BCG:2024AI價值研究報告(英文版)(24頁).pdf

編號:179309 PDF  中文版  DOCX 24頁 13.23MB 下載積分:VIP專享
下載報告請您先登錄!

BCG:2024AI價值研究報告(英文版)(24頁).pdf

1、Wheres the Value in AI?October 2024 By Nicolas de Bellefonds,Tauseef Charanya,Marc Roman Franke,Patrick Forth,Michael Grebe,Romain de Laubier,Vlad Lukic,Amanda Luther,Clemens Nopp,and Joe SassineContents01 Whos Getting Results from AI and Why?A Steep Curve What Leaders Do Differently Incumbents Reap

2、 Value05 The Surprising Sources of Value from AI In the Core Sector Matters13 The Playbook for Winning with AI Overcoming Tough Challenges The Capabilities Required for Success Jump-Starting Your Journey to AI Value How Two Companies Applied the Playbook for Success17 Appendix20 AcknowledgmentsBOSTO

3、N CONSULTING GROUP 1Whos Getting Results from AI and Why?After all the hype over artificial intelligence(AI),the value is hard to find.CEOs have authorized invest-ments,hired talent,and launched pilotsbut only 26%of companies have advanced beyond the proof-of-concept stage to generate value.This rep

4、ort yields import-ant insights into what AI leaders are doing to drive real value from the technology,where others fall short,where the value is coming from,how individual sectors are per-forming,and how companies can change their own AI trajectories.Consider these examples of the value being create

5、d by AI,including generative AI(GenAI),from companies in three different sectors.A financial institution is committed to achieving$1 billion in productivity improvements,in addi-tion to enhanced risk outcomes and better client and employee experiences,by 2030.A biopharma company is chasing$1 billion

6、 in value potential(revenues and costs)by 2027.A major automaker expects to cut its cost of goods sold by up to 2%and accelerate new product devel-opment time by 30%.2 WHERES THE VALUE IN AI?These results are typical of the value that leaders across industries are achieving by building digital capab

7、ilities to a level at which they can implement AI programs at scale.BCGs latest research into AI adoption,a continuation of our studies into digital transformation and AI maturity,found that of the 98%of companies that are at least exper-imenting with AI,only 26%have developed the necessary capabili

8、ties to move beyond proofs of concept and begin extracting value.(For more on how we define AI and our research methodology,see the Appendix.)And only 4%are at the forefront of AI innovation,systematically building cutting-edge AI capabilities and scaling them across the organization.Heres our lates

9、t look at who the top 26%of companies are and how they are generating superior value from AI.The two chapters of the report that follow look at where com-panies are extracting value and what you need to do to move your company up the AI maturity curve.A Steep CurveBuilding AI capabilities is a compl

10、ex challenge.Our latest research,involving more than 1,000 companies worldwide,shows that only 4%have developed cutting-edge AI capa-bilities across functions and are using them to consistently generate substantial value.(See Exhibit 1.)Another 22%have an AI strategy and advanced capabilities and ar

11、e starting to generate value.We call these companies lead-ers.The remaining 74%have yet to show tangible value from their use of AI.These categorical distinctions are important because lead-ers far outperform the others.Over the past three years,leaders revenue growth has been 50%greater than the ov

12、erall average.Their total shareholder returns are 60%higher,and they gain 40%higher returns on invested capital.These companies also excel on nonfinancial factors,such as patents filed and employee satisfaction,and they are in pole position to benefit as AI platforms and tools mature.Exhibit 1-Leade

13、rs Have Built the Capabilities Needed to Implement AI at Scale,Reaping Diverse Benefits over Less Mature CompaniesSource:BCG Build for the Future 2024 Global Study(merged with DAI).Note:“Leaders”include AI future-built and AI scaling companies;“less mature”or“other”companies”include AI stagnating an

14、d AI emerging companies.RoIC=return on invested capital;TSR=total shareholder return.AI stagnating2549224AI emergingAI scalingAI future-builtValue achievedRevenue50%higher revenue growth(3-year average)60%higher 3-year TSR40%higher RoIC(3-year average)1.9xmore patents1.4xbetter overall Glassdoor ind

15、icatorTotal shareholderreturnReturnsInnovationEmployeesatisfaction0Are taking minimal or no AI action,lack foundational capabilities,and are not generating valueMaturity stage(%of companies)Have developed foundational capabilities and started initial experimentation but are struggling to scale and g

16、enerate valueHave developed an AI strategy and advanced capabilities,and are scaling them effectively while starting to generate valueAre at the forefront of AI innovation,systematically building cutting-edge AI capabilities across functions and consistently generating substantial value255075100BOST

17、ON CONSULTING GROUP 3What Leaders Do DifferentlyLeaders have six differentiating characteristics.They focus on the core business processes as well as support functions.A common misconception is that AIs value lies mainly in streamlining operations and reducing costs in support functions.In fact,its

18、greatest value lies in core business processes,where leaders are generating 62%of the value.Leveraging AI in both core business and support functions gives these companies competitive advantage.They are more ambitious.Leaders expectations for revenue growth from AI by 2027 are 60%higher than those o

19、f other companies,and they expect to reduce costs by almost 50%more.Three-quarters of the most forward-looking companies focus on company-level innovation core to the business.In contrast,only 10%of other companies do soand if they leverage AI at all,it is mainly for produc-tivity.Leaders look beyon

20、d pure productivity plays and back their ambitions with investment in AI and workforce en-ablement,doubling down on several aspects of AI,relative to their peers.(See Exhibit 2.)They make twice the invest-ment in digital,twice the people allocation,and twice the number of AI solutions scaled.They in

21、vest strategically in a few high-priority oppor-tunities to scale and maximize AIs value.Data on AI adoption shows that leaders pursue,on average,only about half as many opportunities as their less advanced peers.Leaders focus on the most promising initiatives,and they expect more than twice the RoI

22、 in 2024 that other companies do.In addition,leaders successfully scale more than twice as many AI products and services across their organizations.They integrate AI in efforts both to lower costs and to generate revenue.Almost 45%of leaders integrate AI in their cost transformation efforts across f

23、unctions(com-pared with only 10%of nonleaders).And more than a third of leaders focus on revenue generation from AI,compared with only a quarter of other companies.(See Exhibit 3.)“We have a program under which every business unit is required to submit three to five projects each yearand since 2020,

24、they have all had to focus on AI,”said the enterprise product director of an alternative energy compa-ny.“These projects need to demonstrate how they would improve the company either through cost savings,in-creased operational efficiency,or revenue generation.”Exhibit 2-Compared with Their Peers,Lea

25、ders Are Allocating More of Their Budget and Resources to Digital and AI Capabilities in 2024Source:BCG Build for the Future 2024 Global Study(merged with DAI).Note:FTEs=full-time equivalent employees.2.0 x2.2xBudgetAI stagnating or AI emergingAI scaling or AI future-builtRevenue shareinvested indig

26、ital and AI2024 increase in AI/GenAI investments vs 2023Share of FTEs dedicated to digital andAI workShare of digital FTEs dedicated to AI/GenAI roles todayShare of FTEs to be upskilledin AI/GenAITime to market for new digital and AI productsShare of AI/GenAI products scaledacross the organization 5

27、.0%10.1%2.0 x4.6%9.1%1.9x5.1%9.6%1.6x8.9%13.8%8.2%18.2%1.1x6.3months5.5months1.6x12.3%19.2%PeopleInnovation4 WHERES THE VALUE IN AI?They direct their efforts more toward people and processes than toward technology and algorithms.Leaders follow the rule of putting 10%of their resources into algorithm

28、s,20%into technology and data,and 70%into people and processes,which our data shows are the key capabilities underpinning success.They have moved quickly to focus on GenAI.Leaders use both predictive AI and GenAI,and they are faster in adopting GenAI,which opens opportunities in content creation,qua

29、litative reasoning,and connecting other tools and platformsin part because their more advanced capabilities facilitate putting the prerequisites(such as large language models)in place.Incumbents Reap ValueNot all AI leaders are hyperscalers and digital natives,com-panies that include AI as part of t

30、heir product or services offering.More than half of the top-performing 26%,includ-ing the ones described at the beginning of this chapter,are traditional incumbents that have strengthened their capa-bilities and are using them to build differentiated competi-tive advantage.The sectors with the bigge

31、st percentages of AI leaders tend to be those that were among the first to experience digital disruption a decade and half ago and got the earliest start on building digital capabilities.They include fintech(49%are leaders),software(46%),and banking(35%).But AIs impact extends to all industries.For

32、example,a leading automaker used GenAI to accelerate tender docu-ment drafting and adjustments by 50%while improving document quality and consistency.GenAI also increased the automakers speed in analyzing competing offers(by 50%)and reduced the time necessary to search knowledge assets(by 50%to 75%)

33、.Leaders are blazing the AI trail,but other companies can catch up if they take a page from the leaders playbook and focus on the areas that offer them the best opportunities and on the capabilities they need to build in order to capi-talize.We explore these factors in the next two chapters.Exhibit

34、3-Leaders Integrate AI with Broader Cost Transformation Efforts and Have a Greater Focus on RevenueIntegration of AI with broader cost transformation efforts(%)AI stagnatingAI emergingAI leadersAI stagnatingAI emergingAI leadersAI investment split between cost reduction and revenue growth(%)47524427

35、2120262736533017More AIintegrationGreaterrevenuefocus43455401543Without GenAIExploratoryMultiple functionsRevenueEqualCostSource:BCG Build for the Future 2024 Global Study(merged with DAI).BOSTON CONSULTING GROUP 5Leading companies are dreaming big.By 2027,the top 26%of companies in our survey of AI

36、 maturity expect to achieve 45%more value via cost reduction and 60%more value via revenue growth than other firms.Even in 2024,leaders expect to realize more than twice the RoI from AI initiatives than other companies do,resulting in a 5%reduction in addressable operational expenses and a 5%increas

37、e in addressable revenues.The common narrative for AI involves support functionsHR,IT,legal,and the likewhere automating relatively low-level and repetitive functions creates significant value.But the companies that are generating the most value are not only deploying productivity plays in support f

38、unctions but also focusing on reshaping their core business process-es and inventing new revenue streams.They are achieving results from AI across a wide range of functions,from R&D to operations and from sales and marketing to customer service.Because they have built the necessary capabilities,they

39、 can more readily identify,pilot,and scale up value-creating use cases.For example,one chemicals company expects to create more than$500 million in value from an end-to-end transformation that will implement AI across operations,site services,and procurement.The Surprising Sources of Value from AI6

40、WHERES THE VALUE IN AI?In the CoreOverall,the companies in our survey derive 62%of the value they obtain from AI and generative AI in core busi-ness functions,including operations(23%),sales and marketing(20%),and R&D(13%).Support functions gener-ate 38%of the value,with customer service(12%),IT(7%,

41、)and procurement(7%)leading the way.In some sectors the spread between core and support is even wider.(See Exhibit 4.)Software,media,fintech,insurance,telecommunications,and biopharma generate 70%to 90%of their AI-related value in core business processes.Although we found wide variation among sector

42、s,the overall results are consistenteven most of the sectors in the bottom quartile generate 40%to 60%of AI value in core processes.Sector MattersCompanies in different sectors also benefit from identify-ing the domains in which AI can produce the most value.Our research shows that they vary widely

43、by industry.(See“AI in Insurance and Biopharma.”)Sales and marketing,for example,is fast emerging as a major source of AI value in such sectors as software(31%of AI value generated),travel and tourism(31%),media(26%),and telecommunications(25%).Specific roles and the scale of impact differ by indust

44、ry,but AI offers compa-nies a near-term opportunity to reshape the sales function with next-best action recommendations,talk tracks,and basic workflow automation.In the medium term,AI and GenAI will enable real-time assisted selling and autono-mous selling via digital sales avatars,with limited huma

45、n involvement.Such automation will permit human staff to focus on strategic and relationship selling,while virtual assistants cover more transactional tasks.As predictive smart selling becomes the norm,traditional silos dividing marketing,sales,and pricing will dissolve.Our experience indicates that

46、 resulting increases in customer lifetime value and go-to-market efficiencies could almost double profit margins.Exhibit 4-To Realize Value from AI,Companies Focus on Core Business Processes,with Sector-Specific VariabilitySource:BCG Build for the Future 2024 Global Study(merged with DAI).Core busin

47、ess functionSupport functionWhere companies are achieving or see business valueSectorsSoftwareCore business processes(%)Global average:62%Support functions(%)MediaFintechInsuranceTelecommunicationsBiopharmaBankingAirlinesRetailAutomotiveTransport and logisticsMedtechConsumer productsOil and gasChemi

48、calsMachines and automationPower,utilities,and renewables9487857771706865636261575949484021613152329303235373839434151526079Leaders are not only deploying productivity plays but reshaping core business processes and inventing new revenue streams.8 WHERES THE VALUE IN AI?The impact on marketing will

49、be equally profound and will encompass four key processes:Insight to Innovation.Automated data collection and analysis will speed identification of market oppor-tunities and increase marketers ability to develop new product design.Concept to Creation.Workflows will accelerate asset creation and feed

50、back loops,seamlessly adapting,local-izing,and disseminating content.Campaign Setup and Execution.Hyper-segmentation and real-time execution that responds to trends and feedback will speed campaign creation and automatical-ly track progress against key objectives.Marketer Productivity.Marketers will

51、 spend less time on time-consuming,repetitive,administrative tasks and more time on strategic decision making.For example,a leading North American telco is already using AI to analyze call recordings to identify opportunities for cost savings and higher customer satisfaction.The company has reduced

52、call center interaction time by 20%and cut call transfers to live agents by 25%.AI-powered chatbots now handle 30%of calls,and the telco expects to reduce total costs in the relevant business unit by 25%.Predictably,AI is having a big impact in R&D in research-intensive sectors such as biopharma(27%

53、of value creat-ed),medtech(19%),and automotive(29%,in an industry undergoing a major transition to software-driven vehicles).A medtech company vice president told us,“Generative AI has allowed us to generate images for training purposes that mimic real diseases that humans can have.We start-ed deep

54、diving into generating thousands of images that arent coming from patients but are being generated by the generative model mimicking real-life cases.Our predictive AI model improved accuracy by 4%to 5%because of this generative AI approach.”In the R&D function of the future,we expect individual-,tea

55、m-,and company-level changes to improve concept R&D,product development and industrialization,and prod-uct evolution.AI will accelerate and automate each step by shortening iteration loops,democratizing access to exper-tise across teams and organizations,fast-tracking explora-tion of new concepts,si

56、mulating product designs,and forecasting procurement orders,among other changes.In one current instance,a global pharmaceuticals compa-ny is using AI to accelerate its drug discovery capabilities.The initial vision was to build,test,and validate an AI prototype with chemists to quantify the value im

57、pact in the discovery workflow.The company assessed the potential of state-of-the-art models to find new preclinical candidates faster,and then it built its own machine learning algorithm to rapidly screen over 1 billion drug compounds and a genetic algorithm to power a lead optimization pipeline fo

58、r molecular chemists.The project generated value of$100 million a year through faster launches,including a 25%reduction in cycle time.The company expanded its library of molecules by 100 times,increasing the visibility of novel compounds to its researchers.Customer service is already a significant s

59、ource of AI-generated value in insurance(24%of the value created)and banking(18%).Companies are using AI to boost pro-ductivity,reducing the need for multiskilled frontline teams and redesigned agent journeys.We are seeing near-term increases of 30%to 40%in productivity and a profit-and-loss impact

60、of 10%to 20%for the function.Ambitions run much bigger.Leading companies expect to realize long-term increases in productivity of up to 60%.The impact of integrating AI into customer service process-es will reverberate throughout the value chain.Customer service functions will be able to preempt iss

61、ues and self-heal by fixing problems before customers detect them,and they will enable customers to resolve their own issues through self-help.If the customer still needs human assis-tance,AI will support the agents response with augmented capabilities such as optimizing the conversation in real-tim

62、e by considering the customers needs in context and making offers where relevant.A leading international bank needed to modernize its customer management system to improve service quality,reduce operational costs,and enhance revenue generation.It turned to GenAI to reshape both customer interactions

63、 and backend processes,including deploying GenAI for chat support,enhancing agent efficiency,improving service quality,and increasing conversion rates.It also integrated GenAI into its APIs and apps for smooth and scalable opera-tions.Results included a reduction of almost 20%in interac-tion time be

64、tween customers and agents;a drop of 4 min-utes in average service time while retaining similar levels of customer satisfaction,an increase of 28 points in conversion rates,and a doubling in breadth of products sold.BOSTON CONSULTING GROUP 9Consumer products and retail companies are making big gains

65、 with AI-driven personalization(19%of the value created for the former and 22%for the latter).About 30%of consumer companies in our survey have adopted AI for personalized marketing(among other functions)and are seeing productivity gains of about 30%from such activities as marketing content generati

66、on,marketing mix and RoI optimization,and data-driven digital marketing.As a re-sult,leaders are doubling down in other areas at two to four times the rate of slower movers,applying AI to genera-tive product design,and manufacturing optimization.Within each process or function,its critical to define

67、 spe-cific use cases and associated business value.In most sectors,more than half of GenAIs value potential lies in two or three functional domains.In insurance,55%of the value lies in in policy administration,underwriting,and claims management.In biopharma,57%of the value is found in R&D and in sal

68、es and marketing.The critical challenge for companies is to identify the key use cases within each function.For example,43%of insur-ance companies leverage AI in scoring,fraud assessment,and triage while 42%of biopharma companies use AI in systematic protein and drug molecule generation(at least for

69、 pilots and proofs of concept).The highest value use cases typically involve a mix of predictive AI and GenAI.Although companies in each sector may be generating the greatest value from use cases in one or two do-mains,most are still experimentingand obtaining mea-surable results in up to half a doz

70、en domains in the core business,including customer relations and experience,content production and management,and product man-agement.In more than a few sectorsincluding oil and gas,utilities,and machinery and automationsupport functions are a significant source sources of value,too.There are many r

71、outes to value.Chapter 3 explores how your company can efficiently find its most productive paths.10 PUBLICATION TITLEAt the business process,function,and use-case level,value creation from AI is already taking different directions in different sectors,highlighting the importance to each com-pany of

72、 independently identifying where its best opportuni-ties lie.Consider the evidence that our survey gathered in two very different sectors:insurance and biopharma.The average AI maturity of both sectors falls in the middle of the maturity curve,not far off the all-sector average.Companies in both sec

73、tors generate an average of 70%or more of AI value from core business processes and 30%or less from support functions.But the similarities end there.InsuranceInsurers are focusing on operations(policy administration,underwriting,and claims management),customer service,and marketing and sales.(See th

74、e AI factsheet for insur-ance.)So far,the widest adoption of predictive AI at the individual-opportunity level has occurred in the areas of scoring,fraud assessment,and triage and policy automa-tion.Adoption of GenAI is strongest in the use of chatbots to resolve questions and summarize customer int

75、eractions.In line with their overall scores,insurers biggest challenges involve people and processes:improving staff AI literacy,prioritizing opportunities over other concerns,and estab-lishing RoI for identified opportunities.They also wrestle with the tasks of integrating AI with existing IT syste

76、ms and of increasing the accuracy and reliability of AI models.An Asian life and health insurance company with a strong track record in digital transformation sought to demon-strate the benefits that GenAI could have on its operations by identifying and executing a couple of high-impact,high-use cas

77、es.The insurer prioritized the possibilities on the basis of a high-level analysis of potential impact.It select-ed two opportunities,one in customer-service call center operations and the other in sales and marketing.The former achieved a 30%reduction in call center search times and the latter a 30

78、%to 40%reduction in marketing and sales material creation time.AI in Insurance and BiopharmaBOSTON CONSULTING GROUP 11BiopharmaBiopharma tells a different story.More than half of the value in this sector comes from commercial/sales and marketing(30%),and R&D(27%).Biopharma companies are using GenAI

79、for systematic protein,drug,and biological processes generation,real-time hyperpersonalized engage-ment with health care practitioners,and personalized outreach to patients and providers.They are using AI and GenAI together for analyzing and documenting customer interactions and for targeting patien

80、t identification via biological data.(See the AI factsheet for biopharma.)Once again,the biggest challenges in applying the technol-ogy relate to people and processes:prioritizing opportuni-ties over other concerns,advancing staff AI literacy,acquir-ing available talent and skills,and establishing R

81、oI on identified opportunities.The top algorithm and technology issues involve integrating AI with existing IT systems,and maximizing the accuracy and reliability of models.AI Factsheet for InsuranceKey challengesRespondents citing the challenge(%)Where does insurance stand on the AI maturity curve?

82、Main challengesTop challenges across people and processes,technology,and algorithmsMaturity stage(%of companies)Global averageInsurance averageInsurance companies have emerging AI capabilities slightly aheadof the global averageWhere are the value pools in my sector?Distribution of AI value potentia

83、l along functional domains(%)AI stagnating9AI emerging64AI scaling25AI future-built2Lack of accurate/reliable models150255075100253545556575Lack of access to high-quality dataDifficulty integrating with existing IT systemsDifficulty ensuring security and complianceIT budgets limiting investments in

84、AIInsufficient platform capabilities forat-scale testingInsufficient AI literacyDifficulty prioritizing opportunities vs other concernsDifficulty establishing RoI on identified opportunitiesDifficulty reimagining workflows and processesLack of specialized AI engineersLack of available talent and ski

85、llsDifficulty measuring predetermined KPIsDifficulty sequencing opportunities intoa roadmapWeak governance structures to steerresponsible AIDifficulty identifying short-andlong-term next stepsFocus areasBCGs 10-20-7modelAlgorithms10%Technology20%People andprocesses70%Core business functionsCustomer

86、serviceand policyadministration24Underwriting16Claimsmanagement15Productmanagement9Marketing,sales,distribution1377Support functions23HRITLegalProcurementFinance54446Source:BCG Build for the Future 2024 Global Study(merged with DAI).12 WHERES THE VALUE IN AI?AI Factsheet for BiopharmaGlobal averageH

87、ealth care averageBiopharma averageKey challengesRespondents citing the challenge(%)Where does biopharma stand on the AI maturity curve?Maturity stage(%of companies)Biopharma companies have emerging AI capabilities on apar with the global averageWhere are the value pools in my sector?Distribution of

88、 AI value potential along functional domains(%)AI stagnating27AI emerging46AI scaling19AI future-built8Lack of accurate/reliable models150255075100253545556575Lack of access to high-quality dataDifficulty integrating with existing IT systemsDifficulty ensuring security andcomplianceInsufficient plat

89、form capabilities forat-scale testingDifficulty prioritizing opportunities vs other concernsInsufficient AI literacyLack of available talent and skillsDifficulty establishing RoI on identified opportunitiesLack of leadership alignment,communications,and behavior modelingLack of specialized AI engine

90、ersDifficulty making a business case forscaling initiativesLack of a clear AI case for changeDifficulty identifying short-andlong-term next stepsDifficulty reimagining workflowsand implementing processesFocus areasBCGs 10-20-7modelAlgorithms10%Technology20%People andprocesses70%Core business functio

91、nsCommercial/salesand marketing30Manufacturing13Research anddevelopment2770Support functions30FinanceITHRLegal6Customerservice77433ProcurementMain challengesTop challenges across people and processes,technology,and algorithmsSource:BCG Build for the Future 2024 Global Study(merged with DAI).BOSTON C

92、ONSULTING GROUP 13Leading companies are well on their way to creating significant value and advantage from AI.For example,a consumer products company applied GenAI to re-duce costs by$300 million through productivity gains and agency cost savings.A global consumer goods company expects to generate$1

93、00 million in additional sales from a GenAI-powered virtual conversational assistant,the first in its sector.A North American telco achieved a 10%reduc-tion in call handling time and cut the cost of customer retention by more than 30%,leading to$200 million in annualized savings.Meanwhile,the 70%of

94、companies that are struggling,wait-ing,planning,and experimenting have an urgent need to accelerate their efforts to overcome barriers and catch up as their competitors improve their productivity,revenues,and customer experience.As leaders and aspiring leaders ex-pand their AI capabilities and as Ge

95、nAI models and tools mature,less capable companies will fall farther behind.Heres an AI playbook that all companies can follow.The Playbook for Winning with AI14 WHERES THE VALUE IN AI?Overcoming Tough ChallengesOur survey highlights the most difficult challenges that companies face in implementing

96、AI initiatives.They fall into four groups:Difficulties in defining clear priority use cases with com-pelling returns for the anticipated investments A host of issues related to moving from plans to action and delivering value,such as prioritizing investments,scaling solutions across functions and bu

97、sinesses,over-coming resistance to adoption,and realizing the benefits People and skills issues,including building specific AI skills and broader AI literacy Integrating AI solutions with existing IT systems,and enabling access to high-quality data Our experience,corroborated by our new research,ind

98、i-cates that about 70%of the challenges relate to people and process,about 20%are technology issues,and only 10%involve AI algorithms(which often occupy a lot more organizational time and resources).(See Exhibit 5.)The survey confirms our long-held view that when companies undertake digital or AI tr

99、ansformations,they need to focus 70%of their effort and resources on people-related capabil-ities,20%on technology,and 10%on algorithms.Too often,companies make the mistake of prioritizing the technical issues over the human oneswhich helps explain why many of them do not achieve the results they ar

100、e looking for.Challenges evolve over time,of course,as companies build their capabilities.But while less AI-capable companies focus on getting the basics right,leaders are more con-cerned with ensuring security and compliance,implement-ing responsible AI,and resolving technical issues such as guardr

101、ails for large language models,high model latency,and run costs.Exhibit 5-The Biggest Challenges Relate to People and Processes,Such as Prioritizing Opportunities and Establishing RoIKey challengesRespondents citing the challenge(%)Lack of accurate/reliable models2025303540455055606570Lack of access

102、 to high-quality dataDifficulty integrating with existing IT systemsIT budgets limiting investments in AIDifficulty ensuring security and complianceExpensive scaling due to high model run costsDifficulty establishing RoI on identified opportunitiesDifficulty prioritizing opportunities vs other conce

103、rnsDifficulty making a business case for scaling initiativesDifficulty realizing cost takeout/savingsResistance and fear that AI will impact jobsLack of a clear AI case for changeDifficulty measuring predetermined KPIsLack of leadership alignment and communicationsDifficulty reimagining workflows an

104、d processesInsufficient AI literacyLack of specialized AI engineersFocus areasBCGs 10-20-70 modelAlgorithms10%48%43%56%48%46%37%66%59%56%54%48%42%38%37%37%37%37%Technology20%People andprocesses70%Source:BCG Build for the Future 2024 Global Study(merged with DAI);n=1,000.BOSTON CONSULTING GROUP 15The

105、 Capabilities Required for SuccessWe analyzed the self-reported capabilities of AI leaders compared with those of other companies.This assessment revealed empirical evidence about the most important capabilities for implementing AI at scale.Most relate to peoples and processeschange management,produ

106、ct development skills,and workflow capabilities such as new technologies,role clarity,process reimagination,AI talent,and responsible AI governance.(See Exhibit 6.)The most important technology capabilities are related to data and platforms,and the most important algorithm capability is AI model qua

107、lity and performance.As a senior executive of a leading AI player said,“We strongly believe that the key capabilities for success revolve around talent and process excellence.You need to have specific skills,such as data science,general enthusiasm for innovation,and the ability to reimagine and impl

108、ement new approaches.The AI technology is amazing,but we try not to get dazzled by it.”Jump-Starting Your Journey to AI ValueAfter assessing the capabilities and approaches of the lead-ing companies,we have compiled a playbook for how any company can drive value quickly and effectively from AI.The a

109、pproach has seven critical steps:1.Set a bold strategic commitment from the top,and be prepared to support it over multiple years.2.Maximize the potential value of AI with a balanced portfolio of initiatives that include streamlining everyday business processes,transforming entire business func-tion

110、s,and developing AI-native offerings that unlock new business models.3.Focus on fewer but higher-impact lighthouse programs,starting with implementation with one to three high-RoI,easy-to-implement initiatives to fund the journey.BCGs 10-20-70 modelRelative importance of capabilities1Data science ca

111、pabilitiesto develop andimplement algorithmsModel quality and performanceData analyticsData managementAI platformsCybersecurityAI toolsSecure ML/LLM operationsData security and protectionThird-party risk managementChange managementProduct development pipeline and cyclesAdoption of emerging technolog

112、iesRoles and responsibilitiesProcess reimaginationAI talentResponsible AI governanceRisk-informed cultureAI model guardrailsAI implementation guardrailsInnovative cultureData governanceProduct/platform orientationAI strategyFurther capabilities2Scalable and modernizedstack that supportsbusiness need

113、sEffective processessupported bytalent and changemanagement practicesAlgorithms10%Technology20%People andprocesses70%8%4.9%3.5%7.2%4.8%4.4%2.2%1.9%1.0%0.4%8.4%7.1%7.0%6.9%5.5%5.1%5.1%5.1%3.6%3.1%3.0%3.0%2.6%2.4%2.0%22%70%Exhibit 6-To Get an AI Transformation Right,70%of the Focus Should Be on People

114、 and ProcessesSource:BCG 2024 Global Study on AI and Digital maturity;n=1,000.Note:LLM=large language model;ML=machine learning.1Based on regression against probability of being an AI and GenAI value creator,defined as the average of expected cost savings and revenue uplift from AI and GenAI initiat

115、ives being 5%.2“Further capabilities”summarizes all capabilities that fall into the“people and processes”category but individually received an importance score of less than 2%.16 WHERES THE VALUE IN AI?4.Ensure that the minimal viable infrastructure required for these initiatives exists,especially w

116、ith respect to inte-gration with IT systems and access to quality data.5.Identify your companys capability gaps vis-vis the leaders in the known critical capabilities for success,and invest in parallel to build these capabilities.It may be necessary initially to focus on issues related to tech-nolog

117、y and data,but capabilities involving people and processes are critical and demand close and prolonged attention.6.Ensure that implementation governance focuses on end-to-end transformation and on people and processes,including redesigning ways of working,cultivating talent,reimagining processes,str

118、engthening effective decision making,and addressing reluctance to adopting new solutions.7.Set up guardrails to deploy AI responsibly in all initia-tives through transparency,control,and accountability to ensure ethical and legal compliance and to manage business risks.How Two Companies Applied the

119、Playbook for Success Heres how two big companies in very different industries and states of maturity used this playbook to jump-start their AI journeys.A global financial institution is applying GenAI to increase value of its data by boosting productivity and consistency in data governance,improving

120、 the employee data-management work experience,and reimagining its data management operation.Its ability to use data to create business value was constrained by governance procedures that covered only 10%of its available data.Scalability was limited by manual processes to generate metadata and captur

121、e data lineage.The company turned to GenAI to automate and optimize data management processes,such as capturing data lin-eage,generating business metadata,and tagging sensitive data elements.It implemented the project in three waves of use-case deployment:building foundations,enriching context,and g

122、enerating insights.The first phase involved a ten-week proof of concept to test feasibility and demon-strate value using automated metadata labeling and accel-erating the capture of cross-system data lineage(source and quality)information.On the basis of the lessons learned in the pilots,the company

123、 embarked on a multi-month journey to scale the solution in production and to pilot it within critical data domains to measure the impact delivered and determine the further investment required for expansion.So far,the company has built the production-grade solution and plan to deploy,and it has ach

124、ieved a productivity gain on the order of 40%to 70%in specific activities such as metadata generation and lineage creation.It realized a net productivity gain of 20%to 25%in onboarding data with end-to-end data governance controls and has accelerated the inclusion of data under governance by more th

125、an five years.It also reimagined the operating model to be imple-mented as part of a pilot program,further accelerating impact delivery across more than 10 data domains and more than 200 data management experts globally.A major European automaker faced heightened competi-tion from rivals in Asia who

126、se required time to market had dropped by as much as 45%.The company was well along its path toward digital transformation and was ready for the next stagea journey to leverage AI to reshape its R&D function and become a next-generation OEM leader.It defined a clear ambition to identify where and ho

127、w to integrate AI to achieve gains in scale and speed,with a particular focus on R&D.The automaker prioritized 11 high-impact changes,including high-fidelity simulation,AI generation of initial designs,and accelerated software performance checking.It then brought its well-established process improve

128、ment capabilities to bear on the challenge.The company so far has reduced time from idea to produc-tion by 30%(equivalent to one year)and saved up to 40%in the industrialization ramp-up of new products.It has also reduced its cost of goods sold by 1.5%to 2%overall.Three-quarters of companies have ye

129、t to generate value from AI.They need to act or risk falling far behind.The good news is that AI leaders are showing the way forward in adopting valuable AI solutions at scale.The myriad challenges are clear,as are the ways to address them.Companies in any sector and at any level of AI maturity can

130、tailor our playbook,which is compiled from a trove of empirical evidence,to their particular needs.They can start by conducting an AI maturity assessmenta focused health check for AI readiness across the organizationto helps the C-suite understand the companys starting point and how to move from pil

131、ots to scale.As AI technologies continue to mature,and as adoption increases,time is of the essence for companies to make rapid progress.BOSTON CONSULTING GROUP 17DefinitionsIn this report,when we refer to AI,generative AI(GenAI),and predictive AI,we are using the definitions detailed below.AI refer

132、s to all artificial intelligence technologies and applications.Predictive AI refers to the use of artificial intelligence products and systems to analyze historical and current data to make predictions about future events or trends.These systems use data analytics,machine learning,and various statis

133、tical algorithms to identify patterns and rela-tionships in data,which can then be used to forecast out-comes with a certain level of probability.GenAI refers to the use of products and programs that can generate new realistic content,such as text and images.Examples include ChatGPT for text generat

134、ion and DALL-E for image generation.Essential to GenAI are foundational models that include large language models(LLMs)a subset of deep-learning algorithms that leverage break-through algorithm development in self-supervised and transfer learning.AppendixDefinitions and Methodology18 WHERES THE VALU

135、E IN AI?MethodologyWe designed our 2024 Build for the Future survey following the AI Tritad,which focuses on the AI capabilities necessary to support strategic objectives,deliver significant business value,and identify and capitalize on new market possibilities.In this context,our comprehensive AI m

136、aturity score is built on 30 enterprise foundational capabilities,each mea-sured along four clearly defined maturity stages.(See the exhibit.)We then applied robust statistical methods to calculate individual weights for each capability,on the basis of their overall contribution to the AI value gene

137、ra-tion that respondents reported.Next we sorted the weight-ed scores into four categories:AI stagnating:score of 025 AI emerging:score of 2550)AI scaling:score of 5075)AI future-built:score of 75100)When we refer to AI leaders in the report,we are combining the top two categories of AI-scaling and

138、AI future-built companies.In our survey,we asked 1,000 CxOs and senior executives across more than 20 sectors to estimate their companies AI maturity along the 30 foundational capabilities.In addi-tion,they assessed outcomes in ten dimensions in re-sponse to sector-specific questions.Respondents cam

139、e from 59 countries in Asia,Europe,and North America and from ten industries:consumer goods,energy,financial services,health care,industrial goods,insurance,public sector,technology,media,and telecommunications.Methodology:Underlying Framework Along Enterprise Foundational Capabilities and OutcomesA

140、I and GenAI enterprise foundational capabilitiesOutcomesCustomer experienceGovernanceTalent and skillsCulture and changeOperating modelGenAI pricingCustomer journeyCustomer serviceDigital marketingNext-generation salesSecure ML/LLM operationsProduct/platform orientationRoles and responsibilities Par

141、tnership ecosystemC-suite expertiseTalent sourcing and skills planAI talentInnovative cultureRisk-informed cultureChange managementProcess reimaginationAI and GenAI strategy Data governance Responsible AI Tech innovationOperationsDigital supply chainDigital support functionIndustry 4.0ProcurementSer

142、vice process reimaginationAI delivery officeAI and GenAI portfolio Partner/vendor selection AI deployment guardrailsRapid ideation and testing New product buildCybersecurityRisk and responsible AIpractices and toolsData and AI/GenAI platformData analyticsData managementAI and GenAI platformsModel qu

143、ality performanceThird-party risk managementData security and protectionCybersecurity,including AIand GenAIAI and GenAI toolsGenAI compliance(RAI)GenAI model guardrailsSource:BCG Build for the Future 2024 Global Study(merged with DAI),n=1,000.Note:LLM=large language model;ML=machine learning;RAI=res

144、ponsible AI.BOSTON CONSULTING GROUP 19About the Authors Nicolas de Bellefonds is a managing director and senior partner in the Paris office of Boston Consulting Group.You may contact him by email at .Marc Roman Franke is a partner and associate director,AI and digital transformation,in BCGs Berlin o

145、ffice.You may contact him by email at .Michael Grebe is a managing director and senior partner in BCGs Munich office.You may contact him by email at .Tauseef Charanya is an offer director,(Gen)AI and digital transformation,in the firms Austin office.You may contact him by email at .Patrick Forth is

146、a senior advisor and senior partner emeritus in the firms Sydney office.You may contact him by email at .Romain de Laubier is a managing director and senior partner in the firms Singapore office.You may contact him by email at .20 WHERES THE VALUE IN AI?Vlad Lukic is a managing director and senior p

147、artner in BCGs Boston office.You may contact him by email at .Clemens Nopp is an offer senior manager,AI and digital strategy,in BCGs Vienna office.You may contact him by email at .Amanda Luther is a managing director and partner in the firms Austin office.You may contact her by email at .Joe Sassin

148、e is a project leader in the firms New York office.You may contact him by email at .For Further ContactIf you would like to discuss this report,please contact the authors.AcknowledgmentsThe authors would like to thank Abhinaba Dam,Ignacio Gonzalez,Anagha Kumar,Michael Leyh,Clementine Remy,and Julia

149、Tristan for their valuable contributions to this report.Boston Consulting Group partners with leaders in business and society to tackle their most important challenges and capture their greatest opportunities.BCG was the pioneer in business strategy when it was founded in 1963.Today,we work closely

150、with clients to embrace a transformational approach aimed at benefiting all stakeholdersempowering organizations to grow,build sustainable competitive advantage,and drive positive societal impact.Our diverse,global teams bring deep industry and functional expertise and a range of perspectives that q

151、uestion the status quo and spark change.BCG delivers solutions through leading-edge management consulting,technology and design,and corporate and digital ventures.We work in a uniquely collaborative model across the firm and throughout all levels of the client organization,fueled by the goal of help

152、ing our clients thrive and enabling them to make the world a better place.For information or permission to reprint,please contact BCG at .To find the latest BCG content and register to receive e-alerts on this topic or others,please visit .Follow Boston Consulting Group on Facebook and X(formerly known as Twitter).Boston Consulting Group 2024.All rights reserved.10/24 22 WHERES THE VALUE IN AI?

友情提示

1、下載報告失敗解決辦法
2、PDF文件下載后,可能會被瀏覽器默認打開,此種情況可以點擊瀏覽器菜單,保存網頁到桌面,就可以正常下載了。
3、本站不支持迅雷下載,請使用電腦自帶的IE瀏覽器,或者360瀏覽器、谷歌瀏覽器下載即可。
4、本站報告下載后的文檔和圖紙-無水印,預覽文檔經過壓縮,下載后原文更清晰。

本文(BCG:2024AI價值研究報告(英文版)(24頁).pdf)為本站 (白日夢派對) 主動上傳,三個皮匠報告文庫僅提供信息存儲空間,僅對用戶上傳內容的表現方式做保護處理,對上載內容本身不做任何修改或編輯。 若此文所含內容侵犯了您的版權或隱私,請立即通知三個皮匠報告文庫(點擊聯系客服),我們立即給予刪除!

溫馨提示:如果因為網速或其他原因下載失敗請重新下載,重復下載不扣分。
客服
商務合作
小程序
服務號
折疊
午夜网日韩中文字幕,日韩Av中文字幕久久,亚洲中文字幕在线一区二区,最新中文字幕在线视频网站