① Journal · Benchmarks
Software AG's Adabas and Natural are now a standalone business carved out under Silver Lake, and the estates that run them rarely have a quick exit. That captivity shapes the bill. When the data and the application logic both live in a stack only one vendor maintains, the benchmark is less about market rate and more about how much pricing power the lock in hands the vendor. Here is how to read Software AG spend honestly.
48 hour mobilization Audit notice or renewal under 18 months out? We mobilize within 48 hours.
Get expert help →Software AG (Software GmbH) carved Adabas and Natural into a standalone mainframe business in early 2025, after selling its webMethods and StreamSets lines to IBM and reorganizing under Silver Lake ownership. For customers, the products did not change, but the strategic posture did. Adabas is the database, Natural is the application language, and together they form a tightly coupled stack where the data model and the business logic are written for each other. Migrating off is a multi year program, which means the renewal arrives with the vendor holding most of the leverage.
That is the heart of the Software AG benchmark. The license is typically metered on mainframe MSU capacity, with the license record capturing the operating system, CPU identity, LPAR, and MSU. But the rate you pay reflects the vendor's read of how stuck you are. Benchmarking honestly means separating the part of the bill that tracks genuine capacity from the part that is a premium on captivity.
The Software AG bill concentrates around capacity and the add ons bolted to it. The benchmark is to test each driver against whether it reflects real need or unchallenged renewal inertia:
| Spend driver | How it works | Where captivity inflates it |
|---|---|---|
| Core Adabas and Natural | MSU capacity based license | Capacity figure never re-baselined as the estate shrinks |
| Add on products (replication, security, tooling) | Priced add ons to the core license | Bolt ons kept on the bill long after the project ended |
| Support and maintenance | Annual percentage of license value | Uplift applied yearly with no cap, compounding on a captive base |
| Renewal uplift | Vendor read of switching cost | The premium that exists purely because exit is slow |
The recurring pattern is the support line rising quietly every year on a base the customer cannot easily walk away from. Because migration is genuinely hard, buyers often concede the uplift rather than fight it. The most effective benchmark response is to re-baseline the capacity to the current estate, strip the add ons that no longer serve a live project, and cap the annual uplift, rather than to pretend a credible exit exists when it does not.
A defensible Software AG benchmark is honest about leverage. It re-baselines the licensed MSU against what the estate actually runs today, since captive estates often shrink in workload while the licensed capacity stays frozen at an old high. It inventories the add ons and retires the ones tied to finished work. And it fights the support uplift on the one ground that holds, a negotiated cap, rather than threatening a migration the vendor knows will not happen on a renewal timeline. External rate bands inform the capacity discussion, but the decisive lever on a captive estate is disciplined housekeeping plus a hard cap on the annual increase.
The general method is in the guide on benchmarking your mainframe software spend, and the estate itself is profiled in Adabas in 2026. The sibling reads are benchmarking IBM spend and benchmarking Rocket Software spend. When the Software AG renewal lands and the uplift is on the table, our Software AG cost optimization work caps the increase and re-baselines the capacity. The estate view sits in the Software AG licensing hub.