Picture this: your procurement team walks out of an IBM negotiation holding what looks like a pre-paid gift card. IBM has promised a "heavily discounted" ELA renewal in exchange for accepting it. Everyone shakes hands. Everyone smiles. What could possibly go wrong?
That gift card is IBM's Aggregate Cap Program, better known as ACP. And on a three-year, $50M ELA renewal, we routinely see ACP buckets carrying $20M of unused credits by the time the term ends. Money your company has already committed to. Money IBM has already booked. Money that quietly evaporates.
What is ACP, really?
ACP is a pool of pre-paid entitlement sitting inside your ELA: a capped bucket you can draw against for eligible IBM products during the term. In theory, it gives you flexibility. In practice, it gives IBM a very effective way to inflate deal size today and reset the baseline higher tomorrow.
Why so much ACP goes unused
After hundreds of IBM engagements, the reasons are almost always the same:
- The renewal wasn't planned, and IBM was pulling the strings. When the seller sets the tempo, the customer accepts the structure. ACP volume gets sized to hit IBM's revenue target, not your consumption forecast.
- Everything was bought upfront. New project licenses that you might use "sometime in the next three years" get loaded into the $50M renewal on day one. Meanwhile the ACP bucket, the money you already paid, sits there waiting. Why buy upfront when the ACP could have absorbed that same purchase later?
- Net-new purchases happen alongside the ACP. We've seen customers cut fresh POs for IBM products while a fully funded ACP bucket sat unused, because nobody read the ELA contract closely enough to realize the new purchase was already covered. IBM will not remind you. It doesn't work in their favor.
- Stakeholders forget they're the customer. The dynamic shifts from "we're buying" to "we're being sold to," and the leverage evaporates before the signature is dry.
The trap underneath the discount
Here's the part IBM's slide deck won't show you. Products purchased under an ELA with ACP are typically priced at list, not discounted. The S&S you renew for the remainder of the term? Also list. So the "heavy discount" you thought you negotiated is really a rebate you pre-paid for, drawn from a bucket you may never fully consume.
And when the next renewal cycle arrives, that $50M ELA has quietly become $60M, because the baseline now includes everything you loaded upfront, all at list price, with an even bigger ACP proposed on top.
So, ACP or no ACP?
ACP isn't inherently bad. Used deliberately, it can be a genuine flexibility mechanism. The real questions are:
- How large should the ACP bucket actually be, given realistic consumption?
- Which purchases should be loaded upfront vs. drawn from ACP over the term?
- How do you make sure every eligible net-new purchase pulls from ACP first?
- What discount structure protects the next renewal from ballooning?
How MooseTech helps
We work as an independent partner alongside your application owners and procurement team. We map current and forecasted IBM usage, benchmark discounts across a wide range of customers, and give you the ammunition to push back on IBM with data instead of adjectives. The goal is simple: keep the next ELA renewal flat or lower, instead of watching it drift upward every three years.
If you have an ACP bucket you're not sure how to use, or an ELA renewal on the horizon and want a second set of eyes before you sign, get in touch.