SAP's clean core level concept has been around for about a year now. It replaces the old 3-tier model and no longer rates existing Z code in a binary way as either "clean" or "not clean," but instead in four stages from A to D. At the very bottom sits Level D: things that SAP actively advises against: the use of "noAPI" objects, modifications, implicit enhancements, direct table access. In other words, exactly the kind of thing that has accumulated in abundance over twenty years of an organically grown code base.

In a training session, a customer recently asked me: "And how do the others handle this?" Good question. So I launched a survey in the DSAG forum – deliberately aimed only at SAP customer companies, not consulting firms. 65 responses have come in by now. The result is interesting.
The numbers
How does your company deal with Level D code?
| Answer | Votes | Share |
|---|---|---|
| Active reduction: planned refactoring toward higher levels | 9 | 13.8 % |
| Case-by-case: rebuilding only when an object is touched anyway | 21 | 32.3 % |
| Deliberate retention: it runs, it doesn't get touched | 8 | 12.3 % |
| No conscious strategy in place | 20 | 30.8 % |
| What does Level D mean? 🙂 | 7 | 10.8 % |
| We are on S/4HANA Public Cloud – Level D doesn't exist for us | 0 | 0 % |
Five observations
1. Only one in seven is actively cleaning up
13.8 % carry out planned refactoring toward higher levels. This is the group that sees Level D as technical debt and has the time and money to tackle the topic. Everyone else either has good reasons to wait – or no reasons at all, because they've never asked themselves the question.
2. "Case-by-case" is the secret winner
At 32.3 %, the pragmatic middle path is the most common conscious strategy: whoever touches an object anyway raises it to a higher level while doing so. This is Robert C. Martin's Boy Scout Rule from the book "Clean Code." Together with the active cleaners, around 46 % thus see a fundamental need for action. At least in the very long term, the debt is being paid down.
The majority of 54 % doesn't concern itself with the topic.
3. At least 41 % are driving without a map
This is where it gets interesting: 30.8 % have no conscious strategy, and a further 10.8 % aren't even familiar with the level concept. Together that's over 41 % who have neither a plan nor a decision to counter their Level D code. In terms of the outcome at the next upgrade, the difference from "deliberate retention" isn't large.
4. One in ten has never heard of it
One year after the introduction of the level concept, 10.8 % of respondents don't know what Level D means. And that's within the DSAG forum. One may assume: outside the active DSAG forum visitors, the rate looks considerably worse. SAP's communication apparently still has room for improvement.
5. Public Cloud: zero
Not a single response from Public Cloud customers. Of course: the survey ran in an environment dominated by classic legacy systems, and anyone without a Z code base probably doesn't even click on the survey. But the zero is a data point nonetheless. The world in which, by definition, there is no Level D is simply not yet visible at DSAG.
Assessment
The friendly reading: almost half of the companies have a conscious position on Level D – from active reduction to calculated retention.
The other reading: more than 50 % are not tackling the topic. And it's precisely these companies that get hit hardest later – during release upgrades, when moving to ABAP Cloud, or when the 2009 modification suddenly blocks a project. And the Level D code is only the tip of the iceberg. Level C is also questionable.
For companies where everything went according to plan during the S/4HANA migration and the effort was low, there is no need for action – a lack of strategy is fine there too. For everyone else, one can only recommend engaging with the concept and at least tackling the tip of the level iceberg.
The survey is still running – any SAP customer company that hasn't voted yet: Here's the link to the survey. Status of the numbers: July 2026, n = 65.



