《麥肯錫：什么最重要？2024年企業領導者的八大優先事項（英文版）（95頁）.pdf》由會員分享，可在線閱讀，更多相關《麥肯錫：什么最重要？2024年企業領導者的八大優先事項（英文版）（95頁）.pdf（95頁珍藏版）》請在三個皮匠報告上搜索。

1、What matters most?Eight priorities for CEOs in 2024Forming the right agenda is not getting easier.Heres our annual attempt to cut through the clutter and zero in on things that matter to CEOs.CompendiumDecember 2023Cover image:Wang Yukun/Getty Images All interior images:Getty ImagesCopyright 2023 Mc

2、Kinsey&Company.All rights reserved.This publication is not intended to be used as the basis for trading in the shares of any company or for undertaking any other complex or significant financial transaction without consulting appropriate professional advisers.No part of this publication may be copie

3、d or redistributed in any form without the prior written consent of McKinsey&Company.What matters most?Its a question weve been investigating for a few years now(here are reports from 2022 and 2021).This year,were reminded that what matters most are family,friends,values,principles,and commitments.O

4、ne of our commitments is to CEOs.Its a tough job and getting tougher all the time.Just in the past few years,theyve had to cope with a global pandemic,busted supply chains,war,stubborn inflation,and many other disruptions.Any one of these is enough to derail a CEOs agenda.Taken together,its the most

5、 difficult operating environment we can remember.We talk to hundreds of CEOs every year,and many of our colleagues do the same.We admire how CEOs are leading their companies for the benefit of all stakeholders.Weve consolidated the views that have come out of these conversations and are pleased to o

6、ffer what weve heard about how companies can do better for society,communities,and employeesand the prosaic business of how they can pay for it all and reward investors too.Here are eight priorities for CEOs in 2024.Generative AI goes from proof of concept to scaleThe biggest story of this year(or d

7、ecade)was the arrival of generative AI(gen AI).This is the real deal,folks.Thousands of companies in every industry and in every part of the world are already using a simple gen AI interface to radically transform every imaginable business activity.But while innovators dominate headlines,its scalers

8、 that dominate markets.CEOs need to figure out three things,posthaste:which parts of the business can benefit,how to scale from one application to many,and how the new tools will reshape their industry.How to outcompete with technologyGen AI grabs all the headlines,but lets not forget the“digital re

9、volution,”if we can so describe something that started 30 or 40 years ago.Digitization might be on a slow boil,but given enough time,all the frogs will still be cooked.And theres a risk that paying too much attention to gen AI could set a company back on its digital transformation.How to escape the

10、boiling pot?This year,our colleagues published a best-selling book,Rewired:The McKinsey Guide to Outcompeting in the Age of Digital and AI.Its a collection of our best insights for digitizing the enterprise.Digital winners grow revenues and cut costs faster than others.The biggest capital reallocati

11、on in our lifetimeThats what we said last year about the energy transition.The bill has only gone up since then,for the simple reason that amid uncertainty,investors and companies have held back from committing their capital,even as the Earth grows hotter.Lets be clear:what needs to happen is the cr

12、eation of thousands of new green-technology businesses,in every part of the emerging business system.We have ideas about where,how,and when companies should invest.The road to growthIts a funny thing:growth is always job one for CEOs,but the path to get there is never clear.Sometimes its about seizi

13、ng market share;sometimes its about expanding into new markets;sometimes its about making a left turn into something completely new.The one constant is the ten rules of growth.How will the rules play out in 2024?For many,it will mean rule four:turbocharge your core,by using technology to power growt

14、h.For others,it might mean rule six:grow where you know,by improving sales productivity.And,as always,the most constant of all is rule nineacquire programmaticallyas IntroductioniWhat matters most?Eight priorities for CEOs in 2024the latest installment of our 20-year research effort demonstrates.Wha

15、ts your superpower?Think of any company you admire,and you can likely rattle off one or two superpowers that make it uniquely successful.Toyota and its Toyota Production System.LVMH and its exquisite craftsmanship and the entrepreneurship of its brand leaders.Disney and imaginative customer experien

16、ces.A distinctive capability can lift a company out of the mire of clogged,commoditized markets and onto the high ground of outperformance.Exceptional implementation is part and parcel of building a new capability.Learn to love your middle managersWaffle House,an American restaurant chain,is famous

17、for never closing;some say its doors have no locks.It should also be famous for its management philosophy.The restaurants grill operators are the stars of the show;after years of training,the best get to be called“Elvis of the grill.”After that,they dont get promoted;how do you top being King?But mo

18、st other companies would likely promote such people into senior management roles that they dont want and are not suited for.Companies need to rethink their philosophy about middle managers and recognize them for what they actually are:the core of the company.Geopolitics:Beating the oddsAs Niels Bohr

19、 once said,its very hard to make predictions,especially about the future.As CEOs watch the changes unfolding in the global geopolitical order,all agree with the sentiment.What comes next?One thing is for sure:events have an uncanny way of defying the expectations of experts.In the face of that,manag

20、ement teams and boards should consider black swans and gray rhinos in their scenarios and build geopolitical resilience that will serve them well,no matter which side of the coin comes up.A new lens on the macroeconomyNearly four years after COVID-19 rewrote history,some CEOs are still waiting for m

21、acroeconomic certainty.Thats unlikely to happenand thats okay.Leading firms capitalize on uncertainty:they assess their risk appetite,then invest near the bottom of cycles.Most rely on scenario planning,not least because the exercise usually reveals the core actions that companies need to take no ma

22、tter which way the economy trends.CEOs might want to populate their models with the new scenarios weve developed to look at the ways the global balance sheet might develop.Over the past two decades,assets on the global balance sheet grew much faster than GDPthe real economy.But the continuation of t

23、hat trend is uncertain.Yet another curveball is the rapid shift of assets from the banking system to private markets and what that means for public companies.CEOs need a broad range of contradictory perspectives:outside in and inside out,a telescope to see the world and a microscope to break it down

24、,a snapshot view of the immediate issues and a time-lapse series to see into the future.We hope this article and the in-depth readings available within it give CEOs and executives the clarity they seek.iiWhat matters most?Eight priorities for CEOs in 2024Gen AI:The start of something bigInnovators d

25、ominate headlines.Scalers dominate markets.1 Technologys generational moment with generative AI:A CIO and CTO guideOutcompeting with technologyMost digital transformations fail to deliver the expected impact.Our playbook has helped hundreds complete their reinvention and find the missing value.13 Re

26、wired to outcompete (McKinsey Quarterly)The energy transition:Time is shortCommitments are abundant;actions,not so much.Leaders can make the next moves.22 Full throttle on net zero:Creating value in the face of uncertainty28 Scaling green businesses:Next moves for leaders35 Decarbonize and create va

27、lue:How incumbents can tackle the steep challengeNavigating the road to courageous growthAll roads do not lead to Rome.Leaders need to find the path that works for them.45 The ten rules of growth57 Five paths to TSR outperformanceWhats your superpower?Institutional capability building:sounds dull,ri

28、ght?Its anything but.Find out how distinctive companies get that way.63 Whats your superpower?How companies can build an institutional capability to achieve competitive advantageLearn to love your middle managersStop thinking of middle management as a way station.Instead,make it a destination.69 Mid

29、dle managers are the heart of your company (McKinsey Quarterly)Geopolitics:Beating the oddsA new world order might be emerging from the current upheaval.Leaders can anticipate some shifts and position their companies for success.76 Black swans,gray rhinos,and silver linings:Anticipating geopolitical

30、 risks(and openings)A new lens on the macroeconomyHard landing,soft landing,no landing at all?While analysts debate,smart leaders are moving on to consider the bigger picture.81 Why the path of global wealth and growth matters for strategy(McKinsey Quarterly)Contents14725836iiiWhat matters most?Eigh

31、t priorities for CEOs in 2024Technologys generational moment with generative AI:A CIO and CTO guideCIOs and CTOs can take nine actions to reimagine business and technology with generative AI.This article is a collaborative effort by Aamer Baig,Sven Blumberg,Eva Li,Douglas Merrill,Adi Pradhan,Megha S

32、inha,Alexander Sukharevsky,and Stephen Xu,representing views from McKinsey Digital.July 2023Hardly a day goes by without some new business-busting development related to generative AI surfacing in the media.The excitement is well deservedMcKinsey research estimates that generative AI could add the e

33、quivalent of$2.6trillion to$4.4 trillion of value annually.CIOs and chief technology officers(CTOs)have a critical role in capturing that value,but its worth remembering weve seen this movie before.New technologies emergedthe internet,mobile,social mediathat set off a melee of experiments and pilots

34、,though significant business value often proved harder to come by.Many of the lessons learned from those developments still apply,especially when it comes to getting past the pilot stage to reach scale.For the CIO and CTO,the generative AI boom presents a unique opportunity to apply those lessons to

35、 guide the C-suite in turning the promise of generative AI into sustainable value for the business.Through conversations with dozens of tech leaders and an analysis of generative AI initiatives at more than 50 companies(including our own),we have identified nine actions all technology leaders can ta

36、ke to create value,orchestrate technology and data,scale solutions,and manage risk for generative AI(see sidebar,“A quick primer on key terms”):1.Move quickly to determine the companys posture for the adoption of generative AI,and develop practical communications to,and appropriate access for,employ

37、ees.A quick primer on key termsGenerative AI is a type of AI that can create new content(text,code,images,video)using patterns it has learned by training on exten-sive(public)data with machine learning(ML)techniques.Foundation models(FMs)are deep learning models trained on vast quantities of unstruc

38、tured,unlabeled data that can be used for a wide range of tasks out of the box or adapted to specific tasks through fine-tuning.Examples of these models are GPT-4,PaLM 2,DALLE 2,and Stable Diffusion.Large language models(LLMs)make up a class of foundation models that can process massive amounts of u

39、nstructured text and learn the relationships between words or portions of words,known as tokens.This enables LLMs to generate natural-language text,performing tasks such as summarization or knowledge extraction.Cohere Command is one type of LLM;LaMDA is the LLM behind Bard.Fine-tuning is the process

40、 of adapting a pretrained foundation model to perform better in a specific task.This entails a relatively short period of training on a labeled data set,which is much smaller than the data set the model was initially trained on.This additional training allows the model to learn and adapt to the nuan

41、ces,terminology,and specific patterns found in the smaller data set.Prompt engineering refers to the process of designing,refining,and optimizing input prompts to guide a generative AI model toward producing desired(that is,accurate)outputs.Learn more about generative AI in our explainer“What is gen

42、erative AI”on McK.1 “The economic potential of generative AI:The next productivity frontier,”McKinsey,June 14,2023.2Technologys generational moment with generative AI:A CIO and CTO guide2.Reimagine the business and identify use cases that build value through improved productivity,growth,and new busi

43、ness models.Develop a“financial AI”(FinAI)capability that can estimate the true costs and returns of generative AI.3.Reimagine the technology function,and focus on quickly building generative AI capabilities in software development,accelerating technical debt reduction,and dramatically reducing manu

44、al effort in IT operations.4.Take advantage of existing services or adapt open-source generative AI models to develop proprietary capabilities(building and operating your own generative AI models can cost tens to hundreds of millions of dollars,at least in the near term).5.Upgrade your enterprise te

45、chnology architecture to integrate and manage generative AI models and orchestrate how they operate with each other and existing AI and machine learning(ML)models,applications,and data sources.6.Develop a data architecture to enable access to quality data by processing both structured and unstructur

46、ed data sources.7.Create a centralized,cross-functional generative AI platform team to provide approved models to product and application teams on demand.8.Invest in upskilling key rolessoftware developers,data engineers,MLOps engineers,and security expertsas well as the broader nontech workforce.Bu

47、t you need to tailor the training programs by roles and proficiency levels due to the varying impact of generative AI.9.Evaluate the new risk landscape and establish ongoing mitigation practices to address models,data,and policies.1.Determine the companys posture for the adoption of generative AIAs

48、use of generative AI becomes increasingly widespread,we have seen CIOs and CTOs respond by blocking employee access to publicly available applications to limit risk.In doing so,these companies risk missing out on opportunities for innovation,with some employees even perceiving these moves as limitin

49、g their ability to build important new skills.Instead,CIOs and CTOs should work with risk leaders to balance the real need for risk mitigation with the importance of building generative AI skills in the business.This requires establishing the companys posture regarding generative AI by building cons

50、ensus around the levels of risk with which the business is comfortable and how generative AI fits into the businesss overall strategy.This step allows the business to quickly determine company-wide policies and guidelines.Once policies are clearly defined,leaders should communicate them to the busin

51、ess,with the CIO and CTO providing the organization with appropriate access and user-friendly guidelines.Some companies have rolled out firmwide communications about generative AI,provided broad access to generative AI for specific user groups,created pop-ups that warn users any time they input inte

52、rnal data into a model,and built a guidelines page that appears each time users access a publicly available generative AI service.2.Identify use cases that build value through improved productivity,growth,and new business modelsCIOs and CTOs should be the antidote to the“death by use case”frenzy tha

53、t we already see in many companies.They can be most helpful by working with the CEO,CFO,and other business leaders to think through how generative AI challenges existing business models,opens doors to new ones,and creates new sources of value.With a deep understanding of the technical possibilities,

54、3Technologys generational moment with generative AI:A CIO and CTO guidethe CIO and CTO should identify the most valuable opportunities and issues across the company that can benefit from generative AIand those that cant.In some cases,generative AI is not the best option.McKinsey research,for example

55、,shows generative AI can lift productivity for certain marketing use cases(for example,by analyzing unstructured and abstract data for customer preference)by roughly 10 percent and customer support(for example,through intelligent bots)by up to 40 percent.The CIO and CTO can be particularly helpful i

56、n developing a perspective on how best to cluster use cases either by domain(such as customer journey or business process)or use case type(such as creative content creation or virtual agents)so that generative AI will have the most value.Identifying opportunities wont be the most strategic taskthere

57、 are many generative AI use cases out therebut,given initial limitations of talent and capabilities,the CIO and CTO will need to provide feasibility and resource estimates to help the business sequence generative AI priorities.Providing this level of counsel requires tech leaders to work with the bu

58、siness to develop a FinAI capability to estimate the true costs and returns on generative AI initiatives.Cost calculations can be particularly complex because the unit economics must account for multiple model and vendor costs,model interactions(where a query might require input from multiple models

59、,each with its own fee),ongoing usage fees,and human oversight costs.3.Reimagine the technology function Generative AI has the potential to completely remake how the tech function works.CIOs and CTOs need to make a comprehensive review of the potential impact of generative AI on all areas of tech,bu

60、t its important to take action quickly to build experience and expertise.There are three areas where they can focus their initial energies:Software development:McKinsey research shows generative AI coding support can help software engineers develop code 35 to 45percent faster,refactor code 20 to 30

61、percent faster,and perform code documentation 45 to 50 percent faster.Generative AI can also automate the testing process and simulate edge cases,allowing teams to develop more-resilient software prior to release,and accelerate the onboarding of new developers(for example,by asking generative AI que

62、stions about a code base).Capturing these benefits will require extensive training(see more in action 8)and automation of integration and deployment pipelines through DevSecOps practices to manage the surge in code volume.Technical debt:Technical debt can account for 20 to 40 percent of technology b

63、udgets and significantly slow the pace of development.CIOs and CTOs should review their tech-debt balance sheets to determine how generative AI capabilities such as code refactoring,code translation,and automated test-case generation can accelerate the reduction of technical debt.IT operations(ITOps

64、):CIOs and CTOs will need to review their ITOps productivity efforts to determine how generative AI can accelerate processes.Generative AIs capabilities are particularly helpful in automating such tasks as password resets,status requests,or basic diagnostics through self-serve agents;accelerating tr

65、iage and resolution through improved routing;surfacing useful context,such as topic or priority,and generating suggested responses;improving observability through analysis of vast streams of logs to identify events that truly require attention;and developing documentation,such as standard operating

66、procedures,incident postmortems,or performance reports.2 Ibid.3 Begum Karaci Deniz,Martin Harrysson,Alharith Hussin,and Shivam Srivastava,“Unleashing developer productivity with generative AI,”McKinsey,June 27,2023.4 Vishal Dalal,Krish Krishnakanthan,Bjrn Mnstermann,and Rob Patenge,“Tech debt:Reclai

67、ming tech equity,”McKinsey,October 6,2020.4Technologys generational moment with generative AI:A CIO and CTO guide4.Take advantage of existing services or adapt open-source generative AI models A variation of the classic“rent,buy,or build”decision exists when it comes to strategies for developing gen

68、erative AI capabilities.The basic rule holds true:a company should invest in a generative AI capability where it can create a proprietary advantage for the business and access existing services for those that are more like commodities.The CIO and CTO can think through the implications of these optio

69、ns as three archetypes:Takeruses publicly available models through a chat interface or an API,with little or no customization.Good examples include off-the-shelf solutions to generate code(such as GitHub Copilot)or to assist designers with image generation and editing(such as Adobe Firefly).This is

70、the simplest archetype in terms of both engineering and infrastructure needs and is generally the fastest to get up and running.These models are essentially commodities that rely on feeding data in the form of prompts to the public model.Shaperintegrates models with internal data and systems to gene

71、rate more customized results.One example is a model that supports sales deals by connecting generative AI tools to customer relationship management(CRM)and financial systems to incorporate customers prior sales and engagement history.Another is fine-tuning the model with internal company documents a

72、nd chat history to act as an assistant to a customer support agent.For companies that are looking to scale generative AI capabilities,develop more proprietary capabilities,or meet higher security or compliance needs,the Shaper archetype is appropriate.There are two common approaches for integrating

73、data with generative AI models in this archetype.One is to“bring the model to the data,”where the model is hosted on the organizations infrastructure,either on-premises or in the cloud environment.Cohere,for example,deploys foundation models on clients cloud infrastructure,reducing the need for data

74、 transfers.The other approach is to“bring data to the model,”where an organization can aggregate its data and deploy a copy of the large model on cloud infrastructure.Both approaches achieve the goal of providing access to the foundation models,and choosing between them will come down to the organiz

75、ations workload footprint.Makerbuilds a foundation model to address a discrete business case.Building a foundation model is expensive and complex,requiring huge volumes of data,deep expertise,and massive compute power.This option requires a substantial one-off investmenttens or even hundreds of mill

76、ions of dollarsto build the model and train it.The cost depends on various factors,such as training infrastructure,model architecture choice,number of model parameters,data size,and expert resources.Each archetype has its own costs that tech leaders will need to consider(Exhibit 1).While new develop

77、ments,such as efficient model training approaches and lower graphics processing unit(GPU)compute costs over time,are driving costs down,the inherent complexity of the Maker archetype means that few organizations will adopt it in the short term.Instead,most will turn to some combination of Taker,to q

78、uickly access a commodity service,and Shaper,to build a proprietary capability on top of foundation models.5.Upgrade your enterprise technology architecture to integrate and manage generative AI modelsOrganizations will use many generative AI models of varying size,complexity,and capability.To gener

79、ate value,these models need to be able to work both together and with the businesss existing systems or applications.For this reason,building a separate tech stack for generative AI creates more complexities than it solves.As an example,we can look at a consumer querying customer service at a travel

80、 company to resolve a booking issue(Exhibit2).In interacting with the customer,the generative AI model needs to access multiple applications and data sources.5Technologys generational moment with generative AI:A CIO and CTO guideArchetype Example use cases Estimated total cost of ownershipTaker Off-

81、the-shelf coding assistant for software developers General-purpose customer service chatbot with prompt engineering only and text chat only$0.5 million to$2.0 million,one-time Off-the-shelf coding assistant:$0.5 million for integration.Costs include a team of 6 working for 3 to 4 months.General-purp

82、ose customer service chatbot:$2.0 million for building plug-in layer on top of 3rd-party model API.Costs include a team of 8 working for 9 months.$0.5 million,recurring annually Model inference:Off-the-shelf coding assistant:$0.2 million annually per 1,000 daily users General-purpose customer servic

83、e chatbot:$0.2 million annually,assuming 1,000 customer chats per day and 10,000 tokens per chat Plug-in-layer maintenance:up to$0.2 million annually,assuming 10%of development cost.Shaper Customer service chatbot fine-tuned with sector-specific knowledge and chat history$2.0 million to$10.0 million

84、,one-time unless model is fine-tuned further Data and model pipeline building:$0.5 million.Costs include 5 to 6 machine learning engineers and data engineers working for 16 to 20 weeks to collect and label data and perform data ETL.Model fine-tuning:$0.1 million to$6.0 million per training run Lower

85、 end:costs include compute and 2 data scientists working for 2 months Upper end:compute based on public closed-source model fine-tuning cost Plug-in-layer building:$1.0 million to$3.0 million.Costs include a team of 6 to 8 working for 6 to 12 months.0.5 million to$1.0 million,recurring annually Mode

86、l inference:up to$0.5 million recurring annually.Assume 1,000 chats daily with both audio and texts.Model maintenance:$0.5 million.Assume$100,000 to$250,000 annually for MLOps platform and 1 machine learning engineer spending 50%to 100%of their time monitoring model performance.Plug-in-layer mainten

87、ance:up to$0.3 million recurring annually,assuming 10%of development cost.Maker Foundation model trained for assisting in patient diagnosis$5.0 million to$200.0 million,one-time unless model is fine-tuned or retrained Model development:$0.5 million.Costs include 4 data scientists spending 3 to 4 mon

88、ths on model design,development,and evaluation leveraging existing research.Data and model pipeline:$0.5 million to$1.0 million.Costs include 6 to 8 machine learning engineers and data engineers working for 12 weeks to collect data and perform data ETL.Model training:$4.0 million to$200.0 million pe

89、r training run.Costs include compute and labor cost of 4 to 6 data scientists working for 3 to 6 months.Plug-in-layer building:$1.0 million to$3.0 million.Costs include a team of 6 to 8 working 6 to 12 months.$1.0 million to$5.0 million,recurring annually Model inference:$0.1 million to$1.0 million

90、annually per 1,000 users.Assume each physician sees 20 to 25 patients per day and patient speaks for 6 to 25 minutes per visit.Model maintenance:$1.0 million to$4.0 million recurring annually.Assume$250,000 annually for MLOps platform and 3 to 5 machine learning engineers to monitor model performanc

91、e.Plug-in-layer maintenance:up to$0.3 million recurring annually,assuming 10%of development cost.Exhibit 1 Each archetype has its own costs.Note:Through engineering optimizations,the economics of generative AI are evolving rapidly,and these are high-level estimates based on total cost of ownership(r

92、esources,model training,etc)as of mid-2023.1 Extract,transform,and load.Model is fine-tuned on data set consisting of 100,000 pages of sector-specific documents and 5 years of chat history from 1,000 customer representatives,which is 48 billion tokens.Lower end cost consists of 1%parameters retraine

93、d on open-source models(eg,LLaMA)and upper end on closed-source models.Chatbot can be accessed via both text and audio.Model is optimized after each training run based on use of hyperparameters,data set,and model architecture.Model may be refreshed periodically when needed(eg,with fresh data).Gilad

94、Shaham,“Build or buy your MLOps platform:Main considerations,”LinkedIn,November 3,2021.5 Model is trained on 65 billion to 1 trillion parameters and data set of 1.2 to 2.4 trillion tokens.The tool can be accessed via both text and audio.6Technologys generational moment with generative AI:A CIO and C

95、TO guideExhibit 2 Generative AI is integrated at key touchpoints to enable a tailoredcustomer journey.Agent picks up case andprovides new solutionChatbot pings cus-tomersupportChatbot re-spondsAgent inputs new solu-tion forreview/feedback to modelIllustrative customer journey using travel agent botM

96、odel receives user request and pulls user info in promptLog-in authentif-cation,model/cus-tomer info access authorizationModel checks booking policy and sees cus-tomer can-not make changeDisagreesSelectsoptionModel explains issue and givesalternate optionsModel instructs booking system to complete t

97、askCustomer requests live agentWorkfowmanagementfor bookingBookingmodifcation policymanagementModel instructs customer support system to assign agentCustomer completes book-ing change and drops ofAPI callsWorkfowmanagement for live agentassignmentcustomer journey.Back-end appsGenera-tive AI modelCus

98、tomer logs in and requests to change bookingCustomer reviews optionsCus-tomerCustomer ID dataCustomer history dataPolicy dataBooking system dataAgent assignment dataDatasourceCloud/on-premises infrastructure and computeInfra-structure andcomputeChatbot activatedInter-actionChatbot communi-cates mess

99、age andoptions7Technologys generational moment with generative AI:A CIO and CTO guideFor the Taker archetype,this level of coordination isnt necessary.But for companies looking to scale the advantages of generative AI as Shapers or Makers,CIOs and CTOs need to upgrade their technology architecture.T

100、he prime goal is to integrate generative AI models into internal systems and enterprise applications and to build pipelines to various data sources.Ultimately,its the maturity of the businesss enterprise technology architecture that allows it to integrate and scale its generative AI capabilities.Rec

101、ent advances in integration and orchestration frameworks,such as LangChain and LlamaIndex,have significantly reduced the effort required to connect different generative AI models with other applications and data sources.Several integration patterns are also emerging,including those that enable model

102、s to call APIs when responding to a user queryGPT-4,for example,can invoke functionsand provide contextual data from an external data set as part of a user query,a technique known as retrieval augmented generation.Tech leaders will need to define reference architectures and standard integration patt

103、erns for their organization(such as standard API formats and parameters that identify the user and the model invoking the API).There are five key elements that need to be incorporated into the technology architecture to integrate generative AI effectively(Exhibit 3):Context management and caching to

104、 provide models with relevant information from enterprise data sources.Access to relevant data at the right time is what allows the model to understand the context and produce compelling outputs.Caching stores results to frequently asked questions to enable faster and cheaper responses.Policy manage

105、ment to ensure appropriate access to enterprise data assets.This control ensures that HRs generative AI models that include employee compensation details,for example,cannot be accessed by the rest of the organization.Model hub,which contains trained and approved models that can be provisioned on dem

106、and and acts as a repository for model checkpoints,weights,and parameters.Prompt library,which contains optimized instructions for the generative AI models,including prompt versioning as models are updated.MLOps platform,including upgraded MLOps capabilities,to account for the complexity of generati

107、ve AI models.MLOps pipelines,for example,will need to include instrumentation to measure task-specific performance,such as measuring a models ability to retrieve the right knowledge.In evolving the architecture,CIOs and CTOs will need to navigate a rapidly growing ecosystem of generative AI provider

108、s and tooling.Cloud providers provide extensive access to at-scale hardware and foundation models,as well as a proliferating set of services.MLOps and model hub providers,meanwhile,offer the tools,technologies,and practices to adapt a foundation model and deploy it into production,while other compan

109、ies provide applications directly accessed by users built on top of foundation models to perform specific tasks.CIOs and CTOs will need to assess how these various capabilities are assembled and integrated to deploy and operate generative AI models.6.Develop a data architecture to enable access to q

110、uality dataThe ability of a business to generate and scale value,including cost reductions and improved data and knowledge protections,from generative AI models will depend on how well it takes advantage of its own data.Creating that advantage relies on a data architecture that connects generative A

111、I models to internal data sources,which provide context or help fine-tune the models to create more relevant outputs.8Technologys generational moment with generative AI:A CIO and CTO guideExhibit 3 The tech stack for generative AI is emerging.Illustrative generative AI tech stack1Software as a servi

112、ce.2Direct to consumer.3Enterprise resource planning.4Customer relationship management.McKinsey&CompanyUsersToolingDataModelsAppsInfrastructureApps-as-a-service with embedded foundation modelsEnd-user-facingapplications and founda-tion models accessed through a browserinterface as SaaS(eg,Midjourney

113、)PolicymanagementRole-based accesscontrol and content-based policies to secure enterprise data assets Data sourcesExperience layerAPI gatewayEmbeddings,unstructured data,analytical data,trans-actional dataDataplatformsVectordatabases,datawarehouse,data lakeCloud or on-premises infrastructureand comp

114、ute hardwareQA andobservabilityQA modeloutputs(eg,checks for bias)Closed-source foundation modelsAPI-based,pre-trained models(eg,GPT-4)Model hubPlatforms that allow users to share models and data sets(eg,Hugging Face)Open-/closed-sourcefoundation modelsMLOps platformTrained model that is madeaccessi

115、ble(eg,BLOOM)Prompt libraryDTC or B2B applications(eg,Jasper)Context management and cachingUser and task context retrieved from enterprise data sources to prompt generative AI models,cache for common requestsExisting enterprise platforms(eg,ERP,CRM)9Technologys generational moment with generative AI

116、:A CIO and CTO guideIn this context,CIOs,CTOs,and chief data officers need to work closely together to do the following:Categorize and organize data so it can be used by generative AI models.Tech leaders will need to develop a comprehensive data architecture that encompasses both structured and unst

117、ructured data sources.This requires putting in place standards and guidelines to optimize data for generative AI usefor example,by augmenting training data with synthetic samples to improve diversity and size;converting media types into standardized data formats;adding metadata to improve traceabili

118、ty and data quality;and updating data.Ensure existing infrastructure or cloud services can support the storage and handling of the vast volumes of data needed for generative AI applications.Prioritize the development of data pipelines to connect generative AI models to relevant data sources that pro

119、vide“contextual understanding.”Emerging approaches include the use of vector databases to store and retrieve embeddings(specially formatted knowledge)as input for generative AI models as well as in-context learning approaches,such as“few shot prompting,”where models are provided with examples of goo

120、d answers.7.Create a centralized,cross-functional generative AI platform teamMost tech organizations are on a journey to a product and platform operating model.CIOs and CTOs need to integrate generative AI capabilities into this operating model to build on the existing infrastructure and help to rap

121、idly scale adoption of generative AI.The first step is setting up a generative AI platform team whose core focus is developing and maintaining a platform service where approved generative AI models can be provisioned on demand for use by product and application teams.The platform team also defines p

122、rotocols for how generative AI models integrate with internal systems,enterprise applications,and tools,and also develops and implements standardized approaches to manage risk,such as responsible AI frameworks.CIOs and CTOs need to ensure that the platform team is staffed with people who have the ri

123、ght skills.This team requires a senior technical leader who acts as the general manager.Key roles include software engineers to integrate generative AI models into existing systems,applications,and tools;data engineers to build pipelines that connect models to various systems of record and data sour

124、ces;data scientists to select models and engineer prompts;MLOps engineers to manage deployment and monitoring of multiple models and model versions;ML engineers to fine-tune models with new data sources;and risk experts to manage security issues such as data leakage,access controls,output accuracy,a

125、nd bias.The exact composition of the platform team will depend on the use cases being served across the enterprise.In some instances,such as creating a customer-facing chatbot,strong product management and user experience(UX)resources will be required.Realistically,the platform team will need to wor

126、k initially on a narrow set of priority use cases,gradually expanding the scope of their work as they build reusable capabilities and learn what works best.Technology leaders should work closely with business leads to evaluate which business cases to fund and support.8.Tailor upskilling programs by

127、roles and proficiency levelsGenerative AI has the potential to massively lift employees productivity and augment their capabilities.But the benefits are unevenly distributed depending on roles and skill levels,requiring leaders to rethink how to build the actual skills people need.5 “Unleashing deve

128、loper productivity with generative AI,”June 27,2023.10Technologys generational moment with generative AI:A CIO and CTO guideOur latest empirical research using the generative AI tool GitHub Copilot,for example,helped software engineers write code 35 to 45 percent faster.The benefits,however,varied.H

129、ighly skilled developers saw gains of up to 50 to 80percent,while junior developers experienced a 7 to 10percent decline in speed.Thats because the output of the generative AI tools requires engineers to critique,validate,and improve the code,which inexperienced software engineers struggle to do.Con

130、versely,in less technical roles,such as customer service,generative AI helps low-skill workers significantly,with productivity increasing by 14 percent and staff turnover dropping as well,according to one study.These disparities underscore the need for technology leaders,working with the chief human

131、 resources officer(CHRO),to rethink their talent management strategy to build the workforce of the future.Hiring a core set of top generative AI talent will be important,and,given the increasing scarcity and strategic importance of that talent,tech leaders should put in place retention mechanisms,su

132、ch as competitive salaries and opportunities to be involved in important strategic work for the business.Tech leaders,however,cannot stop at hiring.Because nearly every existing role will be affected by generative AI,a crucial focus should be on upskilling people based on a clear view of what skills

133、 are needed by role,proficiency level,and business goals.Lets look at software developers as an example.Training for novices needs to emphasize accelerating their path to become top code reviewers in addition to code generators.Similar to the difference between writing and editing,code review requir

134、es a different skill set.Software engineers will need to understand what good code looks like;review the code created by generative AI for functionality,complexity,quality,and readability;and scan for vulnerabilities while ensuring they do not themselves introduce quality or security issues in the c

135、ode.Furthermore,software developers will need to learn to think differently when it comes to coding,by better understanding user intent so they can create prompts and define contextual data that help generative AI tools provide better answers.Beyond training up tech talent,the CIO and CTO can play a

136、n important role in building generative AI skills among nontech talent as well.Besides understanding how to use generative AI tools for such basic tasks as email generation and task management,people across the business will need to become comfortable using an array of capabilities to improve perfor

137、mance and outputs.The CIO and CTO can help adapt academy models to provide this training and corresponding certifications.The decreasing value of inexperienced engineers should accelerate the move away from a classic talent pyramid,where the greatest number of people are at a junior level,to a struc

138、ture more like a diamond,where the bulk of the technical workforce is made up of experienced people.Practically speaking,that will mean building the skills of junior employees as quickly as possible while reducing roles dedicated to low-complexity manual tasks(such as writing unit tests).9.Evaluate

139、the new risk landscape and establish ongoing mitigation practices Generative AI presents a fresh set of ethical questions and risks,including“hallucinations,”whereby the generative AI model presents an incorrect response based on the highest-probability response;the accidental release of confidentia

140、l personally identifiable information;inherent bias in the large data sets the models use;and high degrees of uncertainty related to intellectual property(IP).CIOs and CTOs will need to become fluent in ethics,humanitarian,and compliance issues to adhere not just to the letter of the law(which will

141、vary by country)but also to the spirit of responsibly managing their businesss reputation.6 Erik Brynjolfsson,Danielle Li,and Lindsey R.Raymond,Generative AI at work,National Bureau of Economic Research(NBER)working paper,number 31161,April 2023.11Technologys generational moment with generative AI:A

142、 CIO and CTO guideAddressing this new landscape requires a significant review of cyber practices and updating the software development process to evaluate risk and identify mitigation actions before model development begins,which will both reduce issues and ensure the process doesnt slow down.Proven

143、 risk-mitigation actions for hallucinations can include adjusting the level of creativity(known as the“temperature”)of a model when it generates responses;augmenting the model with relevant internal data to provide more context;using libraries that impose guardrails on what can be generated;using“mo

144、deration”models to check outputs;and adding clear disclaimers.Early generative AI use cases should focus on areas where the cost of error is low,to allow the organization to work through inevitable setbacks and incorporate learnings.To protect data privacy,it will be critical to establish and enforc

145、e sensitive data tagging protocols,set up data access controls in different domains(such as HR compensation data),add extra protection when data is used externally,and include privacy safeguards.For example,to mitigate access control risk,some organizations have set up a policy-management layer that

146、 restricts access by role once a prompt is given to the model.To mitigate risk to intellectual property,CIOs and CTOs should insist that providers of foundation models maintain transparency regarding the IP(data sources,licensing,and ownership rights)of the data sets used.Generative AI is poised to

147、be one of the fastest-growing technology categories weve ever seen.Tech leaders cannot afford unnecessary delays in defining and shaping a generative AI strategy.While the space will continue to evolve rapidly,these nine actions can help CIOs and CTOs responsibly and effectively harness the power of

148、 generative AI at scale.Copyright 2023 McKinsey&Company.All rights reserved.Aamer Baig is a senior partner in McKinseys Chicago office;Sven Blumberg is a senior partner in the Dsseldorf office;Eva Li is a consultant in the Bay Area office,where Megha Sinha is a partner;Douglas Merrill is a partner i

149、n the Southern California office;Adi Pradhan and Stephen Xu are associate partners in the Toronto office;and Alexander Sukharevsky is a senior partner in the London office.The authors wish to thank Stephanie Brauckmann,Anusha Dhasarathy,Martin Harrysson,Klemens Hjartar,Alharith Hussin,Naufal Khan,Sa

150、m Nie,Chandrasekhar Panda,Henning Soller,Nikhil Srinidhi,Asin Tavakoli,Niels Van der Wildt,and Anna Wiesinger for their contributions to this article.12Technologys generational moment with generative AI:A CIO and CTO guideRewired to outcompeteSix signature moves led by the C-suite can build organiza

151、tions that will outperform in the age of digital and AI.by Eric Lamarre,Kate Smaje,and Rodney ZemmelJune 2023How companies navigate the technology world to achieve sustainable competitive advantage is the defining business challenge of our time.To be fair,this challenge isnt new.But its an increasin

152、gly pressing one,with deep implications for how companies navigate a world where digital and AI are fundamentally reshaping how we work and live.Companies understand they need to meet the challenge,but most of them are struggling.McKinsey research shows that while 90 percent of companies have launch

153、ed some flavor of digital transformation,only a third of the expected revenue benefits,on average,have been realized.1Yet its also a challenge with enormous potential for the companies that get it right.In the banking sector,for example,where digital and AI transformations have been under way for th

154、e past decade,compelling empirical data shows that digitally transformed banks outperform their peers.We leveraged a unique data set,Finalta by McKinsey,to analyze 20 digital leaders and 20 digital laggards in retail banking between 2018 and 2022.The results were startling.Digital leaders improved t

155、heir return on tangible equity,their P/E ratio,and their total shareholder returns materially more than digital laggards(Exhibit1).Digital excellence is translating into financial outperformance.This outperformance was propelled by a deeper integration of technology across end-to-end core business p

156、rocesses.This,in turn,drove higher digital sales and lower costs in branches Web 2023McKQ-RewiredToOutcompeteExhibit 1 of 220222018202220184.98.18.410.07.29.813.615.515.319.3Retail-banking exampleDigital leaders are spurring more value for their shareholders.McKinsey&CompanyTSRP/E ratioReturn on tan

157、gible equity(ROTE)Weighted peer group average of TSR,201822,CAGR,%Weighted peer group average of P/E ratio,next 12 monthsAdjusted pretax ROTE,peer group average,%Laggards2Leaders11Top 20 retail banks between 2018 and 2022.2Bottom 20 retail banks between 2018 and 2022.Source:S&P Global;Corporate Perf

158、ormance Analytics by McKinsey1 “Three new mandates for capturing a digital transformations full value,”McKinsey,June 15,2022.Exhibit 114Rewired to outcompeteand operations.How did the digital leaders accomplish this?By bringing business,technology,and operations more closely together to digitally in

159、novate;by upskilling their organizations;and by building a distributed technology and data environment to empower hundreds if not thousands of teams to digitally innovate,day in,day out.This gets at the nub of why digital and AI transformations are so difficultcompanies need to get a lot of things r

160、ight.Clearly,for digital and AI to deliver on their business transformation potential,the top team needs to be ready and willing to undertake the organizational“surgery”required to become a digitally capable enterprise.There are no quick fixes.You cant simply implement a system or a technology and b

161、e done.Instead,success means having hundreds of technology-driven solutions(proprietary and off the shelf)working together that you continually improve to create great customer and employee experiences,lower unit costs,and generate value.But creating,managing,and evolving these solutions at enterpri

162、se scale requires a fundamental rewiring of how a company operates.That means getting thousands of people across different units of the organization working together and working differently to digitally innovate,constantly.The lessons learned from our work with more than 200 large companies across m

163、ultiple industries show that capturing this kind of value from digital and AI requires building six critical enterprise capabilities(Exhibit 2).These allow rewired companies to integrate new technologies,such as generative AI,and harness them to create value.While companies may understand this at a

164、high level,they struggle with how to build these capabilities successfully and ensure that they work together across the enterprise.Our new book,Rewired:The McKinsey Guide to Outcompeting in the Age of Digital and AI,is all Exhibit 2Web 2023McKQ-RewiredToOutcompeteExhibit 2 of 2Transformational valu

165、e comes from careful and coordinated execution across all areas of focusSix enterprise capabilities are critical for successful digital and AI transformations.McKinsey&CompanyAlignmenton valueDeliverycapabilitiesChangemanagement1.Business-led digital road mapAlign senior leadership team on the visio

166、n,value,and road map for the transformation;reimagine business domains to deliver outstanding customer experiences and to lower unit costs.Maximize value capture by ensuring the adoption and enterprise scaling ofdigital solutions and by tightly managing the transformation progress and risks.2.Talent

167、3.Operating model4.Technology5.Data6.Adoption and scalingEnsure that you have the right skills and capabilities toinnovate and execute.Increase themetabolic rate ofthe organization by bringing business,operations,andtechnology together.Make technology easier for teams to use so they can innovate at

168、pace.Continually enrich data and make it easily accessible across the organization to help improve customer experience andbusiness performance.15Rewired to outcompeteabout the how.This article is adapted from that book and delineates the core aspects of what it takes for leaders to spur transformati

169、on across all six capabilities.Before we go into detail,its worth highlighting two key findings.First,no digital and AI transformation can be successful without building a baseline of competence across all six capabilities.Second,these elements are interconnected and need to be managed that way:a go

170、od operating model,for example,cant work without the right talent.Similarly,great technology wont make much of an impact if users dont adopt it.You do not have to be a tech company to achieve excellence in digital and AI.Large,established companies can outcompete and capture value,but only when they

171、 are willing to commit to the hard work of rewiring their enterprise.This is a job for the entire C-suite,not just the CEO or the chief information officer(CIO).The cross-functional nature of a digital and AI transformation requires an unparalleled level of collaboration across the C-suite,with ever

172、yone having an important part to play in building these enterprise capabilities.Rewiring the business is an ongoing journey of improvement,not a destination.Lets dig into the details of that journey.Align the C-suite around a business-led road mapWhen evaluating stalled digital and AI transformation

173、s,we find that many of the issues that impede a programs success can be traced back to insufficient planning and alignment.Misunderstanding among leadership at the strategic-planning stage will invariably lead to muddled execution in a companys transformation.Because digital and AI transformations a

174、ffect so many parts of the business,investing the necessary time to help make the transformation a success pays significant dividends in terms of clarity and unified action.The best companies make sure to get these three early moves right:Inspire and align the top team.Take the time to establish a c

175、ommon digital language,learn from other companies that are further along the journey,develop a shared vision among the C-suite,and explicitly agree on a set of commitments that match your ambitions.Consider the example of DBS Bank,one of the worlds most successful digitally transformed banks.CEO Piy

176、ush Gupta and his top leaders visited and learned from top tech companies around the globe and used those lessons to shape a vision around“Making Banking Joyful”and to commit to making DBS a tech leader.This kind of leadership alignment is crucial to ensuring a successful digital and AI transformati

177、on.Get the“bite”size right:business domains.Some companies struggle from the start of their digital and AI transformation by getting the scope of the change wrong.They start too smallbelieving that implementing a few use cases will lower riskor they spread bets and resources too thinly across an unc

178、oordinated set of initiatives.Both approaches typically produce little value.Successful companies,on the other hand,focus their efforts on a few important business domains,such as a production process or the customer journey,and transform them from end to end.As many as 80 percent of successful inte

179、rventions in struggling digital and AI transformations are based on reanchoring the scope to spur a concerted effort against a few well-defined domains.Commit to a contract with the C-suite.Effective rewiring requires companies to tie the transformation outcomes of each business domain to specific i

180、mprovements in operational KPIs,such as reduction in customer churn or improvements in process yield.The team builds a road map where the digital solutions that underpin these KPI improvements are sequenced in a way to produce meaningful value in the short term(say,12 to 18 months)and transformation

181、al value in the medium term(three to five years,for example).The plan explicitly accounts for the buildout of enterprise capabilities,such as hiring digital talent or modernizing data architecture.C-suite leaders commit to these KPI improvements,and the expected benefits are baked into their busines

182、s objectives.Our rule of thumb is that a robust digital road map should deliver EBIT improvement of 20 percent or more.16Rewired to outcompeteWhen business leaders define an ambitious yet realistic transformation of their business domains with technology,they set in motion the flywheel of digital ch

183、ange.The resulting digital road map is their signature move and effectively acts as a contract that they commit to implementing.Build your talent benchNo company can outsource its way to digital excellence.Being digital means having your own bench of digital talentproduct owners,experience designers

184、,cloud engineers,software developers,and so onworking side by side with your business colleagues.Digital transformations are,first and foremost,people transformations.Here are three actions that digital leaders take:Create a cleansheet for your talent.Most companies have digital technologists,but ma

185、ny still face the hard work of reskilling their technology and IT organization.The aspiration should be to have 70 to 80 percent of your digital talent in-house,with 20 to 30 percent coming from outside the company and focused on specialized skills,flexibility,or both.Your talent pyramid should shif

186、t to a diamond shape,with more competent technologists and fewer novices.Thats because there is a step change in productivity from more experienced technologists.You should also have a healthy ratio of hands-on-keyboard technologists versus managerial roles.Rewired leaders target a 4:1 ratio(or bett

187、er)of engineers to managers,versus the 1:1 found at many companies.Get religion about skills.Rewired companies develop very granular skill progression grids supported by credentials.For example,Big Tech companies have up to ten levels of data engineers,each with different skill levels and compensati

188、on ranges.Without a precise calibration of skills,it becomes difficult to recognize distinctive technologists and compensate them accordingly.Skill progression also gets built into expert-based career tracks and in learning and development programs.In short,the whole digital-talent model revolves ar

189、ound fostering excellence in people devoted to their craft.Build the team that will build your digital bench.Many HR organizations are hampered by slow recruiting and onboarding processes,rigid compensation frameworks,and outdated learning and development programs for digital talent.But transforming

190、 your entire HR organization and underlying HR processes to make them digital ready may not be practical.Setting up a special team focused on adapting current HR processes to win digital talent is the most pragmaticand successfulway forward.We call this designated team the Talent Win Room(TWR).The p

191、rimary mission of a TWR is to find technologists with the right skills and to build and continually improve all facets of both the candidate and employee experience.These shifts in talent practices are not simple,but they are fundamental to becoming rewired with the right talent.While every C-suite

192、executive will have a part to play in this talent reinvention,this is often the chief human resources officers signature contribution to the enterprises digital transformation.Adopt a new operating model that can scaleMost companies have succeeded in standing up a handful of cross-functional agile t

193、eams.But scaling up so that hundreds or even thousands of teams work that way,as rewired businesses do,is a daunting challenge.Developing the right operating model to bring business,technology,and operations closer together is perhaps the most complex aspect of a digital and AI transformation becaus

194、e it touches the core of the organization and how people work.Three leading models have emerged:digital factory,product and platform,and enterprise-wide agile.Each of these models is built on two core ideas.The first is that small,multidisciplinary agile teams,or pods,are the most effective and effi

195、cient way to develop software.Second,pods work together most effectively when some are focused on directly improving a customer or user experience(generally called product pods,although they can also be called experience or journey pods)17Rewired to outcompetewhile others focus on creating reusable

196、services to accelerate the work of all pods(called platform pods).Examples of such services could include a customer-360 data set or an easy way for teams to provision compute and storage capacity.The implementation of a new operating model is,in our opinion,one of the most significant pivots a comp

197、any can make to become a rewired enterprise.There are two key moves to getting this right:Select an operating model that supports your strategy.The digital factory is a separate organizational unit where people work together to build digital solutions for the business units or functions that fund th

198、e digital factory.Companies often initially select the digital-factory model because it is a self-contained operating unit and can be implemented relatively quickly(typically 12 to 18 months before its fully operational,though it can get started in a matter of weeks).BHP and Scotiabank,for example,h

199、ave implemented this model.The product and platform model is a more evolved version of the digital factory.While the digital factory might contain 20 to 50 pods,the product and platform model will typically have a few hundred pods,sometimes thousands for large companies.When companies move to a prod

200、uct and platform model,they are making a major strategic decision to realign large parts of the organization to better exploit technology in their core business.Amazon,Google,Ita Unibanco,and JPMorgan Chase have all implemented this model.Finally,the enterprise-wide agile model builds on the product

201、 and platform model and extends the benefit of agile to the entire business,not just the technology-intensive areas.For example,key account sales and R&D can also benefit from working in small,cross-functional teams.Companies adopt this model when they believe that customer centricity,collaboration,

202、and flexible resource deployment are key performance differentiators across the entire enterprise.ING and Spark New Zealand have successfully implemented this model.Professionalize product management.A crucial difference between tech companies and their peers in other sectors is the degree to which

203、they have embedded product management capabilities in their operating models.This capability,in our opinion,makes or breaks the implementation of a new operating model.Some 75 percent of business leaders in a McKinsey survey responded that product management best practices arent being adopted at the

204、ir companies,that product management is a nascent function within their organizations,or that it doesnt exist at all.2 Thats a problem.Its also hard to recruit great product managers because understanding the industry and the company context matters.Most companies end up reskilling and building new

205、career tracks for this rare talent,but this requires substantial investments to ensure good results.The shift to a new operating model is the signature move of CEOs in rewiring the company.Only they can catalyze such large-scale organizational change.Technology for speed and distributed innovationTh

206、e main purpose of technology within a rewired company is to make it easy for hundreds,if not thousands,of pods to constantly develop and release digital innovations.This requires a distributed technology environment where every pod can access the software development tools,data,and applications they

207、 need.While leaders hoping to create that environment have a raft of decisions to make,three priorities stand out:Kit out a technology toolbox.Just like woodworkers,surgeons,or plumbers,software 2 Chandra Gnanasambandam,Martin Harrysson,Jeremy Schneider,and Rikki Singh,“What separates top product ma

208、nagers from the rest of the pack,”McKinsey,January 20,2023.18Rewired to outcompetedevelopers need the proper tools to do their work.As an organization scales from five agile pods to 100,or even more than 1,000,it doesnt make sense for pod members to be calling IT every time they have a basic request

209、,such as additional storage capacity or access to a collaboration tool.Leading companies build a developer platform:a self-service portal that makes it easy to access and use all the standardized and company-approved tools.Use APIs without exception.Once developers have their tools,they need access

210、to data and existing app functionalities to build their solutions.Application programming interfaces(APIs)do that by systematically minimizing dependencies in the architecture by making application functionalities and data easily accessible.Without it,pods will constantly find themselves depending o

211、n other pods.Amazons Jeff Bezos was so adamant about using APIs that he wrote a famous memo about it,which fundamentally changed Amazon and the world of software.The memo essentially said that all teams were expected to expose their data and functionality through service interfaces(that is,APIs)and

212、to communicate with one another through only these interfaces.No other form of inter-process communication would be allowed.No exceptions.Automate software delivery.Have you ever wondered how an app on your phone can be upgraded so frequently?That seamless functionality is made possible by software

213、delivery automation,also known as CI/CD:continuous integration and continuous delivery.This is the method for systematically automating all steps,including quality checks,testing,packaging(that is,containerization),and staged deployment of the solution to the user.With CI/CD,updates that used to tak

214、e weeks or months can now be completed in minutes,allowing pods to release incremental improvements weekly or even daily and thus unleash much faster innovation cycles.You wont be able to achieve distributed digital and AI innovation if pods arent able to release code to a production environment qui

215、ckly and easily.This fixation on automation needs to carry over to AI and machine-learning(ML)models.These models are like living organismsthey need to be constantly recalibrated as new data accumulate and then monitored in real time for drift and biases.When this doesnt happen,AI/ML models fail to

216、transition to full-scale production.Solving for this has required a specialized type of automation called machine learning operations(MLOps).For example,Vistra,a leading energy company,built MLOps automation to support more than 400 AI/ML models deployed to optimize different parts of its power plan

217、t operations.Most CIOs have started their companies journey to build a robust developer platform,decouple the components of the architecture from one another through APIs,and automate their software delivery pipeline.But we know very few companies that have scaled this across their enterprise.The ch

218、ange management efforts are significant,and the software engineering talent required is in short supply.Creating a technology environment that enables distributed digital and AI innovations is a cornerstone capability of rewired enterprises and a signature contribution by the CIO,the chief data offi

219、cer(CDO),or both.Embed data everywhereIn established companies,data is often a source of frustration.As much as 70 percent of the effort involved in developing AI-based solutions can be attributed to wrangling and harmonizing data.Unless data is thoughtfully sorted and organized for easy consumption

220、 and reuse,scaling solutions can be a big challenge.The ability to constantly improve customer experience and drive down unit cost depends on giving each digital and AI team(near)real-time access to data.Companies can focus on three areas to achieve this:Turn to reusable building blocks:data product

221、s.Data products are the secret sauce for scaling AI.They help deliver data-intensive applications as much as 90 percent faster,at 30 percent lower cost,and with a reduced risk and data governance burden.A data product delivers a high-quality,ready-to-use set of data in a way that people and applicat

222、ions across the organization can easily access and consume.For example,a data product 19Rewired to outcompetecould provide a 360-degree view of an important entity,such as customers,employees,product lines,or stores.Companies can prioritize building data products that have the broadest application,t

223、hat are critical for teams developing priority solutions,and that are unique.Building data products requires dedicated teams and investments.Install the data architecture“plumbing.”Data architecture is the system of“pipes”that deliver data from where it is stored to where it is used.When implemented

224、 well,data architecture hastens a companys ability to build reusable and high-quality data products and to put data within reach of any team in the organization.We have seen very rapid technological progress in this field.The emergence of new architectural patterns such as the“data lakehouse”(an inn

225、ovation that combines the capabilities of a data lake and a data warehouse into a single,integrated platform)makes it easier for companies to solve for both their business intelligence and their AI needs.Federate data governance.Data touches all aspects of an organization,so its governance needs to

226、account for that complexity.Rewired companies deploy a federated model where a central function(that is,a data management office)sets policies and standards and provides support and oversight,while business units and functions manage activities such as developing data products and building data pipe

227、lines to enable consumption.A data environment that allows for easy data consumption by hundreds of distributed teams is another signature move of the CIO in collaboration with the CDO.It enables data-driven decisions,feeds real-time decision-making systems,and propels faster continuous-improvement

228、loops.Unlocking adoption and scalingDeveloping a good digital solution can be complex and difficult.But getting customers or business users to adopt that solution as part of their day-to-day activities and then scaling that solution across the enterprise are often the biggest challenges.Successful c

229、ompanies concentrate on the following three moves:Focus equally on adoption and development.User adoption starts with developing great technology solutions that offer an excellent customer experience.But companies often underestimate all the additional elements of the business model that need to be

230、changed to secure adoption.For instance,an insurance company that developed analytic solutions to help agents upsell customers on policies also needed to make changes to pricing algorithms,sales force incentives,distribution and customer engagement models,and metrics and performance indicators.That

231、end-to-end system approach,with a focus on the people side of the equation,is what differentiates digital leaders.They achieve this by making the business accountable for the end-to-end transformation of the domain.As a rule,for every$1 spent on developing digital and AI solutions,plan to spend at l

232、east another$1 to ensure full user adoption and scaling across the enterprise.Scale with“assetizing.”Replicating the adoption of a solution in different environments,such as a network of plants,or in different geographic markets,customer segments,or organizational groups is challenging.Companies oft

233、en find themselves redoing a lot of work and struggling to tailor solutions to local environments.All this extra work is a scale killer,and thats why 72 percent of companies stall at this stage.Digital leaders solve this by“assetizing”solutions,which typically allows 60 to 90 percent of a digital an

234、d AI solution to be reused,leaving just 10 to 40 percent in need of local customization.Track what matters.No one will debate the need to measure the progress of a digital transformation.But the question is what to measure and how.Performance tracking that is poorly designed and lacking the right su

235、pporting tools can quickly crumble under its own weight.Rewired companies take the pods responsible for objectives and key results and link them to operational KPIs,tracking 20Rewired to outcompetethe progression of each pod in a disciplined stage gate review process.The ability to capture the full

236、economic potential of digital innovations is a core differentiator between digital leaders and laggards.Building this capability is the signature move of business unit and function leaders.The capabilities we have laid out for a successful digital and AI transformation present a rich“how to”agenda.Y

237、ou may be wondering where to start your rewiring journey.Why not start where we began this article:by bringing the top team together and having them reflect on your journey thus far?A digital and AI transformation is ultimately an exercise in constant evolution and improvement.If you accept this pre

238、mise,it will change your perspective on how you approach this critical challenge.To borrow Jeff Bezoss expression to Amazon shareholders about the importance of operating like a digital native:its always day one for digital and AI transformation.Copyright 2023 McKinsey&Company.All rights reserved.Er

239、ic Lamarre is a senior partner in McKinseys Boston office,Kate Smaje is a senior partner in the London office,and Rodney Zemmel is a senior partner in the New York office.21Rewired to outcompeteFull throttle on net zero:Creating value in the face of uncertaintyTo thrive amid shocks to the net-zero e

240、conomy,leaders are shifting strategies to position themselves to win when the skies clear up.by Laura Corb,Anna Granskog,Tomas Nauclr,and Daniel PacthodSeptember 2023No question,navigating the net-zero economy has become more complicated over the past 12 months amid higher energy prices,supply chain

241、 pressures,increased interest rates,higher input costs,and lackluster economic growth.Companies are experiencing long lead times,supply shortages,or price spikes for goods,from transformers to bio-based feedstocks,and services such as engineering,procurement,and construction,that could otherwise acc

242、elerate decarbonization.Many leaders feel that creating a clear picture of where the economy is headed has never been as difficult.For some,the current pressures are creating tension between near-term financial performance and commitments toward a net-zero world.However,our research and experience s

243、uggest that there are bold moves leaders can make to create value in the net-zero transition,despite the headwinds.Companies that take disciplined and courageous action on both resilience and sustainability have a unique opportunity:they can reposition themselves ahead of organizations that focus on

244、 just the short-term shocks,or organizations that might even step back from their sustainability commitments.We are seeing that some companies are steadfast in their conviction of pursuing green growth opportunities,while others are questioning whether now is the right time.Some leaders are pursuing

245、 a robust strategy for a range of future scenarios.1McKinsey research on the 200708 financial crisis shows that outperforming companies tended to take a few courses of action to create an earnings advantage,including proactively cutting costs and identifying areas of growth.For navigating the curren

246、t moment of uncertaintywith an eye toward net zerowe have developed a set of priorities that combine the tactics of outperforming companies in the 200708 crisis with moves made by early sustainability leaders.These actions can be applied widely across industries and geographies:push ahead on value c

247、reation with vision and ambition integrate cost and carbon reductions create customer partnerships to be an early winner in the market update the portfolio to secure profitable growth build and scale new green businesses execute at digital speed to create competitive distanceIn this article,we illus

248、trate how companies can still play offense in the net-zero transition despite uncertainty.The rewards for pushing ahead on green growth could be significant:our analysis shows that growing demand for net-zero offerings could generate$9 trillion to$12 trillion of annual sales by 2030.Push ahead on va

249、lue creation with vision and ambitionThe volatile economic environment in many regions makes it even more important for companies to orient their sustainability agendas around value creation in nascent or fast-growing markets.The advantage of being an early mover in these new markets is that compani

250、es can solidify pole position for offering low-carbon goods and build out production capacity before latecomers enter the market.But being early to segments with growth potential often requires vision and ambition.Consider a tier-one automotive supplier that set out to be a first-choice supplier for

251、 leading automotive OEMs looking to decarbonize.To do so,the supplier needed to offer a set of zero-carbon products at a competitive cost.Executing on this agenda has required the company to build leading capabilities in tracking and verifying the carbon content of the materials and components it pr

252、ocures,finding new suppliers,and utilizing carbon as a new element in product design.By investing in these areas,the company now has industry-leading capabilities in enabling Scope 3 emissions reductions.(Scope 3 1 For more,see“Leading through uncertainty in the energy and materials sectors,”McKinse

253、y,July 31,2023.23Full throttle on net zero:Creating value in the face of uncertaintyemissions are indirect emissions that arise across a companys value chain.)Integrate cost and carbon reductionsOur research shows that the companies that fared well coming out of the 200708 financial crisis systemati

254、cally invested in improving the cost competitiveness of their core offerings.In many cases,that means driving down the cost of goods sold(COGS).To date,lowering COGS has often been considered at odds with reducing a products carbon footprint.However,through our work with leading companies across sec

255、tors,we are now seeing that a trade-off between cost and carbon reductions is often not required.Companies in chemicals,pulp and paper,oil and gas,metals,and other process industries are going after the dual benefit of cost and carbon reductions by improving energy efficiency and process yields,as w

256、ell as by shifting to lower-carbon raw materials and feedstocks where possible.Manufacturing companies are addressing the same challenge by making changes in design,material specifications,and supply chain choices.Energy efficiency and yield improvements are not new strategies,but with higher energy

257、 prices,the value of investing in these areas has increased.In addition,sophisticated analytics tools have unlocked even more potential for dual savings.In one example,a leading paper and packaging player set out to reduce energy costs and direct emissions at its largest mill.The company deployed en

258、gineering solutions such as heat integration and steam optimization to reduce energy consumption,as well as advanced analytics to track it.The paper player found an opportunity to reduce its energy costs by 10 to 16 percent and reduce direct emissions by 12 percent.With many industrial players still

259、 facing relatively high energy prices,such energy efficiency opportunities are abundant.In Europe,for example,we estimate that energy-intense industries could create anywhere from 3billion to 12 billion in value by deploying energy efficiency measures such as advanced analytics.In a different case,a

260、 specialty chemicals player is looking to combine cost and carbon reductions in a systematic and highly aspirational sustainable raw-material program.The company has identified a pathway to reduce more than half of its emissions by 2050 while reducing up to hundreds of millions in costs annually.Suc

261、h approaches can come with added benefits down the line.Companies that build carbon reduction competencies across the organization or lock up scarce supply of low-carbon raw materials and components can gain a longer-term edge in the marketplace.Create customer partnerships to be an early winner in

262、the marketIn the current economic cycle,companies in competitive markets may feel a slowdown in their order books and tougher competition for deals.In turn,sales organizations work harder to fill the order pipelines,and pricing decisions become difficult.Meanwhile,for companies that have unique,zero

263、-carbon product offerings,there are opportunities to gain market share.One way to do this is by signing up partners through offtake agreementsthat is,agreements for customers to purchase all or a substantial part of output.Offtake agreements can help solidify an early and disproportionate share of d

264、emand in more nascent markets,and the income can be invested into scaling further capacity.In some cases,partnerships can also help companies earn a price premium.Such steps may require some market shaping to maximize impact.In our experience,calls for offtake commitments are often made at the very

265、top,through CEO-to-CEO dialogue.Offtake agreements can be a strategic advantage for the customer,too,as the customer can lock in supply of early-to-market goods and services ahead of the competition.Once partnerships with offtakers are born,partners can build business ecosystems for the value chain

266、that will allow the category to grow(bringing together raw-material suppliers,technology partners,or regulators,for example).Players that create top-level relationships with their potential customer and 24Full throttle on net zero:Creating value in the face of uncertaintypartner base ahead of others

267、 could have a head start on capturing value from their green offerings.As we have discussed in prior articles,companies that produce sustainable goods can also earn green price premiums through product differentiation.2 What we have learned recently is that green premiums can vary based on a product

268、s carbon credentialsthat is,a zero-carbon offering may earn a higher premium than a lower-carbon one.We have seen this play out in metals:lower-carbon aluminum has been on the market for some time,but the price delta compared with traditional aluminum has been negligible.Zero-carbon or green steel,o

269、n the other hand,has earned a clear price premium compared with any other type of steel.3Update the portfolio to secure profitable growthCompanies that are generating profits with legacy,higher-emissions businesses could face a conundrum:Should they hold on to the legacy business to help finance gre

270、ener investments or pull out of the legacy business proactively?Our analysis shows that companies that came out the strongest from the 200708 financial crisis were the ones that divested early and then acquired businesses ahead of others.4 With this in mind,our perspective is that now is the time fo

271、r companies to take stock of their portfolios with a focus on the long-term outlook of each business.If there is an opportunity to improve the overall growth of the portfolio by rotating out some businesses that are facing diminishing returns due to their carbon emissions,and adding businesses that

272、are propelled by sustainability tailwinds,there may be no reason to hold back.For instance,over the past two decades,NextEra Energy moved out of its thermal-generation portfolio and became a leader in renewable power(in 2020,the company closed its last coal plant in Florida).NextEra is also investin

273、g in clean fuels,hydrogen,and battery storageforms of on-demand,dispatchable generation that can support wind and solar power,which are nondispatchable.5 NextEras subsidiary,Florida Power&Light,plans to convert all of its remaining 16 gigawatts of thermal generation to clean fuels or hydrogen genera

274、tion while driving value to investors and leading the industry in returns and market cap.6The potential value of gearing portfolios toward low-carbon businesses can also be seen at the sector level.A McKinsey review of chemicals companies,for example,revealed that green leaderscompanies with both gr

275、eener product portfolios and exposure to end markets associated with sustainability,including electric vehicles and energy storagesee two to three times higher total shareholder returns compared with laggards.7Additionally,in light of higher interest rates,capital cost is becoming an increasingly im

276、portant factor.For example,research by the University of Oxford suggests that low-carbon electric utilities in Europe have a lower cost of capital than peers with higher-emission portfolios.8 As the net-zero transition continues,executives can look for opportunities in industries where capital costs

277、 are evolving.2“Playing offense to create value in the net-zero transition,”McKinsey,April 13,2022.3 There is not yet a universally set definition for green steel.In one example,the German Steel Association has proposed an approach where steel with emissions below 350 to 450 tons of CO2e per ton of

278、steel(depending on the share of scrap contents)would qualify as A-labeled green steel.4“Somethings coming:How US companies can build resilience,survive a downturn,and thrive in the next cycle,”McKinsey,September 16,2022.5 Wind and solar are considered to be nondispatchable because they rely on exter

279、nal variables(wind or sun).6“NextEra Energy sets industry-leading Real Zero goal to eliminate carbon emissions from its operations,leverage low-cost renewables to drive energy affordability for customers,”NextEra Energy news release,June 14,2022.7 Measuring the“greenness”of a chemical company(or any

280、 company)is not straightforward.To better understand how sustainability in chemicals is actively driving valuation,we segmented our sample of chemical companies along two dimensions:those with“greener”product portfoliosdefined as more than 25 percent of revenues in biologic,recyclable,or low-carbon

281、product portfoliosand those with exposure to end markets supporting sustainability tailwinds,such as electric vehicles,energy storage,water reduction,energy efficiency,natural ingredients,or circular packaging.For more,see“Chemicals and capital markets:Growing sustainably,”McKinsey,April 22,2022.8 X

282、iaoyan Zhou et al.,Energy transition and the changing cost of capital:2023 review,Oxford Sustainable Finance Group and the University of Oxford,March 2023.25Full throttle on net zero:Creating value in the face of uncertaintyBuild and scale new green businessesOur research suggests that companies tha

283、t built new businesses in the last economic downturn outperformed peers by 10 percent during the crisis and 30 percent through the cycle.9 These companies took out their magnifying glasses,identified pockets of growth,and positioned themselves to take advantage.They anticipated market needs and allo

284、cated money to accelerate innovation.It should not come as a surprise that the net-zero transition is creating several pockets of growth.The growing demand for low-emission products,in part propelled by corporate emission reduction commitments,10 is creating opportunities for commercial scaling of a

285、 wide range of climate technologies and related services.This demand is further supported by major regulatory initiatives.Last years Inflation Reduction Act in the United States,for example,allocates about$370 billion for climate and energy spending.Multiple policy packages under the umbrella of the

286、 European Green Deal,including the recent Green Deal Industrial Plan,promise to further accelerate the regions shift toward a net-zero economy by facilitating faster access to funding.Similar to previous regulatory programssuch as early offshore wind auctions in the United Kingdom,the German feed-in

287、 tariff scheme for renewables,and Californias Low Carbon Fuel Standardthese policy packages are setting the stage for companies looking to scale a wide range of zero-carbon technologies and processes,as well as drive down costs.While the climate technology space has largely been known for its start-

288、ups,such as Northvolt,we are already seeing encouraging examples of incumbents tapping into green business building.A German multibillion-dollar revenue technology group has announced that by 2030,new climate technology areas such as hydrogen electrolyzers will account for 70 percent of its business

289、.In Asia,an Indonesian mining company is planning to cut income from coal by 50 percent,while investing hundreds of millions in renewable energy,and build an electric-vehicle ecosystem in the country.In taking on green business building,incumbents are still challenged by start-ups with a DNA of inno

290、vation and ambition.To scale at the pace thats often required to reach competitive cost levels,incumbents will likely need to push themselves beyond their comfort zone.11 When getting started,companies should avoid fragmentation of efforts and investments and resist the tendency to maximize cost syn

291、ergies between the core and growth businesses.For some,a way to steer clear of these pitfalls has been to bring in external investors to the growth ventures.This type of setup provides scaling experience that many incumbents lack,and forces governance thats arms length from the core business.12Execu

292、te at digital speed to create competitive distanceAs discussed above,there are advantages to being first or early to market with low-carbon offerings.Companies that execute quickly and effectively can capture the largest green premiums,bring costs down faster to earn higher margins,and reap the capi

293、tal expenditure benefits from getting projects done faster.Executing at high speeds is often more familiar to digital players.Commercializing green technologies typically requires significant investments in physical assets,which isnt required for software development or digital engineering.Still,gre

294、en 9 Matt Banholzer,Ralf Dreischmeier,Laura LaBerge,and Ari Libarikian,“Business building:The path to resilience in uncertain times,”McKinsey,December 19,2022.10 More than 5,000 companies have made or are in the process of making emission reduction commitments through the Science Based Targets initi

295、ative.11 For more,see Rob Bland,Anna Granskog,and Tomas Nauclr,“Accelerating toward net zero:The green business building opportunity,”McKinsey,June 14,2022.12 For more,see Tomas Beerthuis,Ralf Dreischmeier,Tomas Laboutka,and Nimal Manuel,“A practical guide to new-business building for incumbents,”Mc

296、Kinsey,June 21,2023.26Full throttle on net zero:Creating value in the face of uncertaintybusiness builders can learn lessons from successful digital scale-ups.Executing the actions weve laid out in this article at digital speed takes both the right mindset and the right capabilities.Companies can co

297、nsider a few approaches:Reducing costs and carbon at speed first requires an honest analysis of the trajectory of current emission reduction activities compared with the trajectory required to meet customer demands or net-zero commitments.Companies can then move with urgency to collaborate with supp

298、liers and partners that can help the company meet both their cost and carbon reduction goals.Being fast in forging partnerships to gain market share is first and foremost about being in the right executive-level discussions with potential partners in the early stages of business building.The early d

299、ialogues provide an opportunity to shape the value proposition before all parameters on the product side have been locked in.The speed of portfolio rotation is,of course,contingent on the availability of buyers and sellers and converging views on valuations.That said,players with a well-anchored por

300、tfolio strategy and serial M&A capabilities are likely to find it easier to execute at pace.For them,the decision making related to each deal can focus on the specifics of the transaction at hand,rather than a comprehensive debate on whether it is beneficial to exit or enter a certain business or wh

301、at the true market potential of that business is.In our experience,companies that have built and scaled green businesses successfullyand quicklytend to take a series of key actions.They lead with game-changing ambition,sign up captive demand before scaling,and often build capacity with parallel scal

302、ing.13 For digital start-ups,the funding dynamics often force leaders to challenge conventional wisdom about how quickly an investment project can be planned,engineered,and executed,or how quickly a new concept can be turned into a product available for customers.Green business builders,whether star

303、t-ups or incumbents,could start with such a mentality.Despite the economic uncertainty,there are opportunities for companies to play offense and accelerate value creation in the net-zero transition.Building a set of strategic moves now could set up early movers for cost and carbon reductions,green p

304、remiums,strong market positions,and new capabilities.Players that choose to slow down could find themselves lagging behind.Laura Corb and Daniel Pacthod are senior partners in McKinseys New York office,Anna Granskog is a partner in McKinseys Helsinki office,and Tomas Nauclr is a senior partner in th

305、e Stockholm office.13 For more,see Rob Bland,Anna Granskog,and Tomas Nauclr,“Accelerating toward net zero:The green business building opportunity,”McKinsey,June 14,2022;and“Scaling green businesses:Next moves for leaders,”McKinsey,March 10,2023.Designed by McKinsey Global PublishingCopyright 2023 Mc

306、Kinsey&Company.All rights reserved.27Full throttle on net zero:Creating value in the face of uncertaintyMarch 2023Scaling green businesses:Next moves for leadersNew challengesand opportunitieshave emerged for green business builders.A set of actions could help companies scale during these uncertain

307、times.This article is a collaborative effort by Rob Bland,Laura Corb,Anna Granskog,Tomas Nauclr,and Giulia Siccardo,representing views from McKinseys Sustainability Practice.The transition to net zero is well underway,but it is not happening fast enough.Growth in key climate technologies,including w

308、ind and solar power and electric vehicles(EVs),has helped accelerate decarbonization efforts worldwide.Solutions such as green hydrogen and long-duration energy storage(LDES)are becoming available and,if scaled,could reduce global emissions even further.But the pace of scaling these technologies has

309、 not kept up with projections for a warming planet.Governments and companies have done an admirable job developing and deploying climate technologies to date,but a significant acceleration is required to meet net-zero targetsand stave off the most dire effects of climate change.Last year,we released

310、 a framework for launching and scaling green businesses,based on our work with both incumbents and start-ups.1 A few of the key actions include leading with game-changing ambition,signing up captive demand before scaling,and building capacity with parallel scaling.In the interim,as the economic and

311、geopolitical backdrop has changed,market dynamics for green business builders have shifted in both nuanced and fundamental ways.On the one hand,capital markets and public-sector institutions have started to galvanize behind green investments.Policy,including the Green Deal Industrial Plan in Europe

312、and the Inflation Reduction Act(IRA)in the United States,promises to support companies looking to scale climate technologies.At the same time,inflation,economic uncertainty,and the invasion of Ukraine have all complicated the path to net zero.Three areas have emerged that should now be priorities fo

313、r those navigating the challenges and seeking opportunities:building up supply chains(often through cross-sector partnerships),proactively addressing an emerging skills gap,and exploring different avenues for financing and investments.Many of the unique challenges to scaling green businesses remainh

314、igh capital expenditures on physical assets(compared with building digital businesses),higher short-term costs,and customer education and adoption barriers for many sustainable products.However,the urgency to reach net-zero targets has only grown in many markets,and the industrial economy is now bei

315、ng reinvented around a lower-carbon energy system,circular-economy practices,and other emerging models.Companies that can innovate and scale during these fast-moving,uncertain times could set themselves up for exponential growth.Our analysis shows that growing demand for net-zero offerings could gen

316、erate$9 trillion to$12 trillion of annual sales by 2030 across 11 value pools,including transport,power,and consumer goods.In this article,we lay out the evolving landscape for scaling climate technologies and explore three areas of potential action for green business builders.A significant scaling

317、gapMore than 4,000 companies have set or are in the process of committing to emissions reductions2 and 70-plus countries have set net-zero targets.3 How quickly would key climate technologies need to scale to help meet such goals?To arrive at projections,we conducted an analysis of the current growt

318、h trajectory for climate tech relative to current net-zero commitments.Based on our analysis,even mature technologiesincluding wind and solar powerwould need to scale by a factor of six to 14 times faster to remain on track for a 1.5 pathway by 2030(exhibit).41 See Rob Bland,Anna Granskog,and Tomas

319、Nauclr,“Accelerating toward net zero:The green business building opportunity,”McKinsey,June 14,2022.2“Companies taking action,”Science Based Targets,accessed February 22,2023.3“For a livable climate:Net-zero commitments must be backed by credible action,”United Nations,accessed February 22,2023.4 Ba

320、sed on the McKinsey 1.5C achieved commitments scenario,which represents existing commitments from companies and policies from countries.To conduct this analysis,we estimated the current trajectory of supply of key climate technologies(based on current activity)across four categories of maturity:matu

321、re,early adoption,demonstrated at industrial scale,precommercial;factored in current emissions-reductions commitments from countries and governments;and assessed the supply of these technologies that would be required by 2030 to stay on track for a 1.5 pathway.29Scaling green businesses:Next moves f

322、or leadersHistorically,growth in solar and wind has often outpaced projections,and new players entering the market(oil and gas companies,private equity players,and institutional investors,for example)show signs that the current pace of deployment could speed up.5 Nevertheless,the potential gap for r

323、enewables to meet net-zero targets looks steep.Climate technologies that are high-potential but relatively less advanced in their commercialization(compared with renewables)would need to scale at an even greater rate.Consider hydrogen.Our analysis indicates that supply of green hydrogen,which is pro

324、duced with renewables,would need to grow by a factor of 200 times.Next moves for green business buildersScaling climate technologies often requires companies to think and act in bold and innovative ways.While our seven actions for scaling green businesses hold true,they continue to evolve(for a summ

325、ary of the original framework,see sidebar,“Seven actions for scaling green businesses”).ExhibitWeb Exhibit of Annual deployment of climate technologies needed,1 multiples of current supplyBased on the McKinsey 1.5C achieved commitments scenario,which represents existing commitments from companies an

326、d policies from countries.To con-duct this analysis,we estimated the current trajectory of supply of key climate technologies(based on historic and current activity),factored in current emis-sions-reductions commitments from countries and governments,and assessed the supply of these technologies tha

327、t would be required by 2030 to stay on track for a 1.5 pathway.Source:EV-Volumes;IEA;International Renewable Energy Agency;McKinsey analysisTo reach net-zero targets,a set of existing climate technologies would need to scale exponentially by 2030.McKinsey&Company02015 202120302015 202120302015 20212

328、0302015 202120302015 2021203010020030040050060005001,0001,5002,00001020304050607001020304050020406080100Wind powercapacity,gigawattsSolar powercapacity,gigawattsBattery electriccar sales,millionGreen hydrogenelectrolyzercapacity,gigawattsCarbon capture,utilization,and storage,megatons of CO 10020014

329、146MatureEarly adoption5“Renewable-energy development in a net-zero world,”McKinsey,October 28,2022.30Scaling green businesses:Next moves for leadersEconomic uncertainty,inflation,new public funding,technological risks,and supply chain considerations have altered the landscape for green business bui

330、lding.Actions that have become particularly important for organizations during these volatile times include creatively developing supply chains(including through partnerships),proactively addressing emerging skills gaps in the workforce,and exploring new avenues for financing and investment.Build up

331、 the supply chain through cross-sector partnershipsGreen business building efforts are often supply chain building efforts.For hydrogen-powered vehicles to scale and help decarbonize long-haul freight transport,for example,a supply of hydrogen and hydrogen infrastructure also needs to scale.We are i

332、ncreasingly seeing green business builders develop their supply chains by forging partnerships across sectors and,in some cases,creating a growth strategy with complementary players as collaborators.These partnerships are getting a boost from major climate legislation packages in the United States a

333、nd the European Union.For example,the IRA in the United States allocates$369 billion for climate and energy spending,6 with a focus on ventures that address critical gaps in the North American supply chain.These collaborations happen upstream,downstream,or horizontally in the value chain.Upstream partnerships are operational partnerships that propel vertical integration.They occur when a company p