① Journal · Strategy
Most estates that commission a migration business case stay on the mainframe. The study still earns its keep, just not the way the kickoff deck promised. Here is why exit cases stall, and how a credible one moves the renewal number even when the decision is to stay.
The migration rarely happens. The leverage always could.
It is one of the most reliable patterns in enterprise IT. A new executive arrives, the mainframe bill looks large and unfamiliar, and an exit study is commissioned with real momentum behind it. A year later the study is complete, thorough, and quietly shelved. The estate stays exactly where it was. This happens often enough that the more interesting question is not why the migration failed but what the study was actually for.
The honest answer is that a rigorous exit study is one of the most valuable artifacts a buyer can own, and almost never because it leads to an exit. It leads to a renewal negotiated from strength. The mistake is treating the study as a binary migrate or stay decision and judging it a failure when the answer is stay. Judged as leverage, the same study is a success the moment it becomes credible. Read this with our keep vs exit comparison and our renewal advisory service.
What an honest model surfaces · the costs the kickoff deck left out
| Factor | What it adds | Effect on the case |
|---|---|---|
| Undocumented logic | Decades of business rules live only in code nobody fully owns | Rewrite scope and risk balloon once it is mapped honestly |
| Data gravity | Large, integrated datasets tie dozens of systems to the platform | Migration touches far more than the application in scope |
| Parallel running | Both platforms run for years during a phased cutover | You pay twice while the savings have not started |
| Integration sprawl | Feeds, batch chains, and interfaces multiply the cutover surface | Testing and reconciliation dominate the timeline |
| Skills and risk | Rare knowledge, and a platform that simply does not fail | The risk adjusted case rarely beats staying put |
| Payback horizon | Net savings often arrive only after several years of spend | The horizon exceeds what a board will fund on faith |
Every estate is different and some exits are right; the pattern is that an honest model routinely pushes payback past the funding bar. The study stalls not because the team was wrong but because the numbers told the truth.
A vendor prices against your options. A study is how you build one.
Software pricing rests on an assumption: that the buyer has nowhere to go. Mainframe renewals escalate precisely because the vendor believes the estate is captive, the data is too heavy to move, and the threat to leave is a bluff. A costed exit study is the single artifact that tests that assumption. It does not have to result in a migration to do its job. It only has to be credible enough that the vendor can no longer assume it is empty.
Credible means specific: a scoped target architecture, a realistic timeline, named internal sponsors, and a defensible cost that the vendor's pricing now has to compete with. That is a different conversation from a procurement lead saying the board is unhappy. It is the difference between a verbal threat and a quantified alternative, and it is what turns a renewal from a rubber stamp into a negotiation. The studies that never ship still pay for themselves many times over when they are built to be used. Where they are built only to answer a yes or no question, they are filed and forgotten, and the leverage is left on the table. See how we build the credible walk away in our approach.
An inflated savings case collapses the moment the vendor probes it, and a collapsed case is worse than none. Build the model honest enough to survive scrutiny, then keep it current. A study the other side cannot puncture is the one that holds at the table.
Credibility is the whole asset.
The cloud is not a plan. A usable exit names the target architecture, the migration approach, and the products it displaces, product by product. Specificity is what makes the alternative priceable, and a priceable alternative is what disciplines the renewal.
Vendors price specifics, not slogans.
An exit nobody senior will fund is not a threat. The study carries weight when a named executive is willing to be seen backing it. The vendor reads the org chart as carefully as the cost model, so make the sponsorship as real as the numbers.
A threat needs an owner to be credible.
The study is worth most in the window before a renewal closes, while the vendor still has something to win or lose. Commission it early enough to be complete and tested before the negotiation, not delivered as the contract is already being signed.
Leverage has an expiry date.
⑤ The reframe that pays
An exit study judged only on whether you leave looks like a failure. Judged as leverage, it paid for itself the day it became credible. Build it to be used, not to decide.
Typical reduction negotiated on renewal spend
Mainframe spend negotiated on the buyer side
Engagements delivered since 2019
Because an honest model usually pushes cost, risk, and payback past what a board will fund. Undocumented logic, data gravity, integration sprawl, and years of parallel running tend to outweigh the headline savings. The estate stays not because the team failed but because the numbers were truthful.
No, if it was built to be credible. A rigorous costed exit case is the strongest single input to a renewal even when the decision is to stay. The only wasted study is one too thin to use at the table.
It removes the vendor's assumption that you are captive. A scoped, costed, sponsored alternative forces the pricing to compete with the cost of leaving rather than simply escalate. A quantified walk away disciplines a renewal in a way a verbal threat cannot.
Early enough to be complete and tested before the renewal closes, while the vendor still has something to win or lose. Leverage has an expiry date. See our renewal advisory service.
Related: keep vs exit · exiting Broadcom CA products · the pricing power problem · renewal advisory service
Audit notice or renewal under 18 months out? We mobilize within 48 hours.
Get expert help →