When conversations begin around digital transformation, executives usually describe challenges in terms of technology. They want to automate manual work, introduce artificial intelligence, modernise legacy systems or improve employee productivity. Delivery teams begin discussing workflow platforms, low-code development, enterprise integrations and increasingly sophisticated AI capabilities. Before long, the transformation programme becomes centred around selecting technology that promises to solve those challenges.

The technology almost always delivers what it promises.

The organisation often doesn't.

This isn't because the implementation failed or because the platform lacked capability. More often, it is because the service being automated had already accumulated years of organisational complexity before anyone attempted to improve it. Technology faithfully reproduces that complexity at digital speed, but it cannot determine whether the complexity should have existed in the first place.

That distinction may seem academic until you begin examining how enterprise services actually evolve.

Very few organisations consciously design every service from the ground up. Most services are the product of gradual adaptation. A new regulatory requirement introduces another approval step. An acquisition brings a second process that nobody quite manages to consolidate. Security policies evolve, organisational structures change, responsibilities move between departments and temporary workarounds quietly become permanent operating procedures because replacing them never becomes anyone's highest priority.

Viewed individually, these decisions are entirely rational.

Viewed collectively, they tell a different story.

The service gradually stops reflecting how the organisation wants to operate and starts reflecting everything the organisation has experienced over the previous decade.

Employees rarely notice this transformation because they adapt alongside it. Someone who has worked inside the business for several years instinctively knows which approvals matter and which exist largely out of habit. They understand that one department always prefers email despite the existence of a service portal, that another team expects requests to arrive in a specific order, and that a particular manager invariably approves certain requests regardless of circumstance. This understanding becomes part of everyday work without ever being documented.

New employees do not possess that experience.

Customers certainly do not.

Artificial intelligence cannot infer it.

This is where many automation programmes begin encountering difficulties that technology alone cannot resolve. Modern workflow platforms are exceptionally good at executing processes consistently, but consistency is only valuable when the process itself deserves to be repeated. Automating unnecessary complexity does not eliminate it. It simply ensures that every employee experiences it in exactly the same way.

Some of the most successful transformation programmes we have encountered therefore begin somewhere completely different from where many organisations expect. Before discussing workflows, AI agents or enterprise platforms, they spend time understanding the service itself. Not how it currently operates, but how it should operate if the organisation had the opportunity to design it again from the beginning.

That conversation is often surprisingly revealing.

Processes that appear essential frequently turn out to exist because of constraints that disappeared years ago. Approval chains continue simply because nobody has questioned them since they were introduced. Entire sections of a workflow remain in place because they once compensated for limitations in technology that no longer exists. Once these assumptions become visible, automation changes from being an exercise in digitisation into an opportunity for simplification.

This difference explains why organisations implementing the same technology platform often achieve dramatically different outcomes. Two enterprises may deploy ServiceNow using similar methodologies, comparable budgets and equally experienced implementation teams. One emerges with services that employees genuinely enjoy using, while the other succeeds only in replacing paper forms with digital ones. The platform is identical. The architecture may even be similar. The difference lies almost entirely in the work completed before configuration ever begins.

At its best, service design is not concerned with workflows at all.

It is concerned with experiences.

Employees are not interested in understanding how many departments participate in onboarding them. Customers do not want to learn which applications exchange information before a request reaches completion. They simply expect services to feel intuitive. They expect work to move naturally, decisions to happen without unnecessary delay and interactions with the organisation to require as little effort as possible.

Ironically, creating that simplicity often requires confronting enormous complexity behind the scenes. Departments must agree on ownership. Governance needs to become consistent. Enterprise knowledge must be reliable. Technology platforms need clearly defined responsibilities. None of this is visible to the employee receiving the service, nor should it be. Exceptional service design succeeds precisely because it hides organisational complexity instead of exposing it.

Artificial intelligence is making this discipline even more important. Much of the current conversation surrounding AI focuses on what intelligent systems can do, yet considerably less attention is given to the environment in which they are expected to operate. AI can classify requests, generate documentation, recommend actions and orchestrate sophisticated workflows, but it cannot determine whether those workflows represent the best way of delivering value. It assumes that the enterprise has already answered that question.

Increasingly, we believe that assumption will separate organisations that merely adopt AI from those that genuinely transform through it.

Technology will continue advancing at an extraordinary pace. Workflow platforms will become more intelligent, enterprise search will become more contextual and autonomous agents will participate in increasingly complex operational activities. None of these developments reduce the importance of service design. They amplify it. Every improvement in automation increases the value of a well-designed service, while simultaneously exposing the weaknesses of one that has evolved without clear direction.

This is why we increasingly view service design not as a preliminary phase of implementation, but as one of the defining strategic capabilities of the intelligent enterprise. Organisations that invest time in understanding how value should flow before introducing technology consistently place themselves in a stronger position to adopt whatever innovations emerge next. Rather than repeatedly modernising outdated processes, they build services that are capable of evolving continuously alongside the business itself.