《IIA Research:2023年分析和数据科学团队搭建指引-Part 1(英文版)(19页).pdf》由会员分享,可在线阅读,更多相关《IIA Research:2023年分析和数据科学团队搭建指引-Part 1(英文版)(19页).pdf(19页珍藏版)》请在三个皮匠报告上搜索。
1、Copyright 2023 IIA All Rights ReservedOrganizing Analytics and Data Science Orgs(Part 1)3.6V1Copyright 2023 IIA All Rights ReservedTopic OverviewThis eBook is part of a series on organizing analytics organizations.Part I describes and offers guidance on:The fundamental goals and objectives of organi
2、zational structureA new,continuum-based approach to organizational modelsGuidance for determining where your organization falls on the continuumHow analytics organizations commonly evolve and their pathsQuestions and considerations tied to planning your next structural move2Copyright 2023 IIA All Ri
3、ghts ReservedObjectives of an Organizational StructureDeploying people on the most important and value-adding projects in support of the enterprises business processes and requirements is about doing todays work and maximizing peoples contribution.Developing people so they will have the skills and e
4、xperience for tomorrows work is about maintaining the capability and health of the organization and maximizing peoples individual and collective potential.Maintaining the wide range of analytical processes that are built ensures that they remain accurate,available,and up to date.3Copyright 2023 IIA
5、All Rights ReservedHow the Objectives RelateThese objectives are interrelatedbecause the most important learning and development occurs on the job,and those who build a process understand best what maintenance is required.But they are in opposition whenever theres a choice between:Deploying someone
6、to apply their proven skills to an important new project;Having them help maintain and tune existing processes;orGiving the person a stretch assignment to acquire new skills.Org design boils down to two questions:1.Whats the best way to group people for purposes of deployment,development,and mainten
7、ance?2.What are the best ways to coordinate across groups since no individual group will be perfect for all the needs the organization has?4Copyright 2023 IIA All Rights ReservedThe Organizational Structure ContinuumIn the past,IIA focused on six common organizational models that we saw in the marke
8、tplace.In this update,weve shifted focus to a continuum within which all of those models,and others,fit.This is because it has become clear that mature,successful analytics organizations almost always end up organized with some form of a hybrid model after passing through variations of one or both o
9、f the decentralized and centralized model structures.The remaining companies stick with a centralized model.Well still call out where the old models lie on this continuum,but we think the focus should be on the continuum as opposed to any specific model within it.The graphic illustrates this new vie
10、w,including where the individual models fit.5Copyright 2023 IIA All Rights ReservedThe Centralized Side of the ContinuumIn a centralized model,all analytics and data science resources are part of a single organization within the enterprise.This could be a group that sits at the corporate level and s
11、upports all ofthe organization to some extent or,especially in less mature companies,they can be embedded within a single,forward-thinking business unit.The“centralized”team in a functional model rests within one business unit and often doesnt provide support to any other parts of the organization.V
12、irtually no large companies remain with a functional model in place because,by definition,analyticsarenotacorporate-wideactivity.6Copyright 2023 IIA All Rights ReservedCentralized Side StrengthsCentralized control makes it easier to deploy team members on high-priority strategic projects as well as
13、to coordinate skills development.As a centralized group grows and expands its capabilities,this helps with recruiting new talent by demonstrating the organizations commitment to analytics and providing new hires with an established community to join.In the case of the functional variation of the cen
14、tralized model,the home business unit has dedicated,focused support which will help it succeed and,sincetheunitisfundingtheteam,ithasastrongongoingincentiveto use it effectively.7Copyright 2023 IIA All Rights ReservedCentralized Side WeaknessesA centralized model can create distance between the team
15、,the businesspeople they work with,and the business problems they tackle.Business units tend to feel underserved when they dont have control over resources.This concern of a lack of linkage with stakeholders comes up in every maturity assessmentIIA has executed for a centralized team.In the case of
16、a functional model,units outside of the home unit may have little or no access to analytical talent and are therefore unable to apply analyticseffectivelytothebusiness.Inaconsultingmodel,whentimesturn tough,units will often cut off analytics funding before shrinking their own staff,which can lead to
17、 instability and a constant vacillation between growth and shrinkage of the analytics organization,which is unsustainable long term.8Copyright 2023 IIA All Rights ReservedThe Hybrid Zoneof the ContinuumIn a hybrid model,teams of analytics and data science professionals are in place throughout the or
18、ganization within various business units or functions.A key feature,however,is the existence of some form of central corporate team as well.Central teams can provide strategic direction,standardization,communication,and capability incubation to the broader organization.Considering IIAs traditional m
19、odels,the center of excellence(COE)and federated models land here.In a COE model,one or more COEs exist to address specific needs for the company such as artificial intelligence or analytical technology enablement.In a federatedmodel,the central organization is focusedon supporting the other units a
20、nd will have a range of skill sets.9Copyright 2023 IIA All Rights ReservedHybrid Zone StrengthsA hybrid structure is very good at facilitating innovation and scale.This is because centralized COEs can be created to explore new areas before eventually spreading knowledge more widely.The central team
21、within a hybrid model also enables the pursuit of strategic corporate projects that span multiple business units.In addition,the central team can focus on corporate standards for tools and platforms by both being a single point of purchase for vendors,as well as a resource for training and support f
22、or employees.10Copyright 2023 IIA All Rights ReservedHybrid Zone WeaknessesIn a pure COE model,the COEs often have little formal power and must rely on their relationships and influence.In a federated model governed by a voluntary council,the lack of a single decision-maker can lead to quagmires if
23、an agreement cant be reached within the council.Since there is nobody to“break the tie,”stalemates can be reached that have no clear path to resolution.If there are a lot of distributed teams,career progression can be hindered because many small teams will each have a mid-level leaderIt is worth not
24、ing that hybrid models wont work for smaller organizations because there simply arent enough resources to go around.There is a certain level of scale required before it is feasible to implement a hybrid model.11Copyright 2023 IIA All Rights ReservedThe Decentralized Sideof the ContinuumIn a decentra
25、lized model,analytics and data science resources are housed withinvariouslocationswithinthecompany andthereisnoformaleffortto coordinate oralign them.Each teamis,in effect,a standalone entity that is completely distinct from the others.While many companies begin their analytics journey with this mod
26、el,few remain in this state.Thisendofthecontinuummapstothetraditionaldecentralizedmodelthat IIA discussed in our past research.A decentralized model simply has multiple functional teams that are not coordinated by any strategy instead of a single team.12Copyright 2023 IIA All Rights ReservedDecentra
27、lized Side StrengthsThe primary strength of a decentralized model is that all resources are located very close to the business function that they support and are a part of that function.This means that the analytics team will be very responsive to the needs of their function and there will be no dis
28、tractions from serving that function exclusively.Arguably,a strength is the ability of any business unit to stand up a team if they desire without need to get permission from,or gain alignment with,any other parts of the company.There are few barriers to business units getting started with their own
29、 programs,even if in isolation.13Copyright 2023 IIA All Rights ReservedDecentralized Side WeaknessesThere are lots of weaknesses to go over here.First,there is no mechanism to support corporate or cross-unit requirements.Each team focuses on its own small part of the company and so many initiatives
30、that could be of great value to the company will never happen.Next,there can be a lot of redundant cost and effort as teams individually invest in tools and solution creation.To the extent one team has a solution another team could benefit from,it will not be known to the teams,and each team will re
31、dundantly create their own identical solutions.Next,there is ample room for a divergence in tools,methodologies,and strategy across the teams,which can obviously be bad for the company.Finally,there is limited career opportunity for team members as they can only advance within their own small teaman
32、d they dont have internal resources to collaborate with or learn from other departments.14Copyright 2023 IIA All Rights ReservedEvolving an Organizational Structure Over Time15Thereareseveralcommon pathsthatorganizations follow overtimeasthey mature and grow their analytical capabilities.Well cover
33、4 paths in this eBook.Typically,these paths are not explicitly pursued as much as they naturally occur as analytical demand,success stories,and sophistication increase.However,this does not mean that organizations cannot learn by understanding these common paths,where they fall withinthem,and how th
34、ey might accelerate their next steps along the path the company is on.IIA is not suggesting that these paths are the“correct”paths or the only paths;they are simply some of the most common paths we see taken.Copyright 2023 IIA All Rights ReservedEvolution Path 1:Functional to Centralized to HybridIn
35、 these situations,a company begins with a single,functionally based team.The team is a huge success and has large impacts on the home function.Other functions within the organization hear about this success and start to ask to tap into the team,which can lead to a consulting model where the home fun
36、ction“rents out”analytical resources to other functions to provide services.Over time,however,it is recognized that each business unit or function needs substantive and dedicated support.The need for corporate support and a unified strategy is also recognized.So,a hybrid model eventually evolves wit
37、h a central corporate team.Once a CAO,CDAO,or CDO is hired,they take the helm within the central team and move the organization forward.16Copyright 2023 IIA All Rights ReservedEvolution Path 2:Functional to Decentralized to HybridAs in the prior path,it has substantive and highly visible success.How
38、ever,instead of trying to borrow resources from the initial team,other business functions and units decide to mimic the model and add their own analytics and data science teams.This strategy,of course,quickly leads to a decentralized organization spread out across the company.Over time,the lack of c
39、oordination,redundancy of effort,and lack of consistent strategy is recognized as a problem.This leads either to all teams being combined into a centralized organization or,more commonly,the addition of a central corporate team.Eventually,the central team is given official authority over the dispers
40、ed teams,which is often formalized with the hiring of a CAO,CDAO,or CDO to take the helm.17Copyright 2023 IIA All Rights ReservedBecome an IIA Client to Access the Full Brief IIA clients get access to the full Organizing Analytics and Data Science Organizations research briefs online and as a PDF wh
41、ich contains a walkthrough on how to plan your organizations path forward and when to use each organizational model.Research&Advisory Network(RAN)clients have direct access to the experts that developed this content and framework with on demand inquires.18Copyright 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