Recovering delivery and building a scalable transformation model

At the London Borough of Southwark, I led the recovery of a major website transformation while establishing a wider digital transformation programme designed to work across the hundreds of different services provided by a large London borough.

The immediate priority was getting a significantly delayed website rebuild back under control and delivered. Beyond that, the challenge was to create an approach that could continue beyond the website: standardising customer case management, redesigning problematic high-volume services and beginning to explore how AI could remove some of the administrative work associated with customer contact.

The challenge

Southwark is a large and complex organisation delivering hundreds of services, ranging from relatively simple enquiries and transactions through to long-running cases involving multiple teams and systems.

That scale creates a particular transformation challenge.

Redesigning every service individually is neither affordable nor practical. But simply introducing new technology without changing the underlying service can reproduce existing complexity on a newer platform.

At the same time, a major rebuild of the council website was already underway and had fallen significantly behind plan.

The first task was therefore recovery: establish what was actually required to complete the website, reset delivery around achievable priorities and create sufficient structure and focus to get the programme moving again.

Recovering the website programme

The council was undertaking a complete rebuild of its website using LocalGov Drupal, replacing the existing platform with a modern, accessible digital front door.

When I joined, the programme was slipping badly.

I worked with the team to reset the programme, clarify priorities and dependencies, strengthen delivery management and concentrate effort on what was required to successfully launch the new site.

The recovery was not simply about producing a revised project plan. It required establishing a clearer route through a large body of content, technology, design and organisational dependencies while preventing additional requirements from continually disrupting delivery.

This provided the foundation for completing the LocalGov Drupal implementation while also creating a much clearer distinction between what was required for website delivery and what belonged in the council’s longer-term transformation roadmap.

Moving from a website project to digital transformation

The website was an important intervention, but it was only one part of the resident experience.

Behind the digital front door were hundreds of services operating through different processes, systems, forms, inboxes and ways of managing customer interactions.

Rather than attempting to redesign these one by one, I established a wider digital transformation programme based around repeatable service patterns and common capabilities.

The principle was to identify what was genuinely different about a service while standardising the parts that did not need to be different.

This created a route from individual digital projects towards a transformation model capable of operating across the scale of the council.

A standard approach to case management

A major component of that model was establishing Microsoft Dynamics as a common case-management capability.

The challenge was not simply implementing a CRM. With hundreds of potential services, allowing each team to specify its own implementation would quickly have created hundreds of variations of the same basic solution.

Instead, the programme developed a standardised approach to case management.

Common patterns could be used for receiving a request, creating a case, managing its status, allocating work and maintaining a record of the customer interaction. Services could then adopt those patterns while adding specialist workflow or integration where there was a genuine requirement.

This made it possible to think about CRM adoption at organisational scale.

Rather than asking “how do we build Dynamics for this service?”, the starting point became “how much of this service can use the standard model, and what is genuinely different?”

That distinction significantly reduces the amount of bespoke design and configuration required as transformation moves from one service to the next.

Tackling high-volume housing services

Standardisation cannot solve everything.

Housing represented a particularly important area because of both the volume of resident transactions and the complexity of some of its services.

I worked closely with housing colleagues to identify the journeys generating the greatest problems for residents and staff and to understand where service redesign could have the greatest impact.

The approach was deliberately selective.

Rather than attempting to transform the whole housing service at once, we focused on problematic journeys where poor processes, fragmented information or repeated customer contact created disproportionate demand.

These services provided an opportunity to combine service design with the wider programme capabilities: improved digital journeys at the front, common case management behind them and better integration with specialist systems where required.

It also demonstrated an important distinction in the programme model: standardise the common components, but invest proper design effort where the service problem genuinely warrants it.

Exploring AI for customer email

The programme also began exploring how AI could help with another widespread problem across local government: shared email inboxes.

Many services receive significant volumes of unstructured customer email. Before a service can respond, somebody often has to read each message, understand what it relates to and manually categorise or forward it.

We developed a proof of concept for an AI-assisted email triage tool designed to automate that initial handling.

The proof of concept successfully demonstrated that incoming emails could be interpreted, classified and routed based on their content.

This was deliberately a focused use of AI. Rather than attempting to automate the professional judgement or service response itself, the technology targeted a repetitive administrative task that exists across many different council services.

The successful proof of concept established the basis for developing the idea further as a reusable capability.

Designing for scale

Across the programme, the same principle appeared repeatedly.

A large council cannot sustainably transform hundreds of services through hundreds of bespoke projects.

The website rebuild created a common digital platform. The Dynamics approach created common case-management patterns. The housing work demonstrated where deeper service redesign was justified. The email triage proof of concept showed how emerging technology could potentially address repetitive work shared across many services.

Together, these created a more scalable model for transformation:

• common platforms where the requirement is broadly the same;
• standard service patterns wherever unnecessary variation can be removed;
• targeted service redesign for the journeys with the greatest problems or demand; and
• automation and AI where a repeatable task can be safely removed or reduced.

The objective was not standardisation for its own sake. It was to stop repeatedly solving the same problems and preserve scarce transformation capacity for the services where the problems really were different.

Outcomes

The programme combined immediate recovery with longer-term transformation.

The LocalGov Drupal website programme was brought back under control and delivered against a clearer set of priorities, while a broader transformation approach was established beyond the website.

A standardised Microsoft Dynamics case-management model provided a foundation for progressively transforming services without creating a separate CRM implementation for each one.

Work with housing brought that approach into one of the council’s largest and most transaction-heavy service areas, focusing design effort on journeys causing significant problems for residents and staff.

And the successful email triage proof of concept demonstrated an early practical application of AI that could potentially be reused across multiple council services.

The important shift was from thinking about transformation as a sequence of individual technology projects towards creating a repeatable model for changing services at council scale.