The Silent Erosion
SAP RE-FX Between Generational Change and System Responsibility — A Structural Shift in the DACH Region.

The Situation

SAP RE-FX is running. It has always run. The system posts, the closing is done, the auditors have signed off. For years, RE-FX wasn't an issue —and that was precisely the proof that it works.

At some point, that changed.

Not dramatically. Not with a bang. Slowly: a ticket here, an unresolved issue there, a topic that resurfaces in the next round. Then a second. A third release change without a smooth landing. At some point, RE-FX is on the agenda—not as a project, but as a problem that won't go away.

The pattern isn't the same for every user type. It depends on who is operating the system and for what purpose.

In real estate organisations and public property management agencies, change requests pile up. Tickets accumulate. In between, problems arise that no one can clearly categorise: sudden errors and warnings during input tax allocations in utility billing, FI postings from RE-FX that raise questions during month-end closing, configurations that worked for years and are now inexplicable. Sometimes this becomes apparent shortly after a release change —but technically, it cannot be attributed to that.

For corporates that implemented SAP RE-FX as part of their IFRS 16 adoption, the picture is different—but the result is similar. Six to fourteen months after go-live, something is off. The basic functions are running. But something isn't quite right, and RE-FX comes up regularly in meetings. Not as a success. As an open issue. As a need for explanation that won't go away.

What connects both situations: the decision-maker sees the symptoms. The cause is not visible.

What Has Changed

SAP RE-FX is not a new system. Classic Real Estate was introduced in the mid- 1990s; RE-FX followed starting in 2004. Anyone who was there back then —as an in-house consultant, as a key user, as someone who built the system from the ground up—is now between fifty-five and sixty-five years old.

In the DACH region, this means: this generation is leaving the companies. Not all at once, but at a density that is no coincidence.

Austria experienced this peak earlier. The actual retirement age there is 59.6 years¹—in Germany, employees leave the labour market on average only at 64.4 years of age.² About four years later. The cohort that left in Austria between 2018 and 2023 is now leaving in Germany. This explains why the issue has already been addressed in Austria— and is just reaching its peak in Germany.

Switzerland does not follow this pattern in the same way. The average length of service there is 8.21 years, compared to 10.57 years in Germany.³ Between 2022 and 2023, 14.7 percent of all employed people changed jobs; the average job change occurs after 5.6 years.⁴ Contrary to common perception, job changes in Switzerland are structurally more frequent than in Germany or Austria. What makes the problem less visible there is the culture of documentation: agreements must be set forth in writing, and process changes require justification. Switzerland is therefore considered part of the broader European market, not the DACH-specific generational transition pattern.

What is lost when these people leave is not their job. It is their judgement. Thirty years of decisions—which condition type, which posting logic, why this exception, why not that one—are not documented. They were never codified as knowledge. They were simply there, embodied in a person who knew why.

This is not an isolated phenomenon. In German-speaking countries, 76 percent of companies report massive knowledge loss during staff turnover; 60 percent count knowledge loss among their greatest business risks.⁵ With SAP RE-FX, this risk is structurally higher than in other modules—because the knowledge isn't in the system, but in the decisions that led to its configuration.

The system keeps running. The person is gone. The knowledge, too.

Thirty years of system knowledge should not vanish with the retirement of a single individual. But that is exactly what is happening—across the entire DACH region, at the same time, for the same reasons.

What Comes In When Experience Goes Out

The departure was foreseeable. A successor has been appointed; training sessions have been scheduled. At some point, someone asks how things are going.

The answer: good.

That is the moment when most organisations stopped looking.

The answer isn't a lie. It is structurally worthless. The JobTeaser Career Barometer 2023 is titled: "A Fine Line: Gen Z Between Self-Confidence and Overconfidence." 80 percent of young professionals in the DACH region look optimistically toward their professional future—even though 72 percent do not have a clear career path in mind and 64 percent are simultaneously worried.⁶ 48 percent of HR managers confirm: Gen Z systematically overestimates its own strengths and weaknesses.⁷ The outward appearance is self-confidence. What lies beneath is overconfidence—and the individual in question cannot recognise the difference because they lack the benchmark.

This is not a question of character. The in-house consultant role is, in most organisations, a solo role. No colleague, no internal corrective, no counterpart to point out the gap from the outside. Anyone who takes over a mature RE-FX system after three training sessions sees: it's working. What they don't see are the decisions behind the configuration— because they don't know what to look for. They report what they see. And what they see is: no problem.

The pattern manifests differently—depending on the type of organisation involved.

RE organisations and public real estate management agencies

The expectation: two to three SAP training sessions, around 7,000 euros, and the successor is ready to go. The reality: after that, the overwhelming majority of operational RE-FX tasks cannot be performed independently —not because the successor is unsuitable, but because the module has a depth that even 10 official intensive training days could not fully cover. Independent accounting work in RE-FX is—at best—possible after two years at the earliest—an expert opinion, under favourable conditions, with accompanying external support.⁸ Only 27 percent of DACH companies rate their SAP transformation in terms of user acceptance as good or very good⁹—for a niche module like RE-FX, this figure is the ceiling, not the floor.

The successor reports: everything is running smoothly. He is convinced of it. And that is precisely the problem.

Corporates

Here, the starting point is different. RE-FX was usually introduced later and as part of the IFRS 16 implementation. The person in charge knows IFRS 16— project completed, auditor has signed off, balance sheet is correct. What they don't know is the rest of the module. IFRS 16 by no means covers the full functionality in RE-FX—estimated at only about 20%.¹⁰ Service charge billing, space management, RE Navigator, inventory management, and still a considerable portion of the CLM— all outside the scope of IFRS 16. When tasks arise beyond lease accounting, people improvise. The cognitive trap exacerbates this: the more familiar a domain seems, the greater the overconfidence in one's relative assessment.¹¹ Anyone familiar with IFRS 16 considers RE-FX to be understood. The fine line between self-confidence and overconfidence is crossed here without anyone noticing (see above).

In both cases, the same applies: the system is running. The organisation asks for feedback. The answer is positive. And the gap grows.

The Compensation Trap

What many organisations observe in a vacuum of expertise and know-how— even if only temporarily—is this: the experienced key users— employees with long tenure who have been using the system for years—take the lead. A new in-house consultant, unsure and lacking orientation in an unfamiliar system, aligns themselves with them. In effect, the key users are in charge. That feels like stability.

It is a trap.

Whether a new condition type integrates smoothly into the system, whether account determination promotes or hinders the Clean Core, or whether a customising decision can still be reversed in three years—these are not questions for key users. These are questions for the in-house consultant. Key users know the processes. They do not know the system core. If the consultant nods instead of deciding, people without knowledge of system architecture make decisions with system-architectural impact—and cannot assess the consequences of their own instructions.

The result is two simultaneously active sources of risk. The in-house consultant does not report any problems to upper management and does not resolve customising issues independently. The key users make de facto decisions without grasping the systemic consequences. If, in addition, experienced key users leave the organisation, this fragile buffer disappears—a separate risk operating in parallel. However, this offers the chance that the vacuum will finally be communicated upward.

A real-life case:* A standard problem—old and new contracts—for which RE-FX has a standard solution. Instead, research is conducted. In some RE-FX version in some country, there is a mechanism that seemingly solves the problem. It is examined, deemed acceptable—by people with no real knowledge of the system architecture—and implemented.

The contract ID is a key field in RE-FX—a great deal is referenced to this single ID. Whoever duplicates it does not duplicate contracts. They split the system. If, in the end, every contract bears two or three identical contract numbers, it is no longer repairable —even a greenfield project would reach its limits with that.

We were only notified of this CR after our own project had already been completed. A warning letter—well outside our scope of work. The response came promptly, confidently: We are \_\_\_\_\_\_. If it were critical, someone internally would have warned us. There was no reply to the second email. At the time, we were astonished. Today we would know better. Today, based on our understanding of the mandate, we would write a third and fourth letter.

This is the compensation trap in its purest form: self-confidence after three training sessions, internet research instead of an understanding of architecture, no authority in the room allowed to say no—and external criticism that rolls off the self-assurance.

No one sees the big picture. The system keeps running. And eventually someone asks whether an external agency shouldn't be brought in.

An outsider. Problem solved.

Eventually, the phrase comes up: "Then we'll just bring in someone from outside." The decision-maker believes the problem is solved. That used to be the case. A call to the agency, a profile, a twelve-week assignment—the experienced RE-FX consultant arrived, and the problem was resolved before it could escalate. This model had two prerequisites: there were available, reputable senior RE-FX consultants on the market, and there was a project format that suited them. Changes in both areas make hiring agencies a risk today.

The Module

RE-FX is one of the niche SAP modules that traditionally has a surplus of demand. Training takes three to five years before someone can work independently.¹² Industry-specific process knowledge in the real estate sector is usually a prerequisite. Hourly rates are correspondingly high—which has attracted career changers. The expansion of RE-FX's functionality has also opened the market to FI consultants, often without industry knowledge. SAP does not intervene in this.¹³ Today, the mere mention of "RE-FX" is enough to secure two IFRS 16 projects.¹⁴

Agencies would have the task of addressing this issue and communicating it. But they do not—because a niche module, especially since projects no longer run full-time at 600 hours or more, is no longer lucrative. Anyone who directs RE-FX enquiries to agencies encounters junior contacts. For years. No senior key account manager seems to take on a peripheral topic anymore. RE-FX is delegated —to those who cost the least. These contacts often don't even have basic SAP knowledge. They cannot classify the requirements or make a qualified shortlist. And they are remarkably good at hiding exactly that—during the conversation, nothing stands out, the terminology is correct, confidence is high, the gap remains invisible.¹⁵ At least to us. We assume that this has the same effect on clients.

Conclusion: The Double Uncertainty

A new group of labelled RE-FX consultants meets junior account managers—sometimes even without in-depth SAP knowledge. The new "no longer full-time" model renders needs-based placement —as far as the skills and requirements of senior RE-FX consultants are concerned, who could identify the problem—essentially obsolete.

What is needed today is something else: ongoing governance, one day a week, architectural oversight with a mentoring and handover focus. Someone who accompanies the successor, not replaces them. Someone you can call. Who provides clear advice when a customising decision jeopardises the Clean Core. This format does not exist in the agencies' business model.

The Structural Answer

So the question is not whether an agency wants to solve the problem. The question is whether it can even do so. The answer is structurally: no.

Because the experienced RE-FX consultants are gone. Not vanished—they have often traded freelancing for a permanent position. Contract terms have forced this. Since 2020, the terms have become radically stricter: full liability on the freelancer's part, contractual penalties for non-performance, retention clauses without reciprocal security, payment only after receipt of payment from the end customer.¹⁶ SAP SE itself had already terminated all direct freelance engagements as early as 2014.¹⁷ In March 2024, DHL terminated ongoing freelance contracts in the middle of projects, without prior notice.¹⁸ Any experienced consultant observing this trend is drawing conclusions. Around 36 percent of solo self-employed individuals in Germany are considering relocating, and 27 percent are considering giving up self-employment.¹⁹ The seniors are no longer available—not at any price, but under no conditions.

The contract arrives as a PDF. Sub-sub-sub-level. 10 percent of the text consists of contractual penalties. Payment terms: 45 days. Margin in the chain: 45 percent. Payment is made only if the end client has paid—a circumstance that the consultant cannot even verify in the event of a dispute. In the event of non-payment, there is no negative proof. No bank statement, no confirmation, nothing. You wait—or you sue against a burden of proof you cannot bear.

Or this real-life example: ten hours. Five-month term. On call. After that, the contract automatically renews—for twenty-four months. Ten hours, spread over five months, bind the consultant for over two years. No experienced senior would sign that. The junior would—because he needs the contract. And because he doesn't know what he's signing.

The Consequence

The combination of all three factors—niche module, new format, changed contract terms—makes agencies in this sector obsolete. Agencies are only still interesting as a stepping stone for newcomers. In essence, agencies have already rendered themselves superfluous since COVID, especially in SAP niche markets, due to the shift in risk. They just haven't realised it yet. Their original role—to act as the consultant's advocate even in cases of doubt— is no longer possible at all, especially for small agencies, due to their dependence on the client.

Their business model has always been based on two things: a certain— accepted by all due to a lack of alternatives—black box, and the volume of large placements. Both are eroding. Most agencies have not realised that the new governance model—small, quality-intensive, long-term— will become the standard in the future. Those who fail to recognise this will not survive.²⁰

The Vicious Circle

The external consultant is placed anyway. They look like a good fit. Generative AI reinforces the image—the right terminology, the standard configuration, the official process specifications. What AI cannot yet do in SAP, at least for the next 12–24 months: replace the decisions that have long-term effects. The framework behind the conditions. The why behind the what.

Can the client even verify the qualifications? Real estate organisations and public property management agencies usually recognise the gap—they've typically been familiar with the module long enough. Corporates usually don't—because they themselves only know IFRS 16 and don't know what they don't know. The black box here is not the intention of a single actor. It is a characteristic of the system: the agency market rewards placements, not depth of expertise, and the client has no tool to evaluate the expertise delivered.

In the end, the decision-maker is faced with the same symptoms as at the beginning.

Service contracts with system integrators, based on SLAs, at least offer relative security as a fallback—a different engagement model.

What happens when nothing happens

If you've read this article this far and recognised yourself in more than two paragraphs, then the following lines are written for you. Not for your IT manager. Not for your service provider. For you—the person who signs off in the end.

The symptoms you see today—tickets, open issues, the need for explanations in the monthly closing—are not the problem. They are the surface. Beneath lies a data integrity issue that worsens with every month the system runs without architectural oversight.

The scenario we see most often: a customising decision —made without an understanding of the architecture, without consultation, without an impact assessment—compromises a key field in the database. The system keeps running. But the postings are no longer correct. Not immediately visible, not dramatic—until the auditor asks about the contractual basis of a posting and no one can explain it, or the system stalls technically.

What follows, depending on the case, is not an IT project. It is damage control: suspension of operations or postings, forensic analysis, data cleansing during ongoing operations, two to six months on an unstable system, four to eight external specialists working simultaneously. The costs range—depending on portfolio size—between 450,000 and 850,000 euros. A German study estimates unplanned SAP outages at an average of 25,000 euros per hour. During an S/4HANA migration, 35 to 65 percent of all data records require cleansing—with compromised key fields, this rate is far higher.

The question isn't whether this will happen. The question is whether it will happen at your company before or after you could have prevented it.

Who we don't work for

Anyone who believes that today's agency market is the same as it was in 2017 hasn't read this article—or hasn't understood it.

Anyone convinced that the new generation will figure it out—because they have children of that age themselves, because they believe they can recognise this generation's potential, because they just need to give them time—might as well hire interns as company doctors. Not out of malice. Out of confusing confidence with responsibility.

Anyone who dismisses every reference to structural risks as doom-mongering has made a decision. We respect it. We do not share it.

Disclaimer: This article has been prepared with care. Where translations into languages other than the original German are provided, minor inaccuracies may occur. The views expressed are solely those of the author and reflect personal professional assessment, practical experience, and independent research — unless explicitly supported by cited sources. The author is not affiliated with, employed by, or formally associated with SAP SE. This is not an official SAP publication. All product names mentioned are trademarks of their respective owners.

T.H. Scheer

© Metaneering S.à r.l., Luxembourg. All rights reserved.

Back to Perspectives →