Redefining Work, Workforces, and Workplaces: How Organizations Can Redesign What Gets Done, Who or What Performs It, and Where Value Is Created in the Age of AI

1. Work: the what

What outcomes must be created, which tasks produce them, and how should human and technological capabilities be combined?

2. Workforce: the who

Which employees, contractors, providers, freelancers, AI agents, and machines should perform the work?

3. Workplace: the where

Which physical, digital, hybrid, distributed, or customer-based environments best support the required outcomes? The original Deloitte article argues that the future of work is driven by multiple converging forces rather than technology alone. It describes work becoming less tied to fixed jobs, talent becoming less tied to one employer, and workplaces becoming less tied to one physical location. It recommends that organizations imagine, compose, and activate a preferred future rather than merely reacting to disruption. The first major principle is that organizations should begin with the work itself. Traditional industrial organizations divided complete outcomes into specialized tasks and grouped those tasks into fixed jobs. AI, automation, and digital platforms now make it possible to reorganize those tasks again. Repeatable activities may be automated, information-intensive work may be augmented by AI, and people may focus more heavily on judgment, relationships, creativity, interpretation, and exception management. The second principle is that the workforce should be understood as an ecosystem rather than a payroll list.

Modern organizations can access capability through:

• Permanent employees
• Part-time employees
• Contractors
• Freelancers
• Consultants
• Managed-service providers
• Outsourcing partners
• Crowdsourcing platforms
• AI copilots
• Autonomous software agents
• Robots

Deloitte proposes that talent management should expand beyond attracting, developing, and retaining employees toward accessing, curating, and engaging a broader workforce ecosystem. The third principle is that workplace design should follow work requirements.

The question should not begin with:

How many days should employees be in the office?

It should begin with:

What does this work require to produce the best outcome? Some activities benefit from quiet individual focus. Others benefit from physical proximity, specialized equipment, mentoring, high-trust collaboration, or customer presence. Deloitte’s later workplace research found that leaders widely consider workplace strategy important, but relatively few believe their organizations are highly prepared to design it. The research emphasizes that work design should determine the appropriate mix of physical and digital environments. The framework has become even more important in the AI era. Deloitte’s 2026 human-capital research reports that seven in ten business leaders identify speed and adaptability as their primary competitive strategy for the next three years. Leaders also identify faster orchestration of people and resources and greater organizational adaptability as major success factors. At the same time, workforce transformation remains difficult. The World Economic Forum reports that 63 percent of surveyed employers identify skill gaps as a major barrier to business transformation, while organizational culture and resistance to change are the second most frequently cited barrier.

A practical response requires organizations to redesign all three dimensions together:

Redefine work

• Begin with outcomes.
• Deconstruct jobs into tasks.
• Eliminate unnecessary work.
• Automate appropriate activities.
• Augment human judgment with AI.
• Clarify accountability for agent-operated workflows.

Recompose the workforce

• Identify capabilities rather than merely headcount.
• Decide which knowledge must remain internal.
• Use different talent models intentionally.
• Build internal mobility and skills visibility.
• Govern providers, contractors, agents, and employees as one ecosystem.

Redesign the workplace

• Let the work determine location.
• Design for focus, collaboration, learning, and customer value.
• Create equal access to information and opportunity.
• Protect connection, trust, culture, and social capital.
• Measure outcomes rather than presence.

The central lesson is:

The future of work is not primarily about choosing between humans and machines, employees and contractors, or offices and homes. It is about designing a coherent system in which work, workforce, and workplace reinforce one another.

1. The Future of Work Is a System-Design Challenge

Organizations often discuss the future of work through separate debates:

• Will AI replace jobs?
• Should employees return to the office?
• Should the company use more contractors?
• How should managers supervise remote workers?
• Which skills will be important?

These questions are connected. A decision about AI changes the tasks inside jobs. A change in tasks alters skill requirements. New skill requirements affect whether work should be performed by employees, contractors, service providers, or agents. That decision influences where the work can be performed. Workplace design affects collaboration, learning, culture, and productivity. The future of work is therefore not a collection of independent HR policies. It is an organizational system. Deloitte’s framework is valuable because it separates the system into three dimensions without treating them as isolated. Work, workforce, and workplace influence one another continuously.

2. Work: The What

The first dimension asks:

What must be accomplished, and how should the work be designed? This is more fundamental than asking how many employees are needed.

Organizations exist to create outcomes such as:

• Serving customers
• Designing products
• Operating infrastructure
• Managing risk
• Delivering healthcare
• Producing goods
• Educating people
• Processing transactions

Each outcome is created through a combination of tasks. Traditional job descriptions package many tasks together under one title.

A customer-service representative may:

• Verify identity
• Search records
• Interpret policy
• explain options
• update systems
• calm an upset customer
• escalate an exception

AI and automation can affect each activity differently. The job should therefore be deconstructed before leaders decide how technology will affect it.

3. From Complete Craft to Fragmented Job

Deloitte’s historical framing is useful. Before industrialization, many workers were responsible for complete outcomes. A craftsperson might interact with the customer, select materials, produce the item, inspect the result, and make final adjustments. Industrialization divided complete outcomes into repeatable tasks. This increased efficiency and scale, but it also created jobs that were collections of narrow activities rather than complete ownership of an outcome. The cognitive and AI era now creates an opportunity to reorganize work again. Machines can increasingly handle standardized tasks.

People can potentially return to more integrated responsibility for:

• Solving problems
• Interpreting context
• serving customers
• managing relationships
• improving systems
• delivering outcomes

Deloitte argues that human-machine collaboration may shift the center of human work away from routine task completion and toward problem-solving, relationships, interpretation, and service.

4. Jobs Should Be Deconstructed Into Tasks

A practical work-redesign process should examine every major activity according to:

• Frequency
• Time required
• Standardization
• Data availability
• Error consequences
• Human judgment
• Relationship requirements
• Technology readiness
• Regulatory constraints

Tasks can then be classified. Eliminate Some work exists because of historical processes, duplicated controls, weak systems, or unclear ownership.

Examples include:

• Reports nobody uses
• Duplicate data entry
• Repeated approvals
• Manual reconciliation caused by disconnected systems
• Meetings without a decision purpose

Eliminating unnecessary work is often more valuable than automating it. Automate Structured, repeatable, low-ambiguity work may be performed through conventional automation.

Examples include:

• Data transfer
• Standard routing
• Scheduling
• Repetitive calculations
• System provisioning

Augment AI can support a person while the person retains responsibility.

Examples include:

• Drafting
• research
• summarization
• anomaly detection
• software generation
• scenario analysis

Delegate to an agent A governed agent may complete a multistep workflow within defined limits.

Examples include:

• Resolving routine service requests
• Monitoring systems
• preparing standard procurement actions
• updating approved records
• coordinating follow-ups

Preserve as human-led

Human control remains especially important when the activity requires:

• Empathy
• ethical judgment
• leadership
• legal accountability
• complex negotiation
• trust
• ambiguous strategy
• safety-critical decisions

5. Redefining Work Is More Than Increasing Efficiency

Organizations often begin automation with a cost target. Cost matters, but work redesign can produce broader value.

It may improve:

• Customer experience
• speed
• quality
• innovation
• employee capacity
• safety
• resilience
• accessibility

A contact-center AI system, for example, may be designed primarily to reduce staffing.

A broader redesign might instead use AI to:

• Resolve routine requests automatically
• give employees better customer context
• identify emotional escalation
• reduce repetitive documentation
• allow people to focus on complex cases

The second design creates both efficiency and higher-value human work. Deloitte warns that organizations can choose to use technology narrowly for efficiency and cost reduction or more broadly to increase value and meaning for businesses, customers, and workers.

6. AI Should Be Applied to Workflows, Not Isolated Tasks

Automating one task does not guarantee an improved outcome. Suppose AI reduces report-writing time from four hours to thirty minutes.

If the report still:

• Requires several manual data extracts
• passes through six approvals
• contains metrics no one uses
• arrives too late for the decision

then little value has been created. Work should be redesigned end to end.

The process should ask:

1. What outcome is required?

2. Which steps contribute to it?

3. Which steps should stop?

4. Where is judgment necessary?

5. Which technology can help?

6. Who remains accountable?

7. How will the result be measured?

7. Human-Machine Collaboration Needs Explicit Design

A vague statement that “AI will assist employees” is not enough.

For every AI-enabled workflow, define:

• What the AI performs
• What the employee performs
• Which outputs require review
• Who manages exceptions
• Who owns the final decision
• Which actions the system may execute
• When the workflow must stop
• How failures are reported

Human oversight is meaningful only when the person has:

• Time
• information
• authority
• competence
• accountability

Without these elements, “human in the loop” may become ceremonial approval.

8. Work Design Should Protect Learning

Routine tasks are often where people gain foundational experience.

Junior employees learn through:

• Research
• drafting
• basic analysis
• documentation
• customer interaction
• routine coding

AI may absorb many of these activities.

Organizations should preserve their developmental function through:

• Apprenticeships
• simulations
• rotations
• supervised AI work
• structured review
• mentoring
• customer exposure

A company that removes entry-level work without creating new learning pathways may improve short-term efficiency while weakening its future expert pipeline. Workforce: The Who

9. The Workforce Is Becoming an Ecosystem

The second dimension asks:

Who or what should perform the work? The answer is no longer limited to full-time employees.

Organizations may use:

• Employees
• Temporary staff
• Contractors
• Independent professionals
• Consulting firms
• Managed-service providers
• Outsourcing partners
• Crowdsourcing communities
• AI agents
• Robots

Deloitte describes a continuum of talent models and argues that organizations need new ways to orchestrate the growing complexity of alternative labor sources. This creates flexibility. It also creates significant governance challenges.

10. The Workforce Should Be Planned Around Capability

Traditional workforce planning often focuses on:

• Positions
• headcount
• salary budgets
• vacancies

A workforce-ecosystem approach focuses on capabilities.

For each capability, leaders should ask:

• Is it strategically differentiating?
• Is it needed continuously?
• Does it involve sensitive knowledge?
• Is demand stable or temporary?
• Can the company develop it internally?
• Can a provider offer superior scale?
• Can AI perform part of the work?
• How reversible is the sourcing decision?

This produces a more deliberate workforce mix.

11. Permanent Employees

Permanent employment is often appropriate when the capability:

• Creates strategic differentiation
• Requires deep company knowledge
• Is needed continuously
• Includes leadership or accountability
• Should remain under internal control

Examples may include:

• Product ownership
• enterprise architecture
• cybersecurity leadership
• data governance
• customer strategy
• AI governance

12. Contractors and Freelancers

Independent talent may be useful when the need is:

• Temporary
• specialized
• variable
• project-based
• urgent

These workers can provide flexibility and expertise.

Risks include:

• Knowledge loss
• unclear intellectual-property ownership
• security exposure
• dependence on individuals
• weaker organizational connection

13. Managed Services and Outsourcing

Providers may be appropriate for capabilities requiring:

• Around-the-clock operation
• specialist platforms
• global coverage
• standardized delivery
• economies of scale

The company should retain enough internal expertise to:

• Set strategy
• evaluate performance
• manage risk
• control architecture
• change providers

Execution can be externalized. Understanding and accountability should not be.

14. AI Agents as Workforce Participants

AI agents should not be treated as employees, but they should be included in capacity and operating-model planning.

For each agent, organizations should define:

• Tasks
• permissions
• data access
• escalation
• monitoring
• human ownership
• operating cost
• failure procedures

The workforce plan should estimate the productive capacity created by agents while avoiding unrealistic assumptions about perfect autonomy.

15. From Attract, Develop, Retain to Access, Curate, Engage

Deloitte proposes a broader talent-management model. Access How can the organization reach needed capability across internal and external ecosystems?

This may include:

• Internal talent marketplaces
• contractors
• partners
• providers
• crowdsourcing
• digital platforms

Curate How can workers receive meaningful development through work experiences, learning, mobility, and career pathways? Development should not be limited to formal courses.

It should include:

• Projects
• rotations
• mentoring
• communities
• assignments
• AI-supported learning

Engage How does the organization build productive and compelling relationships with different workforce groups? Permanent employees, contractors, providers, and partners may have different relationships with the organization, but all need:

• Clear outcomes
• information
• coordination
• trust
• appropriate support

This broader model reflects the reality that value may be created by people who do not follow one traditional employee life cycle.

16. The Workforce Ecosystem Needs Clear Accountability

A blended workforce creates ambiguity unless decision rights are explicit.

For every critical outcome, define:

• Business owner
• product owner
• technical owner
• risk owner
• provider responsibilities
• agent permissions
• escalation authority

The fact that several parties contribute does not mean accountability can be distributed vaguely among them.

17. Internal Talent Mobility Becomes Essential

Organizations frequently possess relevant skills that remain hidden because employees are represented mainly by their job titles.

A person may have:

• Previous industry experience
• certifications
• project work
• technical knowledge
• career interests

Internal talent marketplaces can connect employees with:

• Permanent roles
• temporary projects
• stretch assignments
• mentors
• learning opportunities

This allows organizations to access existing capability before defaulting to external hiring.

18. Skills Data Must Be More Reliable

Organizations need better information about:

• Current skills
• proficiency
• experience
• interests
• mobility
• future role requirements

Useful evidence may include:

• Work history
• project outcomes
• certifications
• assessments
• manager input
• employee validation
• work samples

Skills systems should not become opaque employee-scoring mechanisms. Workers should be able to review and correct their information.

19. Managers Must Learn to Orchestrate Mixed Teams

Future managers may supervise combinations of:

• Employees
• contractors
• provider teams
• AI copilots
• autonomous agents

They will need to manage:

• Outcomes rather than attendance
• cross-organizational collaboration
• security boundaries
• knowledge transfer
• human-agent accountability
• different incentives and employment relationships

Management becomes less about controlling one co-located team and more about orchestrating a capability network.

20. Culture Must Extend Beyond the Payroll

Organizations often talk about culture as something experienced only by employees.

Yet contractors, partners, and providers may interact directly with:

• Customers
• systems
• data
• products

They influence quality and trust.

The organization should communicate relevant:

• Values
• customer standards
• security expectations
• decision principles
• ethical requirements

This does not mean treating every provider employee as an internal employee. It means recognizing that the customer experiences the complete ecosystem. Workplace: The Where

21. Workplace Is No Longer Synonymous With Office

The third dimension asks:

Where and through which environments should work occur?

The answer may include:

• Corporate offices
• homes
• customer locations
• factories
• hospitals
• laboratories
• coworking spaces
• virtual platforms
• digital twins
• mobile environments

Deloitte argues that workplace design should evolve as both the work and the workforce change. Digital communication and collaboration systems allow teams to operate across a continuum from co-located to fully distributed.

22. Let the Work Determine the Workplace

A universal office policy assumes that all work has similar needs. It does not. Different activities may require different environments. Focus work

Examples:

• Coding
• writing
• analysis
• research

These activities may benefit from quiet and limited interruption. Collaborative work

Examples:

• Product discovery
• strategy
• complex design
• crisis response

These may benefit from rapid interaction and shared context. Relationship work

Examples:

• Mentoring
• team formation
• negotiation
• difficult conversations

These may benefit from richer human interaction. Physical or specialized work

Examples:

• Manufacturing
• clinical care
• laboratory research
• equipment maintenance

These require access to specific places, people, or tools. Deloitte’s workplace research explicitly recommends beginning with the question, “What does the work require?” rather than applying one universal workplace model.

23. Hybrid Work Is a Design Problem, Not a Scheduling Compromise

Many hybrid models are built by selecting a number of office days. That is a scheduling decision, not a complete work design.

A stronger hybrid model defines:

• Which activities are synchronous
• which are asynchronous
• which require physical presence
• how decisions are documented
• how information is shared
• how mentoring occurs
• how remote workers access opportunity

Without these elements, hybrid work may produce:

• Excessive meetings
• duplicated communication
• proximity bias
• weak onboarding
• employee isolation

24. The Office Needs a Clear Purpose

If employees can perform routine computer work elsewhere, the office must provide distinctive value.

Possible purposes include:

• High-trust collaboration
• mentoring
• social connection
• innovation
• customer engagement
• specialized technology
• cultural rituals
• team formation

The office should not be evaluated only through occupancy.

It should be evaluated through outcomes such as:

• Better collaboration
• faster decisions
• stronger relationships
• improved learning
• employee connection

25. Digital Workplaces Need Architecture

A digital workplace is more than a collection of communication tools.

It requires coherent design across:

• Identity
• knowledge
• collaboration
• workflow
• search
• documentation
• security
• employee support

Poor digital design creates:

• Information fragmentation
• notification overload
• duplicate documents
• inaccessible knowledge
• meeting dependence

The digital environment should make it easy to:

• Find information
• understand decisions
• collaborate asynchronously
• complete work
• preserve organizational knowledge

26. Workplace Design Must Include Frontline Workers

Future-workplace discussions often focus on office workers.

Frontline workers in:

• Healthcare
• retail
• logistics
• manufacturing
• construction
• field service

also need better technology and work environments.

Digital tools can support:

• Scheduling
• communication
• safety
• knowledge access
• remote assistance
• training
• workflow coordination

Deloitte’s workplace research emphasizes that workplace reinvention should extend beyond knowledge workers to digitally empowered frontline work.

27. Culture Cannot Depend Only on Physical Proximity

Organizations sometimes assume that culture naturally exists when people share a building. Poor culture can exist in an office. Strong culture can exist in a distributed organization.

Culture is created through:

• Leadership behavior
• incentives
• decisions
• recognition
• communication
• fairness
• shared experiences

Physical interaction may strengthen relationships, but it does not replace good management. Deloitte emphasizes the importance of connection and community as work becomes more virtual and contingent.

28. Social Capital Must Be Built Intentionally

Social capital includes:

• Relationships
• trust
• access to information
• informal support
• reputation
• professional networks

Distributed work can weaken accidental contact.

Organizations should design:

• Mentoring
• onboarding cohorts
• communities of practice
• cross-team projects
• periodic gatherings
• informal interaction
• knowledge-sharing rituals

Social connection should be treated as productive infrastructure rather than an optional cultural benefit.

29. Protect Against Proximity Bias

When some employees are frequently visible to leaders and others work remotely, opportunity may become uneven.

Organizations should monitor:

• Promotions
• project assignments
• performance ratings
• access to leadership
• mentoring
• compensation

Decisions should be documented and based on outcomes rather than physical visibility.

30. Workplace Flexibility Has Boundaries

Flexibility does not mean every individual can choose any arrangement regardless of customer, team, or operational needs.

A workable model balances:

• Individual needs
• team coordination
• customer requirements
• security
• regulation
• physical constraints

The objective is not unlimited individual choice. It is purposeful flexibility aligned with outcomes. Composing the Three Dimensions Together

31. Why Separate Decisions Fail

A company may redesign one dimension while ignoring the others. New technology, old jobs AI is installed, but employees keep the same responsibilities and approvals. Flexible workforce, weak governance Contractors and providers are added, but accountability and knowledge transfer remain unclear. Hybrid workplace, old management Employees work remotely, but managers continue measuring presence and scheduling constant meetings. Transformation succeeds only when the three systems are composed together.

32. Example: Customer Support

Work

• Automate authentication and standard requests.
• Use AI for knowledge retrieval and case summaries.
• Preserve human involvement for emotional, high-value, and complex cases.

Workforce

• Internal employees own customer relationships and exceptions.
• A provider handles standardized global coverage.
• AI agents resolve defined low-risk requests.

Workplace

• Routine work may be distributed.
• Complex-case teams may gather for coaching and review.
• Digital systems provide shared knowledge and complete case history.

33. Example: Software Development

Work

• AI generates routine code and tests.
• Engineers focus more on architecture, integration, security, and product decisions.

Workforce

• Internal teams own architecture and product outcomes.
• Specialists and providers support migrations or niche technologies.
• Agents perform controlled development and operational tasks.

Workplace

• Focused coding may occur remotely.
• Architecture and product discovery may use richer synchronous collaboration.
• Documentation and decisions remain accessible digitally.

34. Example: Healthcare Administration

Work

• Automate scheduling, document classification, and routine updates.
• Use AI to summarize patient information.
• Preserve human review for clinical and sensitive decisions.

Workforce

• Clinical professionals retain accountability.
• Service providers may handle selected administrative operations.
• AI supports information processing.

Workplace

• Administrative work may be distributed.
• Clinical work remains tied to patients and facilities.
• Digital platforms connect remote specialists and onsite teams.

Imagine, Compose, Activate

35. Imagine

Deloitte’s first action is to imagine a preferred future. This means defining an ambition broader than cost reduction.

Questions include:

• What customer value could be created?
• What work could become more meaningful?
• What capabilities could distinguish the organization?
• What constraints could technology remove?
• Which employee frustrations should disappear?

Scenario planning should consider different assumptions about:

• AI adoption
• talent availability
• regulation
• customer behavior
• economic conditions

36. Compose

The second action is to design the combination of:

• Work
• talent
• technology
• locations
• platforms
• partnerships

This is where leaders decide:

• Which tasks change
• which roles remain internal
• which providers are used
• how agents are governed
• where work occurs
• how teams collaborate

The composition should be evaluated across:

• Value
• cost
• speed
• risk
• resilience
• employee experience
• reversibility

37. Activate

The third action is to make the design operational.

Activation requires:

• Leadership
• funding
• role redesign
• learning
• change management
• technology implementation
• governance
• performance measurement

A future-of-work vision without activation remains a presentation. Deloitte’s framework emphasizes aligning leadership, workforce development, and organizational practices so that new forms of work and talent can be used effectively. Governance and Measurement

38. Create Cross-Functional Ownership

The future of work should not be owned solely by HR.

It requires:

• Business leadership
• operations
• technology
• finance
• HR
• real estate
• cybersecurity
• legal and risk

Each function sees part of the system. The business owns outcomes. Technology assesses automation and digital architecture. HR manages skills, mobility, and employee practices. Finance connects the model with investment and value. Real estate and operations shape physical environments. Risk functions govern employment, data, security, and AI exposure.

39. Measure Work Outcomes

Useful measures include:

• Cycle time
• quality
• customer satisfaction
• rework
• error rate
• innovation
• cost per outcome

40. Measure Workforce Outcomes

Useful measures include:

• Critical-skill coverage
• internal mobility
• time to productivity
• provider dependency
• employee retention
• talent-market access
• agent performance

41. Measure Workplace Outcomes

Useful measures include:

• Collaboration quality
• employee connection
• focus time
• meeting load
• workplace utilization
• digital-tool friction
• onboarding effectiveness

Presence by itself is a weak measure of productivity.

42. Measure Human Outcomes

Organizations should also track:

• Trust
• autonomy
• well-being
• learning
• workload
• fairness
• career confidence

The workplace system should not produce business value by quietly degrading human sustainability. Current Labor-Market Context

43. Employment Will Grow Unevenly

The US Bureau of Labor Statistics projects total employment to rise by 5.2 million jobs between 2024 and 2034, reaching approximately 175.2 million. Healthcare and social assistance is projected to add almost two million jobs, while professional, scientific, and technical services is projected to add more than 800,000.

This reinforces an important point:

The future of work is not simply a story of disappearing employment. It is a story of changing demand, uneven growth, occupational redesign, and transitions.

44. Skills and Culture Are Major Constraints

The World Economic Forum reports that skill gaps are the most frequently identified barrier to organizational transformation, cited by 63 percent of surveyed employers. Culture and resistance to change rank second. Organizations therefore cannot solve future-of-work challenges through technology procurement alone.

They need:

• Skills
• management
• trust
• mobility
• organizational redesign
• cultural adaptation

45. Continuous Adaptability Is Becoming a Strategic Capability

Deloitte’s 2026 human-capital research indicates that leaders increasingly view speed, resource orchestration, and adaptability as central competitive priorities. It argues that reinvention is becoming continuous rather than episodic and that organizations need workforces capable of learning and adapting in real time. The future-ready organization is not the one that predicts every change. It is the one that can recompose work, workforce, and workplace repeatedly. Common Failure Patterns

46. Starting With Office Policy

A debate over attendance cannot substitute for understanding the work. Begin with outcomes and activities.

47. Starting With Headcount Reduction

This narrows redesign toward labor removal and can obscure opportunities involving:

• Growth
• quality
• customer value
• innovation
• better jobs

48. Treating Contractors as Invisible

External workers and providers are part of the real workforce.

They should be included in:

• Capacity planning
• security
• knowledge management
• risk
• cost analysis

49. Using AI Without Redesigning Jobs

AI tools added to unchanged roles often create:

• Duplicate effort
• unclear review
• increased workload
• inconsistent accountability

50. Assuming Flexibility Automatically Creates Inclusion

Remote work may improve access for some employees while creating isolation or proximity bias for others. Outcomes must be measured.

51. Treating Culture as Office Attendance

Culture is created through behavior, incentives, leadership, fairness, and relationships. A building cannot compensate for weak leadership.

52. Ignoring Frontline Work

A future-of-work strategy focused only on office employees misses large parts of the workforce.

53. Outsourcing Too Much Knowledge

A workforce ecosystem should expand capability without eliminating internal understanding and control.

54. Measuring Activity Instead of Value

Weak measures include:

• Office attendance
• messages sent
• prompts entered
• hours worked
• training completed

Better measures connect activity to:

• Outcomes
• quality
• customer value
• capability
• sustainability

A 12-Month Implementation Roadmap Months 1 - 2: Define ambition

• Establish strategic outcomes.
• Appoint cross-functional leadership.
• Select priority business areas.
• Define principles for human-AI work.
• Establish workforce and workplace goals.

Months 3 - 4: Map the system

• Map workflows and tasks.
• Inventory internal and external talent.
• Map physical and digital workplaces.
• Identify major pain points.
• Establish baseline measures.

Months 5 - 6: Design the future model

• Classify tasks.
• Define future roles.
• Select appropriate talent models.
• Define workplace requirements.
• Clarify decision rights and accountability.

Months 7 - 8: Pilot

• Redesign one meaningful workflow.
• Create a blended team.
• introduce approved automation or AI.
• Test workplace arrangements.
• Measure outcomes and employee experience.

Months 9 - 10: Build supporting infrastructure

• Improve skills data.
• Establish learning pathways.
• Strengthen provider governance.
• Improve digital workplace architecture.
• Train managers in ecosystem orchestration.

Months 11 - 12: Scale and integrate

• Expand successful patterns.
• Connect workforce planning with budgeting.
• Update workplace policies.
• Establish quarterly reviews.
• Retire ineffective practices.

Key Takeaways

1. The future of work should be understood through work, workforce, and workplace.

2. Work is the outcome and task system, not merely a collection of existing jobs.

3. Jobs should be deconstructed before leaders decide how AI will affect them.

4. Technology can eliminate, automate, augment, or agentize different parts of the same role.

5. Work redesign should pursue value, quality, innovation, and meaning alongside efficiency.

6. The workforce is now an ecosystem of employees, specialists, providers, agents, and machines.

7. Talent decisions should be based on capability, strategic importance, risk, and duration.

8. Deloitte’s access, curate, and engage model expands talent management beyond the traditional employee life cycle.

9. Managers increasingly need to orchestrate mixed human and technological teams.

10. The workplace should be determined by what the work requires.

11. Hybrid work is a work-design and management problem, not merely a scheduling policy.

12. The office needs a clear purpose beyond routine computer work.

13. Digital workplaces require intentional architecture, knowledge management, and security.

14. Frontline workers must be included in workplace reinvention.

15. Culture and social capital must be built intentionally in distributed organizations.

16. Work, workforce, and workplace decisions should be composed together.

17. Deloitte’s imagine, compose, and activate framework provides a useful implementation sequence.

18. Skill gaps and organizational resistance remain major barriers to transformation.

19. Employment growth and displacement will occur simultaneously across different occupations and sectors.

20. The most future-ready organizations will be able to recompose their operating systems continuously.

Frequently Asked Questions

What does redefining work mean?

It means redesigning tasks, workflows, roles, and accountability around required outcomes rather than assuming current jobs must remain unchanged.

What is the difference between work and a job?

Work is the activity required to create an outcome. A job is an organizational package of tasks assigned to one person or role.

What does workforce ecosystem mean?

It refers to the full range of people and technologies contributing to organizational outcomes, including employees, contractors, providers, agents, and robots.

What is a workplace?

A workplace is any physical or digital environment through which work occurs. It is not limited to an office.

What are Deloitte’s three future-of-work dimensions?

They are:

• Work: the what
• Workforce: the who
• Workplace: the where

What are imagine, compose, and activate?

They are three actions proposed by Deloitte:

• Imagine the preferred future.
• Compose the required work, workforce, and workplace.
• Activate the design through leadership, development, and organizational change.

What does access, curate, and engage mean?

It is a broader talent model:

• Access capabilities across the ecosystem.
• Curate meaningful learning and work experiences.
• Engage different workforce groups in productive relationships.

Should every automatable task be automated?

No. Leaders should consider cost, accuracy, risk, customer preference, employee impact, and human accountability.

How should AI and humans divide work?

AI is generally stronger at scalable information processing, pattern recognition, drafting, and standardized execution. People remain particularly important for judgment, empathy, leadership, accountability, relationships, and ambiguous problems.

Are AI agents part of the workforce?

They are not employees, but their capacity, permissions, cost, and risks should be included in workforce and operating-model planning.

Should strategic work remain internal?

Capabilities involving differentiation, architecture, customer ownership, risk, data governance, and strategic judgment generally require strong internal ownership.

What work is suitable for contractors?

Temporary, specialized, variable, or project-based work may be suitable, provided security, knowledge, and accountability are managed.

Should the workplace determine the work?

No. The work should determine the workplace.

How many days should employees work in the office?

There is no universal answer. The decision should depend on activity requirements, customers, team needs, learning, security, and physical constraints.

Is hybrid work always better?

No. A poorly designed hybrid model can increase meetings, fragmentation, isolation, and proximity bias.

What should offices be used for?

Potential uses include:

• Collaboration
• mentoring
• relationship building
• customer engagement
• specialized facilities
• team formation

How can distributed organizations maintain culture?

Through leadership behavior, clear norms, shared experiences, mentoring, communication, fairness, and intentional relationship building.

How can proximity bias be reduced?

Organizations can monitor promotion, assignments, ratings, access, and compensation while using transparent outcome-based decisions.

What metrics should replace attendance?

Relevant measures include:

• Customer results
• quality
• cycle time
• collaboration
• learning
• employee experience
• innovation

Who should own the future-of-work strategy?

It should have executive ownership and participation from business, HR, technology, finance, operations, real estate, and risk.

How should an organization begin?

Start with one important outcome, map the work, identify the workforce needed, define workplace requirements, pilot the model, and measure the complete system.

Conclusion

The future of work is not one decision. It is not an office policy. It is not an AI deployment. It is not an outsourcing strategy.

It is the combined design of:

• What the organization needs to accomplish
• Who or what will perform the work
• Where and how the work will happen

Deloitte’s framework remains powerful because it forces leaders to examine these dimensions together. Work changes when technology takes over routine activities or creates new possibilities. The workforce changes when organizations access capability through employees, providers, contractors, agents, and robots. The workplace changes when digital tools reduce the need for constant physical proximity while increasing the importance of intentional connection, learning, and knowledge-sharing. Organizations that redesign only one dimension will struggle. AI introduced into old workflows creates limited value. Alternative talent used without governance creates dependency and fragmentation. Hybrid work applied without management redesign creates meetings, isolation, and mistrust. The stronger approach begins with purpose. What outcome should improve? What work creates it? Which combination of human judgment, technology, external expertise, and automation is appropriate?

Which environment will support that combination? How will people learn, connect, progress, and remain accountable? The future of work is not predetermined. Organizations can use technology mainly to remove labor and reduce real-estate costs. They can also use it to restore complete ownership, eliminate frustrating work, expand access to capability, improve flexibility, create better customer outcomes, and make human contribution more valuable. The result depends on design.

The defining question is not:

Should work be automated, outsourced, or performed remotely?

It is:

How should we compose work, workforce, and workplace so that the entire system creates greater business value, stronger human capability, and more meaningful outcomes than the system it replaces?

Relevant Articles and Resources

1. Redefining Work, Workforces, and Workplaces - Deloitte Insights

The foundational work-workforce-workplace framework, including the imagine, compose, and activate model and the access, curate, and engage approach to talent ecosystems.

2. 2026 Global Human Capital Trends - Deloitte Insights

Current research on organizational speed, resource orchestration, continuous reinvention, human-machine relationships, and workforce adaptability.

3. Workplace Trends: The Workplace Is No Longer a Specific Place - Deloitte Insights

Research explaining why the work itself should determine the mix of physical, digital, and hybrid environments.

4. The Future of Work - Deloitte Global

A broader collection examining connectivity, robotics, cognitive tools, alternative talent models, and the changing nature of jobs.

5. The Future of Jobs Report 2025: Workforce Strategies - World Economic Forum

Global employer evidence on skill gaps, organizational resistance, talent availability, and planned workforce responses through 2030.

6. The Future of Jobs Report 2025: Skills Outlook - World Economic Forum

Research on changing demand for technological, analytical, creative, leadership, and resilience-related capabilities.

7. US Employment Projections, 2024 - 2034 - Bureau of Labor Statistics

Official projections covering aggregate employment growth and major changes by industry and occupation.