TRANSFORMATION AT SCALE AND SPEED - PART 4

Part 4

Build the skills to make the new model work

An organisation can redesign its workflows, introduce powerful technology and give people new responsibilities before it has helped them develop the judgement those responsibilities require.

An employee who previously prepared an analysis is now expected to assess one generated by AI. A customer service adviser handles fewer straightforward enquiries and more difficult exceptions. A manager becomes responsible for work carried out by a combination of employees and automated systems.

The job title may remain unchanged. The demands of the work have moved significantly.

For senior leaders, this creates a consequential question: have we built the capability required to operate the model we have designed?

In Part 3, I argued that we should redesign the work before redrawing the organisation. The next step is to prepare people to perform that work reliably, with sufficient confidence and appropriate authority.

This is the Activate stage of perpetual transformation. It is where decisions about strategy, structure and technology become changes in everyday performance.

Its success depends on treating workforce development as part of the business investment.

Begin with the work people will need to perform

A general ambition to “upskill the workforce” provides limited guidance for an employee trying to understand what will be expected of them next month.

The starting point should be the changed work.

Which tasks will reduce or disappear? Which will become more important? What new decisions will people make? What must they understand well enough to challenge? Where will they need support from colleagues with deeper expertise?

These questions make the capability requirement specific.

Consider the customer-onboarding example used throughout this series. AI now extracts information, checks for missing fields and prepares a summary. Employees investigate inconsistencies, assess unusual circumstances and communicate decisions.

An effective learning plan would go beyond showing people how to use the interface.

It would help them recognise when a summary omits something material, distinguish a routine discrepancy from a significant concern and explain why a case needs further investigation. It would also establish when they can act independently and when they must escalate.

The required capability is therefore a combination of subject knowledge, technology understanding and practical judgement.

Leaders need to define what competent performance looks like before deciding how to develop it.

Prepare people for changes within their existing roles

Workforce transformation is often discussed through the lens of jobs created and jobs displaced. Senior leaders also need to examine how much work can change within a continuing role.

McKinsey Global Institute’s September 2026 report on the US workforce projects that more than 70% of workers could experience substantial changes in their tasks, using a threshold of new tasks accounting for more than 15% of working time. These are modelled estimates for the US economy, influenced by assumptions about automation and other structural trends. They are not a forecast for an individual organisation. Their relevance is the scale of potential change within occupations, including those where employment remains stable.

An employee may remain in the same team while the basis of good performance changes around them.

Someone valued for producing a high volume of accurate work may increasingly be expected to identify exceptions, question recommendations and improve the process itself. Their experience remains valuable, but they need a clear bridge between what they know and what the redesigned role requires.

That bridge should be visible in work assignments, learning opportunities and performance expectations.

Changing the expectations first and leaving people to discover the capability gap through mistakes is an avoidable leadership failure.

Make judgement a deliberate development priority

“Use your judgement” can sound reassuring. It is also an incomplete instruction.

Judgement depends on understanding the subject, recognising relevant patterns, weighing evidence and appreciating the consequences of being wrong. It develops through experience, feedback and opportunities to examine decisions.

If automation removes tasks through which employees previously gained that experience, leaders must consider how the learning will happen.

Bain’s 2026 operating-model analysis highlights this issue: when routine execution is automated, organisations need to redesign how employees build capability and judgement. Its discussion includes the implications for early-career roles, where people may encounter exception handling and validation earlier in their development.

For the onboarding team, I would build learning around realistic cases.

Some should contain straightforward errors. Others should include conflicting evidence, incomplete information or a plausible recommendation that cannot be adequately supported.

Ask employees to explain what they would do and why. Compare their reasoning with that of experienced colleagues. Discuss what additional evidence would change the decision.

This creates an opportunity to examine the thinking behind an answer.

It also helps identify whether a problem comes from missing knowledge, an unclear policy or a weakness in the workflow. Those require different interventions.

Protect the route through which future expertise develops

There is a longer-term question for the executive team.

If we reduce entry-level work, how will people gain the experience needed to become our future specialists and leaders?

Some routine tasks provide limited developmental value. Others help people learn the underlying logic of a service, recognise unusual cases and understand how an organisation makes decisions.

Leaders should distinguish between them.

The answer may involve supervised casework, rotations, simulations or structured reviews of completed decisions. It may require experienced employees to make their reasoning more visible.

For example, a junior colleague could examine a case before seeing the AI recommendation, then compare the two assessments with a supervisor. The purpose would be to develop independent reasoning alongside effective use of the tool.

These are design choices, and they require time.

A workforce plan that recognises immediate efficiency gains but overlooks the development of future expertise is incomplete. The executive team should ask how today’s changes affect the organisation’s ability to staff demanding roles several years from now.

Fund the time required to become capable

Learning has an opportunity cost.

Employees need time to practise. Managers need time to coach. Experienced specialists need time to review work and provide feedback. During a transition, output may temporarily decline.

Those costs belong in the transformation plan.

BCG’s February 2026 workforce-transformation article reports that the most AI-mature companies in its analysis were four times more likely than laggards to have structured AI-learning programmes and protected learning time. This is an association, rather than proof that a particular training arrangement produces better financial results. It nevertheless supports examining whether workforce ambitions are matched by organisational resources.

A commitment to learning becomes credible when leaders decide what work will be reduced to make room for it.

That might mean retiring a report, narrowing a competing initiative, adjusting service expectations during a controlled transition or providing temporary cover.

The specific choice will depend on the business. The obligation to make a choice remains.

If employees are expected to learn while meeting unchanged targets with unchanged capacity, the organisation has created a conflict. Encouragement alone will not resolve it.

This is where structural empathy becomes practical: leaders recognise the demands their plans create and allocate resources accordingly.

Give managers the capability to develop others

Managers carry much of the responsibility for translating organisational change into daily expectations.

They explain what good performance looks like, allocate work, provide feedback and decide when someone is ready for greater responsibility.

They may also be learning the new model themselves.

A manager can be highly effective in the existing operation while needing support to assess AI-assisted work, recognise limitations in a system or coach employees through uncertainty.

Their development needs should be explicit.

Microsoft’s 2026 Work Trend Index, based in part on a survey of 20,000 AI-using workers across ten countries, found that organisational conditions such as culture, manager support and talent practices were more strongly associated with reported AI impact than individual factors alone. The report distinguishes these statistical associations from causal effects.

The practical leadership question is whether managers have the means to create those conditions.

Do they understand the redesigned workflow? Can they recognise acceptable performance? Do they have access to specialists when a case exceeds their expertise? Are they rewarded for developing people, or only for short-term output?

An organisation that relies on coaching should make coaching an explicit part of a manager’s workload and assess whether it happens well.

Create credible pathways into changing work

A skills programme has greater practical value when employees can see where it leads.

People need to understand which capabilities the organisation will require, how they can develop them and how demonstrated proficiency will influence work assignments or progression.

For some, the pathway will involve greater responsibility within their existing role. For others, it may involve movement into a different team or occupation.

Those transitions need more than a catalogue of courses.

An employee may need supervised experience before qualifying for an internal opportunity. Their manager may need cover before releasing them. A receiving team may need support to take someone who has relevant transferable skills but less direct experience.

Unless leaders address these conditions, internal mobility can remain an aspiration.

I would start with a small number of priority roles or capabilities and make the pathways concrete. Identify the transferable experience already available, the gaps that matter and the practical opportunities through which people can demonstrate readiness.

Keep the approach understandable. An elaborate skills inventory provides limited value if it does not improve an actual staffing or development decision.

Be honest about the implications for people

Employees will assess a transformation through its effect on their work, prospects and security.

Leaders should explain which decisions have been made, what remains uncertain and how people will be involved as the model develops.

Where responsibilities change, say so. Where roles may reduce or disappear, address that possibility directly. Where internal transitions are feasible, explain the support and the requirements.

Avoid promising that development will guarantee a particular outcome when the organisation cannot make that commitment.

At the same time, give people a meaningful opportunity to prepare. Access to learning should account for different schedules, locations and starting points. Employees with less discretionary time should not be disadvantaged by a programme that depends on learning outside working hours.

Trust also requires clarity about assessment. People should know what evidence will be used to judge readiness and have a route to discuss or challenge that judgement.

Fairness needs to be designed into the transition.

Measure whether people can perform the changed work

Training attendance and course completion are useful administrative measures. They provide limited evidence of operational readiness.

For the redesigned onboarding process, I would want to know whether employees can identify material errors, handle exceptions, explain decisions and escalate appropriately.

I would also examine how long it takes them to become proficient, how much support they require and whether performance holds as conditions change.

Use several forms of evidence: practical assessment, observed work, quality reviews and feedback from customers and colleagues.

Interpret the results carefully. An initial increase in escalations may indicate that employees are recognising problems they previously missed. A reduction in handling time may conceal additional correction work elsewhere.

The aim is to understand capability in context.

Assessment should inform support as well as authorisation. If someone is not yet ready, the organisation should be able to explain the gap and provide a practical route towards closing it.

Make capability part of the decision to scale

Before expanding a redesigned workflow, the executive team should ask whether the required capability exists at the scale proposed.

A pilot may depend on enthusiastic volunteers and unusually experienced specialists. A wider rollout must work with a broader range of experience, workloads and support needs.

I would require every significant workflow redesign to have a funded capability plan, a business owner accountable for readiness and clear evidence that people can perform the changed work.

HR and learning specialists should help build that plan. Business leaders must remain responsible for whether it is sufficient to deliver the outcome.

This connects workforce development to investment discipline. If the organisation is not ready, leaders may need to narrow the deployment, strengthen support or change the design.

Perpetual transformation depends on people repeatedly becoming capable of work that is new to them. The organisation must make that development possible while continuing to serve customers and maintain standards.

A new operating model becomes real when people can use it well, recognise when it is failing and help improve it.

Where is your organisation asking people to exercise judgement it has not yet helped them develop—and what will you change to close that gap?

Next in the series: Sustain speed through structural empathy.

Research underpinning this article includes McKinsey Global Institute’s September 2026 report Workforce in Motion; Bain’s May 2026 operating-model analysis; BCG’s February 2026 workforce-transformation article; and Microsoft’s May 2026 Work Trend Index. The onboarding example is illustrative. The proposed leadership and learning practices are a synthesis informed by these sources.