your organization?Download our buyers guide to multilingual support software.Looking to optimize all your customer service channels?Get the three-guide bundle on providing multilingual support over live chat,email and chatbot.