《IIA Research:2023年分析和人工智能战略指南(英文版)(16页).pdf》由会员分享,可在线阅读,更多相关《IIA Research:2023年分析和人工智能战略指南(英文版)(16页).pdf(16页珍藏版)》请在三个皮匠报告上搜索。
1、Copyright 2023 IIA All Rights ReservedAnalytics and AI Strategy Guide3.6V1Copyright 2023 IIA All Rights ReservedThe Objectives of an Analytics StrategyThere are two primary purposes of an analytics and AI strategy.One is to aid the overall organization in its application of these resources.For this
2、purpose,a strategy would address topics such as what kinds of applications or use cases should be the companys focus;what kinds of talent it needs;and what kinds of data it should have.Eventually,analytics and AI should be the province of every function and unit within the organization,and they can
3、all be guided in their AI initiatives by the strategy.The second purpose is to guide analytical and AI professionals within the organization.An effective strategy could help them build the right use cases;employ the right tools and methods;hire the right staff;work in the right ways with other parts
4、 of the organization;and provide the right kinds of value.2Copyright 2022 IIA All Rights ReservedEvolving an Organizational Structure Over TimeCopyright 2023 IIA All Rights ReservedDetermining Areas to beAddressedOne essential question is just what kinds of activities are covered by the strategy.“An
5、alytics”is a broad term and may encompass activities ranging from:Business IntelligenceAdvanced AnalyticsAIData InfrastructureSome companies may include,for example,digital or web analytics in an overall strategy,and some may include that category in a separate domain and strategy.Analytics strategy
6、 may incorporate operational analytics in manufacturing environments.Its important to decide and announce not only what areas are included,but those that are not.4Copyright 2023 IIA All Rights ReservedThe Strategy Process Pt.14 Aspects of the Process:Formality-A strategy can be formal or informal.An
7、 informal strategy would involve discussions with senior executives and analytics/AI leaders about objectives,key projects,funding,and so forth.A more formal approach with a strategy document captures strategic deliberations in some sort of written text and makes broad awareness and action on the st
8、rategy more likely.Participation-This type of strategy development requires marrying deep domain knowledge with extensive awareness of analytics and AI capabilities.Candidates should have the following traits:Knowledge of the most up-to-date analytics and AI practices and how they are used in the bu
9、sinessEffectively communicate to managers in non-technical terms;Understand the key issues of the business overall and its current strategic directionUnderstand design thinking5Copyright 2023 IIA All Rights ReservedThe Strategy Process Pt.24 Aspects of the Process cont.Assessment-Since an analytics
10、and AI strategy is generally intended to improve capabilities and outcomes,it is often a good idea to perform an assessment of existing capabilities before starting a strategy effort.IIA has extensive experience with assessments,and by using its diagnostic tools,companies can compare their existing
11、capabilities with companies in their industry or others.That comparison can provide motivation and insight for an analytics and AI strategy,furnishing the starting point for a detailed roadmap.Link to Business Strategy-Analytics and AI activities should,at a minimum,reflect and support an organizati
12、ons business strategy.However,in companies for which data and analytics are considered strategic resources,the link to business strategy may be more direct.6Copyright 2023 IIA All Rights ReservedOperational ModelsOperational models are often combined with organizational and funding models.Below are
13、several common(but not exhaustive)combinations:Service bureauHub-and-spoke or federal modelsStrategic applications onlyCenter of excellence modelSelf-service modelOutsourced modelThe choice of operational and organizational model is likely to change over time in most companies.One large retailer,for
14、 example,reorganized and changed quickly from a centralized model focused on building important analytics and AI applications and reporting to a chief digital officer on the business side,to a decentralized model reporting to the chief information officer and focused on meeting all needs(service bur
15、eau)for analytics and AI services.7Copyright 2023 IIA All Rights ReservedFunding ModelsIf the operational and organizational model doesnt include a funding model,AI,and analytics strategy should include it to specify the financing of these technologies in the business.Investment in these areas is on
16、 the rise in 92%of firms in the 2022 NewVantagePartners survey.That survey also found that the same percentage was achieving returns on their data and AI investments,almost double the percentage reporting returns in 2017.Funding model decisions should address where the initial funding for executing
17、the strategy will be obtained;how internal(and in some cases external)customers will be charged for their analytics and AI projects;and how return on investment will be measured.8Copyright 2023 IIA All Rights ReservedApplications/Use CasesAlmost every organization must face the issue of is where in
18、the business to apply analytics and AI.What business problems,issues,or opportunities can benefit from the use of these powerful tools,and what applications or use cases can address them?Some analytics and AI applications seem to yield more value than others;a 2021 MIT Sloan Management Review survey
19、 found that companies that used AI primarily to create new value were 2.5 times more likely to feel that AI is helping their company competitively compared with those that said they are using AI primarily to improve existing processes.This decision of which use cases to emphasize should largely be d
20、riven by business strategy.9Copyright 2023 IIA All Rights ReservedAnalytical/AI Service LineAs analytics and AI grow more popular and central to business strategies,there is also a need to produce successful outcomes repeatedly,reliably,and quickly.Analytics and AI groups need to have“service lines”
21、in place.Service lines are necessary because the customers of analytical groups within organizations need both quick and reliable delivery of analytics/AI,but they also need to be familiar with the possibilities for such work.A service line should have the following attributes:Analysis has been prov
22、ided before,ideally several timesand its achieved a good outcomeThere is some sense of a process that is followed to produce the desired resultData is either readily available for the analysis,or it can be found easilyIts clear what the likely decision outcomes are for the analysisLikely customers f
23、or this service within your organization or client have been identifiedThere is some degree of marketing material to describe this service and its benefits.10Copyright 2023 IIA All Rights ReservedTechnologyAnalytics and AI are not one technology,but a collection of themincluding data management tool
24、s,business intelligence tools,visual analytics,statistical machine learning,neural networks,natural language processing and generation,and robotic process automation.In order to make wide-reaching technology decisions,an organization needs not only a clear idea of what business objectives it wants t
25、o accomplishbut also what specific analytical and AI methods it needs to use to solve them.In order to ascertain the right technology to build or buy,a company needs to involve not only executives but also expert analysts,IT professionals,and data scientists.11Copyright 2023 IIA All Rights ReservedT
26、alentThe choices for talent strategy are similar to those for technologies:buy,build,or rent.To buy people is to hire those who already possess the needed skills.Hiring talent will be particularly difficult if your company is not located in large East Coast or West Coast cities,and it is not willing
27、 to pay large compensation figures.To“build”people is to train them in the needed skills.This will be much less difficult if the candidates for such training already have basic statistical and data management skills.It is relatively rare for companies to engage in substantial efforts to train or ret
28、rain their employees in analytics and data science skills,but it should be more common as companies have had great success when undertaking large retraining programs.A third basic option is to“rent”an analytically-informed workforce,which is to hire consultants or vendors to deliver services.This st
29、rategy is widely practiced by companies that dont have the in-house expertise to build analytics and AI applications.12Copyright 2023 IIA All Rights ReservedChange ManagementProjects employing analytics and AI are not just about technical change,but also about changes in organizational culture,behav
30、ior,and attitudes.Change management is one of the most important components in a successful analytics and AI program;without it,models and systems may be developed,but not successfully deployed.Typical change management approaches include stakeholder management,frequent communications,training and u
31、pskilling,and careful planning for deployment.Change management techniques may or may not be a formal component of strategy.Approaches to change management have not changed much over the years,and there is not much debate about their key elements.Change management initiatives may only need to be cal
32、led out in an analytics and AI strategy if they are not well understood or followed.13Copyright 2023 IIA All Rights ReservedDisseminate and Monitor3 Aspects of the Process:Documentation Distribution-An executive summary and in-depth strategy document are necessary for adoption among key stakeholders
33、 and analytics/AI personnel respectively.Briefings/Education-More effective dissemination approaches usually involve one-on-one or small group meetings with senior executives,seminars,and webinars.The stakeholders for a strategy includenot only senior executives,but also members of analytics and AI
34、groups,all employees of an organization,and in some cases even external customers and suppliers.Periodic MonitoringThere are three approaches to analyzing the effectiveness of the strategy:1.Simply create and monitor a set of KPIs tied to the objectives.This would include a description of which use
35、cases have been successfully developed and deployed,funding used,returns on investment,etc.2.Regularly survey key stakeholders for the strategy and assess their satisfaction.3.Record what key decisions were made on the basis of it,including directions and actions not taken.14Copyright 2023 IIA All R
36、ights ReservedBecome an IIA Client to Access the Full Brief IIA clients get access to the full Analytics and AI Strategy research brief online and as a PDF which contains examples of each strategy piece and a walkthrough on how to plan your organizations path forward.Research&Advisory Network(RAN)cl
37、ients have direct access to the experts that developed this content and other frameworks through on-demand inquires.15Copyright 2023 IIA All Rights ReservedDont go it alone.Let IIA guide your |We work with clients to build and grow their analytics organizations.Benefit from our unbiased,unrivaled network of analytics experts,practitioners and thought leaders.3.6V1