AI Strategy, Business Strategy and the Need for System Agility

As AI adoption accelerates, strategy has become essential. Organisations are developing the foundations, governance models, workflows, skills, systems and operating structures needed to apply AI safely and at scale.

But over time, AI should not remain a separate strategy. The business strategy remains the strategy. AI will become one of the key tactics, mechanisms and capabilities used to deliver it.

AI is increasing the pace of change. Organisations that want to respond effectively will need more than isolated use cases, productivity tools or automation projects. They will need core systems that can adapt as markets, operating models, customer expectations and competitive pressures change.

System agility is therefore becoming a strategic requirement.

In the early days of enterprise AI adoption, many AI strategies seemed to repeat the old robotic process automation playbook: get your data right, get your processes right, then use automation to reduce headcount and increase efficiency. AI was positioned as a more powerful automation layer, capable of streamlining work, removing manual tasks and reducing operational cost.

Another version of early AI strategy was to suggest redesigning the organisational hierarchy around AI. Which roles should be performed by AI? Which roles should remain human? Which roles should be hybrid?

Both questions are relevant, but they are not sufficient. Organisations are still learning what AI can reliably do, where it creates value, how people should work with it, what governance is required, and what risks need to be managed.

And yet, the AI conversation evolves quickly.

A year ago, there was a lot of focus on prompt engineering. The belief was that better prompts would lead to better responses and better outcomes. Prompt engineering was seen as a core capability. Today, it is still useful, but it is no longer the central issue. AI systems are becoming better at understanding intent, remembering prior interactions, working within context and adapting to the type of output a user is trying to create.

The focus is now shifting from prompts to context, memory, workflow, governance, data access, permissions, orchestration and business integration. These are the areas where the best AI platforms, AI harnesses and AI-enabled tools are starting to differentiate themselves.

For those developing strategy, the difficulty is: where does this go next?

The honest answer is that nobody can say with certainty. The only thing we can say with confidence is that the next few years will bring accelerated change across multiple fronts, influencing operating models, customer engagement, data, decision-making, service delivery and the economics of work.

This high volume of change creates a different requirement for AI strategy.

The goal should not simply be to identify where AI can reduce cost today. The goal should be to build an organisation that can remain relevant, adaptive and competitive as AI continues to evolve and disrupt established ways of working.

This is where the relationship between business strategy and AI needs to be reframed.

I remember watching a Mark Ritson (marketing professor) lecture where he made the point that digital is not a strategy in itself. The strategy is the marketing strategy. Digital is one of the tactics used to implement that strategy.

I think a similar logic will soon apply to AI.

AI will not sit beside business strategy as a separate strategy in its own right. The business strategy remains the strategy. AI becomes one of the most important tactics available to implement that strategy.

If an organisation wants to enter a new market, improve customer experience, reduce operating cost, increase service quality, improve decision-making or adapt its operating model, AI should be considered as a key mechanism for achieving those outcomes.

Many outcomes that were once difficult to achieve may now become more realistic. Optimising supply chains, personalising customer engagement, improving data quality, enhancing decision support or increasing service consistency may all become more achievable because AI can reduce cost, improve productivity and lift the quality of work.

But those opportunities depend on the organisation’s ability to change.

The pace of AI development means we can expect disruption across established businesses and industries. New AI-enabled business models, new customer expectations, new operating economics and new service delivery models will continue to emerge.

In that environment, organisational agility has never been more important.

Business agility has been discussed for many years. Companies such as Kodak were often used as examples of what can happen when organisations fail to adapt to changing environments. Many organisations responded by introducing agile teams, agile delivery methods and agile ways of working.

But team agility is not the same as organisational agility.

An organisation can have agile teams and still be constrained by systems that are slow, rigid and expensive to change. In many cases, the real barrier to agility is not the people. It is the systems landscape.

A friend once shared the challenges he encountered while working in sales and marketing for a telecommunications company. The company operated in an intensely competitive market, with smaller rivals regularly introducing new offers that forced the business to respond.

His team could develop a new campaign, promotion or counter-offer within days. The challenge came when those ideas had to be implemented through the organisation's systems.

IT typically needed six to eight weeks to develop, test and deploy the required changes. By the time the new offer was ready, the market had already moved on. Meanwhile, smaller competitors continued to chip away at the company's market share.

The organisation had commercial agility, but not system agility. Its people could see the opportunity, but its systems could not move fast enough to support the response.

That is a critical lesson for businesses operating in an AI-enabled world, where new features, capabilities and competitive threats increase the likelihood of disruption.

If AI increases the pace of change, then organisations need systems that can adapt.

This does not mean every system in the organisation needs to be equally adaptable. Some systems are stable by design and may not require frequent change. The strategic priority is to identify the core systems that enable the organisation to operate, compete, make decisions and respond to customers, markets, regulators and internal change.

Where those core systems are rigid, slow to update, poorly integrated or dependent on manual workarounds, they become constraints on organisational agility. In an AI-enabled environment, the potential pace of change increases the risk these constraints pose. Organisations therefore need to understand which systems are strategically important, where flexibility is required, and where current limitations could prevent the business from responding quickly and effectively to new opportunities or risks.

The issue is not whether one part of the organisation has flexible technology. The issue is whether the organisation as a whole has the system agility required to respond to change.

The organisations that succeed with AI will not simply be those that automate the most tasks. They will be the organisations that build the capacity to respond quickly, safely and intelligently to change.

This is also why a narrow focus on process automation can create risk.

Automation can create value, but if it is designed around fixed processes, rigid workflows and today’s operating assumptions, it may reduce flexibility rather than improve it. In a dynamic environment, the organisation may need to change quickly as markets, customer expectations, regulations, technologies and business models evolve.

If automated processes are not supported by an adaptable framework, they can become difficult to modify, expensive to maintain and misaligned with the needs of the business. In that scenario, automation risks creating a new class of stranded assets: AI-enabled processes that were valuable when designed but become constraints when rapid change is required.

The objective should therefore not be automation for its own sake. The objective should be adaptable automation: AI-enabled capabilities that improve efficiency today while preserving the organisation’s ability to change tomorrow.

AI strategy should therefore start with a much broader question:

How do we ensure the long-term viability of the organisation in an age of profound disruption?

That question leads to a very different strategy from one focused mainly on headcount reduction.

It also requires organisations to ask a practical question:

What can we improve today that strengthens our ability to respond to change tomorrow?

The source of disruption is difficult to predict.

Every competitor will apply AI differently. Some will focus on lowering costs, others on personalising customer experiences, while others will use it to accelerate product development, strengthen supply chains, or improve decision-making across the business.

The possibilities are too broad for a narrow AI strategy.

This is why organisations need to think in two related dimensions.

• The first is agility: how do we create an organisation that can respond to change?

• The second is optimisation: how do we use AI to move the organisation closer to its ideal operating state?

AI should not be limited to automating tasks inside existing processes. It should also help organisations question whether those processes, structures and systems are still fit for purpose.

Historically, many improvement programmes were limited by cost. AI changes that cost profile, creating strategic opportunities to address issues and pursue opportunities that were previously too costly to deliver.

If your data is not fit for purpose, AI can help improve data quality, identify issues, support remediation and strengthen the data pipeline.

If your ERP is too rigid, the organisation should not simply accept that constraint. It should ask whether it should keep core transactions in the ERP while building more flexible auxiliary capabilities around it, modernise parts of the architecture, or consider whether the existing platform remains fit for purpose.

If your planning systems cannot adapt to a dynamic environment, then change them.

If your reporting is too slow, backward-looking or disconnected from decision-making, reimagine it.

AI strategy is not about chasing every AI opportunity. It is about creating a business strategy that recognises AI as one of the key mechanisms for improving performance, increasing adaptability and delivering outcomes that may previously have been out of reach.

Business strategy now needs to include an AI lens. Leaders should be asking:

1. What business outcomes are we trying to achieve, and how could AI change what is possible?

2. How might AI affect our market, our competitors, our customers and our operating model?

3. Where could AI materially improve speed, quality, cost, scale, service delivery or decision-making?

4. Which outcomes were previously too costly, too slow, too complex or too difficult, but may now be achievable with AI?

5. What are the implications for our customer proposition, pricing model, service delivery model and competitive position?

6. Which core systems, processes, data structures, approval paths or manual dependencies are too rigid to support future change?

7. What can we improve today that strengthens our ability to respond to change tomorrow?

AI strategy should then focus on how the organisation will use AI to support and execute that business strategy. Leaders should be asking:

1. What are our highest-priority AI opportunities, and how do they link to our business strategy?

2. Which AI use cases can create value today, and which capabilities create optionality for tomorrow?

3. What foundational capabilities do we need to strengthen, including data quality, integration, security, permissions, governance, workflow, architecture and skills?

4. What governance, controls and accountability do we need so AI can be used safely, responsibly and effectively?

5. How do we evaluate how humans and AI work together across key roles, processes and decisions?

6. How will we measure value, learn from implementation and adjust as AI capabilities evolve?

7. How do we ensure AI supports long-term organisational agility?

Organisational agility has never been more important. Unfortunately, many organisations do not fully understand how rigid their systems are, where the constraints exist, or why those constraints prevent change. In an AI-enabled environment, identifying and addressing these constraints can no longer be treated as a future improvement opportunity. It is now a strategic imperative.

AI can reduce the delivery cost of many technology and business improvement initiatives. It can make capabilities that were previously too expensive, too complex or too slow more viable. It can help organisations improve the foundations they need to compete in a more dynamic environment.

AI strategy is not about predicting exactly what AI will do in five years. It is about building an organisation that is ready for whatever comes next.