From Oracle Implementation to AI Adoption: Why Programmes Are Falling Short

As sponsors at Oracle AI World, we had the opportunity to engage with leaders, practitioners, and programme teams who are shaping what AI‑enabled transformation looks like in the Oracle ecosystem. Across dozens of conversations at our stand, four themes consistently emerged:

• Organisations have a huge desire to adopt SaaS solutions like Oracle, but are not prepared for the realities of transformation programmes, their complexity and the journey required to truly adopt life in the cloud.
• These struggles have manifested in clients live for some years who require support to get more from their solutions as organisations move from process automation to agentic driven automation with Oracle Agentic Apps.
• Adoption of new ways of working driven from these systems is not always strong, with organisations struggling to enable their people using traditional methods.
• There is a huge need for enhanced controls within AI powered environments and in the current regulatory landscape, with a need to move to predictive, not reactive controls. 

What the data told us: 

Speaking with clients and practitioners at the event, we drew some interesting insights – 75% of respondents felt that their controls position as an organisation was strong,  despite auditors and regulators repeatedly highlighting gaps across the market.

This could be a reflection of a time lag between Oracle teams and their auditors and regulators, with organisations in more heavily regulated spaces like financial services feeling the pinch of a lack of controls earlier than their counterparts in ‘softer’ industries like retail and consumer goods.

This trend aligns with Grant Thornton’s recent work, where more regulated firms are engaging us earlier in their Oracle lifecycle to mature their controls environment.

We also saw some interesting disparities in programme readiness alongside adoption readiness. Whilst nine in ten respondents indicated they felt ready for their Oracle programme, only a third felt that their adoption and training approach was ready to meet users.

This data is reflective of our experience as a practice in working with clients at different stage of the transformation lifecycle. Whilst technical teams, (particularly in more mature organisations using tier one solutions like Oracle) often feel ready for programme delivery – this tends to come from an outdated mentality on ‘lifting and shifting’ business processes.

The reality is that users require more support to transform, and we often see a lack of value extracted from Oracle programmes where the focus has leant too much on the technical delivery and not enough on the transformation delivery. When surveyed in a previous insight for Grant Thornton, just 12% of respondents felt that their ERP programme had been a success after the fact. These statistics are instructive of the reality of programmes like these. 

Why These Gaps Matter Now: 

Taken together, these findings reveal a widening gap between what organisations think they are ready for and what is actually required to deliver sustainable, AI‑enabled transformation. Whilst the industry has been consumed with ‘adopt not adapt’ messaging in these deliveries for some years now, the reality is that the shift required to truly ‘adopt’ AI Agents as part of your business is bigger than it was before.

User adoption can no longer be treated as secondary. Supporting users in understanding where Agents can enable them, and how work evolves in a more automated landscape will be key to driving the value that solutions like Oracle bring in reporting and analysis, which for now will still be driven by your human teams.

The controls space is a great example of this, as with increasing regulation (in the form of the recent UK FRC guidance and otherwise) driving a need for organisations to respond. In an Oracle world, this is further necessitated as organisations adopt agents within their finance environments, and drive transactions in new ways with much less human involvement in core processing.

With this in mind, the conversations at Oracle AI World naturally centred on four core areas where organisations most need clarity and practical support. 

How do we drive user adoption and make implementation really sing from a training and change perspective?  

Driving user adoption requires treating change and training as strategic enablers, not downstream activities. Successful Oracle programmes focus on enabling people to work differently - building confidence, ownership and trust through early engagement, role‑based design and hands‑on involvement. Leading organisations are moving away from one‑off system training towards blended, scenario‑led learning, supported by change champions and continuous learning models that evolve alongside AI capabilities. By measuring adoption, confidence and value realisation beyond go‑live, organisations ensure Oracle becomes embedded in everyday work and delivers sustained, people‑led transformation. 

How do we ensure our programmes are correctly resourced and scoped for effective delivery? 

Programme planning and premobilisation is all about understanding your scope, resourcing and skills, and the pace at which your organisation can absorb change. Adoption of standard process is a great goal to have on your programme, but it always comes with limitations. Understanding the areas where your business will naturally diverge from standard, and the resources you need from partners to support internal teams to make these changes will be key in ensuring your programme moves at a pace right for your organisation. 

Where have we not got it right on Oracle and how can we improve and extend Oracle to do more for our business? 

Oracle Cloud Apps have evolved substantially since their launch back in 2011, and many of our clients have changed too. Changes in how we work today require change to how Oracle supports that work, and at Grant Thornton we’ve developed our ‘Measure Success’ methodology to support organisations in optimising and evolving their Oracle estate to meet current business needs, and prepare you for the AI driven future. 

How do we implement strong controls and think about Oracle in the context of our wider operating model and business need? 

Regulators are continuously evolving their approach to controls management and reporting, with the UK FRC making some key changes in the last 18 months. For Oracle clients, Grant Thornton have developed an accelerated controls solution with a library of 450 pre-built controls with supporting dashboards to support our clients in developing their controls stance, ready for the adoption of AI Agents and automation within and beyond their Oracle estates.

If you’d like to explore any of these themes further, please reach out to our Oracle team. We’d be happy to guide you through the next steps:

• David Reynolds
• Zoe Bratton (Change)
• Neal Dempsey (Public Sector)
• Alex Miller (Tax)
• Jim Jeffries (GRC)