A vendor announces a price change, or a platform reaches end of support, or an integrator arrives with a proposal. The question lands on someone’s desk: do we modernize, or do we renegotiate and wait?
It is rarely answered on the merits, and the reason is structural rather than anyone’s fault.
Everyone advising you is paid by the answer
The vendor proposing the migration earns from the migration. The integrator confirming it is necessary earns from the implementation. The internal team may be entirely right, and may also be defending an architecture decision they made two years ago and would rather not reopen.
None of these people is behaving badly. They are simply not positioned to give you a disinterested answer, and treating their agreement as corroboration is the most common way this decision goes wrong. Three parties who all benefit from yes is not three opinions, it is one.
What is missing is someone senior enough to read the technical proposal properly, with nothing riding on the conclusion.
The four inputs that actually settle it
1. The cost of the exit, produced by someone who would have to do it. Not a vendor estimate and not an internal guess. Moving the workloads is the visible part and rarely the expensive one. The expensive parts are requalifying the backup chain, replacing the drivers inside the guest systems, and rewriting every operating procedure your team knows. A number that omits those three is not a number.
2. The cost of staying, extended honestly. The new price, times the renewal periods you can foresee, plus whatever the contract permits the vendor to add. And the risk that becomes yours if the platform reaches end of support: an unpatched system is a cost you have not paid yet.
3. What the current setup is actually used for. This is the input that most often changes the answer. The upheaval in the virtualization market produced one genuinely useful side effect: it forced organizations to ask what they actually needed. A meaningful number discovered they had been paying for years for high-availability features nobody used. Start with an inventory of real consumption, not of licences held.
4. Whether you can operate the alternative. The decisive criterion, and it is organizational rather than technical. A cheaper platform that nobody on your team can run is not cheaper. A poorly monitored system with untested backups is more dangerous than an expensive one that is properly held.
Why the costed migration is worth having either way
Here is the part that surprises people: produce the number even if you intend to stay.
A renewal negotiated without a costed alternative is not a negotiation. The vendor knows whether you have an option, and prices accordingly. Walking into that conversation with a real figure, produced by someone who would have to execute it, changes what you can credibly say.
Several of our assessments have ended with the client staying on their existing platform and paying materially less. The deliverable that made that possible was a migration plan they never executed. It was not wasted work; it was the only leverage available.
The three answers, and what each looks like
Renegotiate and stay. Right when the environment is deeply tooled around the current stack, the increase is absorbable, and the platform has years of support left. The work is the costed alternative plus a hard look at what you actually consume.
Modernize now. Right when support is ending, when the price trajectory is unsustainable over three years, or when the current platform blocks something the business needs. The work is a sequenced migration with a rollback window per wave, and backups requalified as a milestone rather than a formality.
Do neither yet, and fix the foundation. The answer we give more often than either of the others. If the backups have never been restored, if nobody can list what the company actually runs, or if the monitoring shows green while something is quietly broken, then moving the estate adds risk without resolving anything. Restore test first, then monitoring that sees what matters, then modernization. Inverting that order is the most common way to spend money while increasing exposure.
What a good assessment produces
Not a slide deck. Four things, and they should fit in a document you can hand to a board.
The real inventory, including what nobody remembered was running. The two costed options, with the assumptions written down so they can be challenged. The recommendation, with the reasoning that would let someone disagree with it intelligently. And the sequence, if the answer is to move, with the rollback plan per wave.
The document should be yours whether or not you continue with whoever produced it. That is the test of whether it was an assessment or an avant-vente: ours are priced and delivered as a standalone deliverable for exactly this reason. If the conclusion is renegotiate, we are paid the same.
Further reading
- Technical due diligence: the assessment format, and what it produces
- Legacy modernization: the engagement, if the answer is to move
- Leaving VMware: the real alternatives: the same decision on a concrete case
- Your backups run. Do they restore?: the foundation to fix first
- Fractional CTO vs interim vs consultant: the three ways to buy the outside read
- What a fractional CTO costs: pricing the leadership half of the decision
- Adapting an existing platform instead of building new: a real engagement where the calendar decided