AI adoption, work design
AI adoption, work design
The AI tools are here. Now we need to redesign the work.
The AI tools are here. Now we need to redesign the work.
Teams already use AI, often one task at a time. The next step is to turn these separate uses into a new way of organising work, designed with the people who do it.
Teams already use AI, often one task at a time. The next step is to turn these separate uses into a new way of organising work, designed with the people who do it.
In many businesses, AI adoption is already under way: faster research, data analysis and automation across the lead-to-cash process. Uses are spreading from one department to another, but the gains often remain individual. The next challenge is a shared one: decide which tasks AI should handle, where it should support people, which processes need redesigning and how to use the time released.
Gallup’s research in the first half of 2026 found that only 17% of weekly AI users in organisations that had adopted AI gave it the highest rating for its impact on productivity and efficiency. That figure reached 50% among frequent users who also had a clear plan, active manager support and strong engagement in their work. These are reported outcomes and associations. They suggest that making a tool available and deciding how it changes a team’s work are separate steps. Many businesses are now tackling the second.
Move from individual uses to work design
An individual use may save a few minutes. At team level, different questions arise: how tasks are divided, how work passes between departments, who approves what, which skills need strengthening and how the time released will be used. These require decisions beyond individual tool use.
In an article published in late August by the Harvard Business Review, Faisal Hoque, Tom Davenport and Paul Scade propose analysing tasks rather than whole jobs. A role combines different activities. Some can be assigned to AI; others depend on coordination, relationships, context or judgement. Looking at them separately helps identify what can change while keeping the overall process effective.
AI therefore creates an opportunity to look beyond automation. In many businesses, work has evolved by accumulation: a procedure added here, a tool introduced there, an approval retained out of habit. The whole process has rarely been reconsidered. AI gives leaders a reason to examine each task: should we reduce the burden, support the person doing it, reorganise it or protect its human contribution?
Bring two perspectives together
Leaders see objectives, costs, results and dependencies across the business. Teams know the detail: tasks that take longer than expected, information entered repeatedly across systems, exceptions missing from the procedure and small interventions that protect the quality of the final result.
Leaders see the whole; teams see the detail. A sound roadmap needs both.
This combined view captures more than the most visible tasks in a role. It reveals the supervision, relationships, exceptions and tacit knowledge that keep a process working. Making that work visible is essential before deciding what AI should take on. Involving teams improves the evidence and the decisions built on it.
Gallup’s research in the first half of 2026 found that only 17% of weekly AI users in organisations that had adopted AI gave it the highest rating for its impact on productivity and efficiency. That figure reached 50% among frequent users who also had a clear plan, active manager support and strong engagement in their work. These are reported outcomes and associations. They suggest that making a tool available and deciding how it changes a team’s work are separate steps. Many businesses are now tackling the second.
Move from individual uses to work design
An individual use may save a few minutes. At team level, different questions arise: how tasks are divided, how work passes between departments, who approves what, which skills need strengthening and how the time released will be used. These require decisions beyond individual tool use.
In an article published in late August by the Harvard Business Review, Faisal Hoque, Tom Davenport and Paul Scade propose analysing tasks rather than whole jobs. A role combines different activities. Some can be assigned to AI; others depend on coordination, relationships, context or judgement. Looking at them separately helps identify what can change while keeping the overall process effective.
AI therefore creates an opportunity to look beyond automation. In many businesses, work has evolved by accumulation: a procedure added here, a tool introduced there, an approval retained out of habit. The whole process has rarely been reconsidered. AI gives leaders a reason to examine each task: should we reduce the burden, support the person doing it, reorganise it or protect its human contribution?
Bring two perspectives together
Leaders see objectives, costs, results and dependencies across the business. Teams know the detail: tasks that take longer than expected, information entered repeatedly across systems, exceptions missing from the procedure and small interventions that protect the quality of the final result.
Leaders see the whole; teams see the detail. A sound roadmap needs both.
This combined view captures more than the most visible tasks in a role. It reveals the supervision, relationships, exceptions and tacit knowledge that keep a process working. Making that work visible is essential before deciding what AI should take on. Involving teams improves the evidence and the decisions built on it.
An AI roadmap becomes more precise when it combines two perspectives: leaders see the whole, and teams know how the work gets done.
An AI roadmap becomes more precise when it combines two perspectives: leaders see the whole, and teams know how the work gets done.
Article
Article
An AI roadmap becomes more precise when it combines two perspectives: leaders see the whole, and teams know how the work gets done.
Article
Four decisions for each task
Once the work is visible, examine each task through two questions. How does the team experience it: as a burden or as rewarding work? And where does its value lie: can it be automated, or does it depend on a human contribution worth preserving?
The answers lead to four decisions.
Reduce the burden. Demanding, repetitive tasks that can be automated are early candidates. Addressing them frees up time and provides the programme’s first practical evidence of value.
Augment. For rewarding tasks that can be partly automated, AI acts as an assistant. It prepares, researches or drafts; the expert remains responsible for the outcome and its quality.
Reorganise. Some difficult tasks cannot be assigned to AI. Their difficulty often comes from roles, rules, handovers or timing. The process itself needs to change.
Protect. Rewarding activities whose value depends on people are central to the business. They embody what customers value and the work people want to do. Time released elsewhere can be reinvested in them.
The roadmap becomes a set of explicit decisions about the future of work, with an owner for each workstream.
Decide how to use the time released
Automating a task frees up time. That time creates value when the business decides how to use it: deepening customer relationships, improving quality, bringing an offer to market sooner, training teams or giving complex decisions more attention.
The HBR article frames this as a business strategy that integrates AI. The gains may reduce some resource needs while strengthening teams where human involvement creates more value.
The strategic question is what comes after automation: what do we want to achieve more of, do better or do differently with the time recovered? This is the idea behind Spentia’s approach: automate what weighs teams down and protect what makes the business distinctive.
Design with teams and make responsibilities clear
Participation works when decision-making responsibilities are clear. Teams describe the work and propose workstreams. Managers coordinate implementation, support learning and share the effects with other teams. Leaders set the ambition, allocate resources and decide how to reinvest the time released.
Managers connect these levels. Gallup found that 48% of employees who reported active manager support for AI use were engaged, compared with 30% among those without such support. A manager’s role is to translate broad direction into practical rules: where to use AI, where to avoid it, how to assess results and when people should take over.
Teams contribute their knowledge of the work. Leaders make decisions and provide the means to act. The roadmap is developed collectively, with clear accountability.
Build a roadmap that can change
AI capabilities will continue to evolve, and the roadmap should evolve with them. HBR proposes treating reversibility as a design principle: move in stages, measure results and retain the ability to change a decision as the technology or evidence changes.
A monthly review makes this practical. A new tool prompts a fresh look at the relevant tasks. A workstream that misses its objectives is reviewed. Every time saving should have an identifiable use. The map helps organise change over time.
A lasting advantage comes from the ability to redesign work regularly with the people who understand it, protect what makes the business distinctive and reinvest gains in more valuable activities. The tools are here. The work around them now needs to be designed with those who do it.
Sources:
Jim Harter, “Employee Engagement Remains Flat as AI Adoption Accelerates”, Gallup, 21 July 2026: Gallup research
Faisal Hoque, Tom Davenport and Paul Scade, “AI Transformation Requires Redesigning Work, Not Cutting Roles”, Harvard Business Review, 28 August 2026: HBR article
Spentia, “AI adoption programme”: Our approach to AI adoption
Four decisions for each task
Once the work is visible, examine each task through two questions. How does the team experience it: as a burden or as rewarding work? And where does its value lie: can it be automated, or does it depend on a human contribution worth preserving?
The answers lead to four decisions.
Reduce the burden. Demanding, repetitive tasks that can be automated are early candidates. Addressing them frees up time and provides the programme’s first practical evidence of value.
Augment. For rewarding tasks that can be partly automated, AI acts as an assistant. It prepares, researches or drafts; the expert remains responsible for the outcome and its quality.
Reorganise. Some difficult tasks cannot be assigned to AI. Their difficulty often comes from roles, rules, handovers or timing. The process itself needs to change.
Protect. Rewarding activities whose value depends on people are central to the business. They embody what customers value and the work people want to do. Time released elsewhere can be reinvested in them.
The roadmap becomes a set of explicit decisions about the future of work, with an owner for each workstream.
Decide how to use the time released
Automating a task frees up time. That time creates value when the business decides how to use it: deepening customer relationships, improving quality, bringing an offer to market sooner, training teams or giving complex decisions more attention.
The HBR article frames this as a business strategy that integrates AI. The gains may reduce some resource needs while strengthening teams where human involvement creates more value.
The strategic question is what comes after automation: what do we want to achieve more of, do better or do differently with the time recovered? This is the idea behind Spentia’s approach: automate what weighs teams down and protect what makes the business distinctive.
Design with teams and make responsibilities clear
Participation works when decision-making responsibilities are clear. Teams describe the work and propose workstreams. Managers coordinate implementation, support learning and share the effects with other teams. Leaders set the ambition, allocate resources and decide how to reinvest the time released.
Managers connect these levels. Gallup found that 48% of employees who reported active manager support for AI use were engaged, compared with 30% among those without such support. A manager’s role is to translate broad direction into practical rules: where to use AI, where to avoid it, how to assess results and when people should take over.
Teams contribute their knowledge of the work. Leaders make decisions and provide the means to act. The roadmap is developed collectively, with clear accountability.
Build a roadmap that can change
AI capabilities will continue to evolve, and the roadmap should evolve with them. HBR proposes treating reversibility as a design principle: move in stages, measure results and retain the ability to change a decision as the technology or evidence changes.
A monthly review makes this practical. A new tool prompts a fresh look at the relevant tasks. A workstream that misses its objectives is reviewed. Every time saving should have an identifiable use. The map helps organise change over time.
A lasting advantage comes from the ability to redesign work regularly with the people who understand it, protect what makes the business distinctive and reinvest gains in more valuable activities. The tools are here. The work around them now needs to be designed with those who do it.
Sources:
Jim Harter, “Employee Engagement Remains Flat as AI Adoption Accelerates”, Gallup, 21 July 2026: Gallup research
Faisal Hoque, Tom Davenport and Paul Scade, “AI Transformation Requires Redesigning Work, Not Cutting Roles”, Harvard Business Review, 28 August 2026: HBR article
Spentia, “AI adoption programme”: Our approach to AI adoption
