If Trusted Data Is the New Oil, Your SAP Custom Code Philosophy Matters Now More Than Ever

For years, we’ve heard the phrase: “Data is the new oil.”

It’s a compelling analogy. Trusted Data fuels insight, innovation, and competitive advantage. And for SAP customers running on platforms like SAP S/4HANA, data sits at the very heart of the enterprise. There is no doubt that Trusted Data can unlock AI.

But there’s a critical piece missing from that analogy, one that I see every day working with global organizations. Unmanaged Custom Code is now what’s holding it back.

If Trusted Data is the new oil, then your SAP system is the refinery that is highly fine-tuned. And your custom code is the infrastructure inside it.

The Hidden Complexity Inside Your Refinery

Over the past three decades, companies have invested heavily in customizing their SAP environments. These customizations were often necessary to support unique processes, differentiate the business, or bridge functional gaps. Perhaps an unpopular opinion, but some customizations are still very necessary in 2026 and will be for a long time outside of the SAP core.

But over time, those incremental changes have added up. Across 400 customer landscapes, smartShift has remediated 4B lines of code across 12M custom objects, averaging ~10M lines and ~30K objects per SAP system across SAP ERP modules.

What was intended to be a clean, efficient system has become something far more complex:

• Thousands of custom objects
• Millions upon millions of lines of code
• Interdependent logic embedded deep in core processes
• Limited documentation and institutional knowledge

In other words, a refinery that’s been modified again and again until it’s difficult to upgrade, difficult to maintain, and increasingly difficult to modernize.

Why This Matters Now

We’re entering a new era of enterprise technology – one defined by:

• Continuous innovation cycles
• Cloud-based hybrid architectures
• And, most importantly, AI-driven business processes which hopefully deliver enormous amounts of business benefit

SAP is delivering powerful new capabilities in these areas. But there’s a catch:

They assume a level of standardization that many systems simply don’t have.

I think we have learned over the last 40 years that when unmanaged custom code dominates the landscape:

• Upgrades become slow, expensive, and risky
• New capabilities and innovation are difficult to adopt
• AI struggles to access and interpret fragmented data and processes

So while organizations are sitting on vast reserves of “oil,” they lack the modern refinery needed to turn it into value.

Here are a few examples of where agentic workflows fail:

A custom report is written with a calculated Key Performance Indicator:

• The KPI is not in the data model and the custom logic lives in the custom code
• The KPI is needed to support decisions and other processes
• The KPI is not available and the critical calculation is unknown and buried
• The agent does not have this context and delivers the incorrect outcome

Consider sales pricing, which is one of the most commonly customized areas in SAP:

• The agent has access to all the sales data
• The pricing engine has been heavily customized because the standard model does not fit the customer’s pricing requirements
• The agent is not aware of the custom logic embedded in the custom code
• The agent delivers the incorrect outcome

The Custom Code Barrier to AI

It is not just about the data that is physically stored, but also the data that is buried in coding logic. AI doesn’t just require data, It requires understandable, accessible, and Trusted Data.

In heavily customized SAP environments, that’s rarely the case.

Unmanaged custom code introduces:

• Inconsistent data models
• Non-standard process flows
• Logic that bypasses modern APIs and integration points

The result?

Before AI can deliver value, organizations are forced into a costly and time-consuming preparation phase, cleaning, restructuring, and reinterpreting what already exists.

That’s not innovation. That’s remediation.

A Different Approach to Transformation and Continuous Modernization

At smartShift, we have always believed there’s a better way than a manual approach with limited resources. Automation has been in our DNA since the beginning.

Instead of treating custom code as a manual problem that requires armies of developers and years of effort we approached it over 20 years ago as an automation challenge.

Using advanced automation and AI, we:

• Have been delivering Automated Repeatable Solutions with Guaranteed Outcomes for over 20 years
• Continuously analyze entire custom code landscapes at scale faster than anybody in the world
• Identify what’s necessary and what’s not faster than anybody in the world
• Automatically transform code for modern SAP environments faster than anybody in the world
• Optimize and simplify along the way faster than anybody in the world

See the pattern here. This isn’t just about making code compatible.

It’s about re-architecting the refinery so it operates like a modern, efficient system again. A modern, efficient system similar to my golf swing. Just checking to see if anybody is still reading this.

From Constraint to Catalyst

When organizations address custom code effectively with a highly automated continuous approach, something powerful happens.

Upgrades accelerate.
Risk declines.
Innovation becomes achievable again.

And most importantly:

AI stops being a future aspiration and starts delivering real business outcomes.

The Path Forward

The conversation around SAP transformation has evolved.

It’s no longer just about moving to S/4HANA. That is table stakes. We handle the entire custom code work stream with our eyes closed. Ask our customers. It’s about enabling a platform that can continuously adapt, integrate, and innovate. That requires a shift in mindset:

• From managing custom code manually as part of a 12 to 18 month upgrade cycle to automated continuous custom code transformation and management
• From customization to standardization, if at all possible
• From manual effort to intelligent automation
• From maintaining complexity to eliminating it, if at all possible

Because in the end:

Data may be the new oil but without a modern, streamlined refinery, its value remains untapped.

Final Thought

Every organization I speak with understands the importance of data. SAP certainly does as evidenced by their recent acquisition of Reltio.

Fewer recognize that custom code is the single biggest factor determining whether that data can power the future or hold it back. We must change this. Now.

It is a solvable problem and one that smartShift has been addressing for decades. I have been referring customers, SI partners, and SAP to smartShift since 2012. I used to spend quite a bit of time convincing customers and partners that what we do is not too good to be true. It never gets old. Now is the time to install not only “Sesame Street Simple” guard rails and governance for AI and data but also for custom code. Not doing so is likely a big mistake. The good news is that the word is getting out about smartShift and solving the custom code factor is one of the biggest elements that is key to unlocking everything that comes next!