① Guide · Software asset management
A mature SAM practice can have the distributed and cloud estates fully metered and still be blind to the mainframe, the single most expensive line in the portfolio. The frame breaks standard SAM in six specific ways. This is how to extend your practice to fit it.
Your SAM tool sees the whole estate except the most expensive part.
Software asset management platforms were built for the distributed and cloud world, where licensing is per install, per core, or per user, and agent based discovery sees almost everything. The mainframe does not work that way. It licenses on capacity and usage metrics, commonly MSU under sub-capacity with the charge set by the rolling four hour average, and the authoritative data lives in SCRT and SMF rather than in any discovery agent. The result is a familiar gap: a SAM practice that reports confident compliance across thousands of distributed titles and effectively cannot see the mainframe license position at all.
That gap matters because the mainframe is routinely the largest single software cost in the enterprise and the highest audit exposure. Closing it does not require abandoning your SAM discipline; it requires extending it onto metrics and data sources the standard tooling never modeled. This guide maps the six breaks and the method that fits. Pair it with building your mainframe software inventory.
Distributed SAM assumption vs mainframe reality
| Dimension | Distributed SAM assumes | Mainframe reality |
|---|---|---|
| Licensing metric | Per install, per core, or per user | MSU and MIPS capacity, usage driven |
| What drives the charge | Count of installs or devices | The rolling four hour average peak |
| Source of truth | Agent based discovery | SCRT, SMF, and ISV usage reports |
| Installed vs used | Installed is roughly the entitlement | Installed says little; consumption is the bill |
| Contract vehicle | Simple per seat or per core terms | MLC, Tailored Fit Pricing, ISV consumption models |
| Audit basis | Install counts reconciled to entitlement | Measured peaks reconciled to contracted capacity |
Every row is a place a distributed SAM workflow silently produces the wrong answer on the frame. The fix is to model the usage metric and pull the real data, not to bolt the mainframe onto a per install schema.
MSU and MIPS as capacity, the rolling four hour average as the charge driver, and the contract vehicles above them. These are usage and capacity measures, not install counts, so the whole SAM record has to track consumption against entitlement. See our explainer on sub-capacity vs full capacity.
Consumption against entitlement, not installs.
Authoritative usage lives in SCRT and SMF, with ISV products reporting through their own meters. Get those feeds into the SAM record rather than relying on discovery agents that cannot see capacity consumption. The data source is the difference between a real position and a guess.
SCRT and SMF, not discovery agents.
Entitlement from contracts, installed product data from SMP/E, and measured consumption from the usage feeds. The gaps between them are the whole story: overlicensing is savings, underlicensing is exposure, unmapped running products are the first audit finding.
The gaps are the audit and the savings.
Refresh capacity and consumption monthly, because the rolling four hour average moves with workload, and capture entitlement changes the moment they happen. A live mainframe SAM record is both the cost baseline and the audit defense; a stale one is the auditor's opening.
A live record is the audit defense.
④ The SAM gap that costs the most
The distributed estate is metered to the decimal. The frame, the biggest bill, is unseen. Close that gap and SAM finally covers the whole portfolio.
Typical reduction negotiated on renewal spend
Mainframe spend negotiated on the buyer side
Engagements delivered since 2019
Most SAM platforms discover and meter distributed and cloud estates, where licensing is per install or per core. The mainframe licenses on MSU under sub-capacity with the charge driven by the rolling four hour average, and the real usage data lives in SCRT and SMF, not in discovery agents. The frame is the blind spot in an otherwise mature practice.
MSU and MIPS as capacity, the rolling four hour average that drives sub-capacity charges, and the contract vehicles on top, IBM MLC, Tailored Fit Pricing, and ISV consumption models. These are usage driven, so the record tracks consumption against entitlement, not installs.
Reconcile three independent sources: entitlement from contracts, installed data from SMP/E, and measured consumption from SCRT, SMF, and ISV reports. The gaps are the story. A record built this way is both the cost baseline and the audit defense.
SAM builds the baseline; the renewal is where it pays off. The same reconciled record that defends an audit is the evidence base that wins a negotiation. We commonly partner with SAM teams on both. See our license negotiation service.
Related: building your mainframe software inventory · sub-capacity vs full capacity · audit rights clauses to negotiate · license negotiation service
Audit notice or renewal under 18 months out? We mobilize within 48 hours.
Get expert help →