① Licensing · Contractual vs Consumed MSU
Every estate carries a gap between the MSU it is licensed for and the MSU it actually uses. The gap is pure waste, it is invisible on a budget line, and it only ever drifts wider because contracts ratchet up and almost never down. It is frequently the single largest piece of avoidable spend on an estate, and it closes for exactly one reason: somebody finally measured it.
Licensed, but not used.
Contractual MSU is the capacity you are licensed for and paying against under your agreements. Consumed MSU is the capacity your workload actually uses. When the contractual figure sits above the consumed figure, the difference is capacity you pay for and do not use, and on most estates that gap is real, persistent, and larger than anyone inside assumes.
The gap drifts one way for a structural reason: contracts ratchet up but rarely down. Capacity gets added for growth, a peak, a merger, or a hardware refresh, and then the licensed figure stays high after the driver passes. Few agreements automatically reduce the commitment when consumption falls. So unless the buyer actively measures and resets it, the gap only widens, and it widens silently because no single number on a budget reveals it. This is the same gap the MSU baseline locks in when it is set on an inflated estate.
What the gap is worth.
Lay contracted MSU next to consumed MSU per product and the waste becomes visible for the first time. Take a four product slice of an estate, with an illustrative annual cost per MSU applied to the unused portion to size what the gap is worth.
| Product | Contracted MSU | Consumed MSU | Gap (annualized) |
|---|---|---|---|
| Product A | 600 | 470 | 130 MSU |
| Product B | 420 | 350 | 70 MSU |
| Product C | 300 | 300 | 0 MSU |
| Product D | 250 | 160 | 90 MSU |
| Unused capacity, paid for | 1,570 | 1,280 | 290 MSU |
Illustrative MSU figures. Apply your own contracted per MSU cost to the gap column to size the annual waste for your estate.
Read the bottom row. This four product slice is licensed for 1,570 MSU and uses 1,280, a 290 MSU gap that bills every year through the licensing and support attached to that unused capacity. Product C is matched, which proves the point: the gap is not inevitable, it is just unmeasured. Products A, B, and D drifted because capacity was added and never reclaimed. Across a full portfolio, with every product priced against capacity, the gap compounds into one of the largest avoidable lines on the estate, sitting unbilled in plain sight until the contracted and consumed columns are finally put side by side.
The ratchet only turns one way.
The gap survives because nothing in a standard agreement works to close it, and several things work to widen it.
| Driver | What it does to the gap |
|---|---|
| Capacity added for a peak or merger | Widens it, then stays after the driver passes |
| Hardware refresh with a higher rating | Can widen it even with flat workload |
| Conservative provisioning | Licenses headroom that is never used |
| No automatic true down | Leaves the commitment high when consumption falls |
| No owner of the number | Keeps the gap invisible and unmeasured |
Every driver pushes the same way. Nothing in a default contract pushes back.
The decisive line is the last one. The gap is not primarily a contracting failure or a technical one; it is a measurement failure. No single person owns the comparison between contracted and consumed MSU, so the number that would expose the waste is never produced, and capacity that should have been reclaimed at the last hardware refresh or workload move is still on the bill years later. The fix is not clever. It is producing the number, then acting on it.
How to close it.
The four levers that close the gap
Across 500+ engagements and $180M+ of negotiated mainframe spend, the gap between contracted and consumed MSU is one of the most reliable sources of recovered cost we find, precisely because it hides in the absence of a single number. The capacity is licensed, the workload does not use it, and the bill keeps arriving. Measure the gap, reset the commitment, and the waste stops. For the engagement that does this end to end, see mainframe cost optimization.
Contractual MSU is the capacity you are licensed for and paying against. Consumed MSU is what your workload actually uses. When the contractual figure sits above the consumed figure, the difference is capacity you pay for but do not use, and on most estates that gap is real and persistent.
Because contracts ratchet up but rarely down. Capacity is added for growth, peaks, mergers, or refreshes and then the licensed figure stays high after the driver passes. Few agreements automatically reduce the commitment when consumption falls, so the gap widens over time unless the buyer actively measures and resets it.
The licensing and support attached to every MSU of unused contracted capacity, every year, across every product priced against that capacity. Because it is invisible on a budget line, it is often the largest piece of avoidable spend on an estate, and it compounds quietly because nobody owns the number that would reveal it.
By measuring consumed against contracted MSU per product, then resetting the commitment toward real use through renewal renegotiation, true down provisions, decommissioning credits where products are retired, and baseline resets on consumption contracts. The gap closes only when someone measures it; it does not close on its own.
Audit notice or renewal under 18 months out? We mobilize within 48 hours.
Get expert help →