SAP's product roadmap reads like a sci-fi novel right now — BTP, Joule, Business AI, Gen AI, Next Gen Procurement, CPI... the list never ends. Honestly even consultants are having an existential crisis trying to keep up.

But then you walk into a client meeting and ask "so, are you thinking about adopting any of this?"

And every single time:

> "What's the ROI?"

> "Our current system is working fine."

> "Let's revisit this next year."

Next year comes. Same answer. 😂

The honest ground reality: SAP is racing toward the future, but a huge chunk of their customer base doesn't even know half these products exist. And the ones who DO know? They're nodding politely in roadmap sessions while mentally calculating how long they can stay on ECC.

Curious what others are seeing out there —

**How aware are your clients actually about new SAP tech? Are they genuinely interested or just waiting for you to stop talking so they can go back to their tickets?**

Drop your most painfully accurate client response in the comments 👇

Does anyone know where is a good area to find partners that can build various integrations (i.e. SAP Ariba, Oracle Fusion) to a platform that serves many users?

I’ve been learning SAP CPI recently, and today I spent some time understanding graphical (message) mapping. Sharing a few takeaways in case it helps someone else who’s also starting out.

From what I understood, CPI gives us three main mapping approaches:

• Groovy scripting
• XSLT mapping
• Graphical mapping (built-in tool)

Graphical mapping seems to be the most commonly used one for day-to-day scenarios, especially when dealing with source → target structure transformations.

A few things that stood out to me:

• You always need a mapping sheet (basically source fields → target fields + logic)
• Along with that:
  • Input XSD (source structure)
  • Output XSD (target structure)
  • Sample input XML
  • Expected output XML

Without these, it’s pretty hard to even start mapping properly.

Also, CPI graphical mapping comes with a lot of built-in functions like:

• Text functions
• Boolean functions
• Date functions
• Arithmetic functions
• Node functions

And if those aren’t enough, we can use user-defined functions (Groovy) — but those work more at element level rather than full structure mapping.

One thing I found interesting:

The concept of occurrences in XSD (minOccurs / maxOccurs)

• If not defined → default is 1 to 1 (mandatory)
• If it’s mandatory in target and not mapped → mapping won’t even save
• Optional fields (0..1 or 0..unbounded) are more flexible

My current understanding:

Graphical mapping is basically:

>

Sounds simple, but when multiple fields, conditions, and functions come in, it gets complex quickly.

Question for experienced folks:

• In real projects, how often do you use graphical mapping vs Groovy?
• And do you strictly follow mapping sheets, or is it more flexible in practice?

Would love to hear how this is handled in real-time scenarios.