Replatform decisions split because each function is optimizing a different risk: IT fears outages and skill gaps, HR fears attrition and change fatigue, Finance fears forecast variance and unclear return. To get buy-in, write one business case that answers each fear in its own language, then govern the programme with a shared RACI and a standing FinOps council.
How do replatform decisions usually break down between IT, HR and Finance?
They break down because the three groups are rarely asked the same question. IT is asked "can we do this safely?", Finance is asked "what does it cost and when does it pay back?", and HR is often not asked until the plan is already approved. Each then reasonably objects to the part they were not consulted on.
| Stakeholder | What they say | What they actually worry about | What answers it |
|---|---|---|---|
| IT / engineering | "It is too risky" | Outages during cutover, skill gaps, security and compliance exposure, a long tail of legacy dependencies | A wave plan with rollback criteria, a dependency map, a pilot, a named skills plan |
| HR | "The team is not ready" | Reskilling load, role changes, attrition of people with in-demand cloud skills, morale during a long programme | Training budget and timeline, role mapping, retention plan, honest communication about which roles change |
| Finance | "The numbers do not add up" | Capex-to-opex shift, forecast variance, double-running costs, a TCO that quietly excludes people and egress | A TCO with transition costs included, scenario ranges, unit cost targets, governance on commitments |
| Business owners | "Will it disrupt my quarter?" | Freeze windows, feature delays | A schedule built around their calendar |
What does IT really worry about?
Risk and skills. The technical risk is usually manageable; the unmanaged part is the long tail of undocumented dependencies and the fact that the team that built the legacy system is often the same team asked to migrate it while keeping it running. Ask IT to define the rollback criteria themselves; people support plans they have authored the safety net for. For the mechanics of sequencing, see our guide on how to migrate an enterprise workload to the cloud.
What does HR really worry about?
Skills transition and retention. Cloud-skilled engineers are mobile in the Indian market, and a visible reskilling programme can either retain people or make them more employable elsewhere. Neither outcome is certain, so plan for both: budget for training and certifications, define career paths for the new roles, and decide in advance how you will backfill if people leave mid-programme. HR also owns the honest conversation about which roles (for example, data-centre operations) shrink and what the redeployment options are. Hiding that conversation tends to cost more than having it.
What does Finance really worry about?
Three things. First, the capex-to-opex shift: spend moves from depreciated assets to monthly bills, which changes how it appears in budgets, approvals, and variance reports. Second, forecast variance: cloud bills respond to usage, so a fixed annual number is a poor model, and a range with named drivers is more honest. Third, TCO completeness: a credible comparison includes migration labour, parallel running, training, egress, licensing changes, and ongoing operations, not just compute versus servers.
How do I justify the spend and get buy-in?
Write a one-page business case with this structure. If it does not fit on a page, it is not yet clear enough to decide on.
- Decision requested. One sentence: what you want approved, for how much, by when.
- Why now. The forcing function: hardware refresh, end of support, licence renewal, capacity limits, regulatory change.
- Options considered. Including "do nothing" and a partial option. Show why you rejected them.
- Total cost of ownership. Three-to-five-year view of current state vs target state, with transition costs, people costs, and egress listed separately rather than buried.
- Benefits, split by confidence. Hard savings (contracts that end), soft savings (productivity), and strategic value (speed, resilience). Finance will discount the last two; label them yourself.
- Range, not a point. Low, base, high with the two or three assumptions that drive the spread.
- Risks and mitigations. Top five, each with an owner.
- People plan. Reskilling, role changes, hiring, retention. This is the section most business cases omit, and the one HR will look for first.
- Governance and checkpoints. Stage gates where spend can be paused, and the unit-cost targets that define success after go-live.
Presenting this jointly, with IT, HR and Finance each sponsoring their own section, is usually more persuasive than any one function presenting alone. For the vendor and SLA side of the numbers, see how to evaluate cloud vendors on SLA and ROI.
Which cloud provider is the safest long-term bet for a large Indian enterprise?
"Safest provider" is the wrong frame. Every major provider is financially strong and unlikely to disappear in ten years. The risks that actually hurt are different:
- Concentration risk. Depending on one provider for critical services means that provider's outage, price change, or policy change is your problem.
- Lock-in through managed services. The deeper you adopt proprietary data and AI services, the more costly leaving becomes. That is sometimes a fair trade, but it should be a decision, not an accident.
- Commercial leverage. A credible alternative changes how a renewal negotiation goes.
- Regulatory fit. Some regulators expect an exit and continuity plan for critical outsourced services.
A better question is: "What is our exit plan, and which workloads would we be willing to be locked into?" Practical answers: standardize on open interfaces where cheap to do so (containers, Kubernetes, standard databases, open data formats), accept proprietary services where the value is high, document an exit path for critical workloads, and test it. A primary provider with a deliberately chosen second one for specific workloads is common. Multi-cloud everywhere is expensive and usually unnecessary. Choose providers on workload fit, local region availability, support in your geography, and commercial terms, and re-check current offerings, since they change.
How do you implement a single governance model for all clouds?
You govern the outcomes, not the consoles. The tooling differs by cloud, but the model can be one.
- Common taxonomy. One set of tags or labels (owner, application, environment, cost centre) enforced by policy on every cloud.
- Common cost data. Normalize billing into one view; the FOCUS standard exists for this purpose.
- Common policy. One written standard for identity, network, encryption, and data classification, implemented per provider.
- Common decision rights. A RACI that applies regardless of provider.
- A standing council. A FinOps council that meets monthly.
A starting RACI:
| Decision | Engineering | Finance | HR | Security/Risk | Business owner | Council |
|---|---|---|---|---|---|---|
| Workload placement | R | C | I | C | A | I |
| Commitments and discounts | C | R | I | I | C | A |
| Budgets and unit-cost targets | C | R | I | I | A | C |
| Reskilling and role changes | C | I | R/A | I | C | I |
| Policy exceptions | R | I | I | A | C | I |
| Exit plan and concentration review | R | C | I | A | C | C |
(R = responsible, A = accountable, C = consulted, I = informed. Adjust to your org; the point is that every decision has exactly one accountable owner.)
The FinOps council should include a senior sponsor from Finance and Engineering, a delegate from HR for the people plan, and security. Its monthly agenda: unit-cost trend, forecast versus actual, commitment coverage, policy exceptions, and the exit-plan review. Our cloud cost optimization page outlines how this connects to day-to-day cost work.
What India-specific factors affect a replatform decision?
- Data residency. Confirm which workloads must stay in India and check that the services you plan to use are available in Indian regions. Sector rules differ.
- DPDP Act. India's Digital Personal Data Protection Act, 2023 shapes how personal data is processed, stored, and transferred. Implementation details have been phased through rules; have your legal team confirm current obligations rather than relying on a blog summary.
- BFSI. Banks, NBFCs, insurers, and brokers face RBI and SEBI requirements on outsourcing, data localisation, audit access, and business continuity, which directly affect provider choice and exit planning. Treat those requirements as design inputs from week one, and confirm current circulars with your compliance team.
- Procurement and billing. INR billing, GST handling, and local contracting entities affect the TCO and the forecast.
- Talent. Hiring for cloud skills in metros is competitive, so the HR plan carries real cost.
When does this approach not apply?
If you are a small team with one application and no formal approval layers, a one-page business case and a council are more process than you need; a short note and a budget alert will do. If the replatform is mandated by regulation or a hard end-of-support date, the question becomes execution rather than justification. And none of this is legal or regulatory advice; confirm compliance points with your own counsel.
Where Nuvika fits
Nuvika Technologies helps Indian enterprises with FinOps-led IT cost optimization and cloud migration, including a three-week assessment that produces a quantified baseline and roadmap you can put in front of Finance, priced in INR for India clients. Our cloud migration page has the details. We are not the right fit if you want a provider-selection or legal-compliance opinion, a full-scale change-management programme led by HR specialists, or a vendor-neutral architecture review of regulated core banking systems; other specialist firms handle those better.
FAQ
How do replatform decisions usually break down between IT, HR and Finance?
IT worries about risk and skills, HR about reskilling and attrition, and Finance about capex-to-opex shift, forecast variance, and TCO completeness. The breakdown usually happens because one group is consulted late or the plan is written in only one group's language.
How do I justify replatform spend and get buy-in?
Write a one-page business case covering the decision requested, options, a full TCO with transition costs, benefits labelled by confidence, a low/base/high range, risks, a people plan, and stage gates. Have IT, HR and Finance each sponsor their own section.
Which cloud provider is the safest long-term bet for a large Indian enterprise?
No provider is meaningfully "safe" in the sense of disappearing; the real risks are concentration, lock-in, and regulatory fit. Choose on workload fit, regional availability, and commercial terms, and keep a documented, tested exit plan for critical workloads.
How do I implement a single governance model for all clouds?
Standardize tags, cost data, policy, and decision rights across providers, assign one accountable owner per decision in a RACI, and run a monthly FinOps council. Implement the controls per provider but report them in one view.
What should BFSI companies in India consider before moving to cloud?
RBI and SEBI expectations on outsourcing, data localisation, audit access, and business continuity should shape provider choice and exit planning from the start. Confirm current circulars with your compliance team, as requirements are updated over time.