A defensible TCO framework — five years, four cost layers.

Direct, indirect, exit, risk-adjusted — the layers most TCO models miss.

Conventional TCO analyses capture direct cost — licence, support, infrastructure — competently. The gap is in the other three layers, which compound differently and create the surprises that derail multi-year forecasts. The framework that produces defensible numbers covers all four:

1. Direct cost: licence fees, support fees, infrastructure (cloud or on-premises), and any vendor-billed managed services.
2. Indirect cost: internal labour, integration cost, training, change management, and consumption of cross-functional capacity.
3. Exit cost: the cost to leave the platform at end of term — data migration, system retraining, transition support, and the implied option value of staying.
4. Risk-adjusted cost: the probability-weighted cost of unfavourable events — audit exposure, scope expansion, price uplift overruns, vendor failure.

Why the four-layer model wins

A TCO model that captures only direct cost almost always favours the incumbent vendor in renewal scenarios — the alternative carries higher indirect costs in year one and the model rewards inertia. A four-layer model surfaces the genuine economic question: across five years, with realistic indirect, exit, and risk costs included, which alternative actually costs less? The answer is frequently different.

Direct cost — the visible 60%.

Direct cost is what shows up on the vendor invoice. The components and the common errors:

• Licence fees: the contracted annual cost, projected with realistic uplift assumptions (7–10% annual is the unconstrained default; 3–4% with negotiated caps).
• Support fees: typically 18–22% of licence cost on-premises, often baked into SaaS price, with their own uplift trajectory. Oracle's 4% support uplift compounds aggressively over five years.
• Infrastructure: on-premises hardware refresh cycles or cloud consumption costs; the latter has its own discounting and reservation structure (see AWS EDP and Microsoft Azure MACC).
• Professional services: implementation, migration, and ongoing professional services billed by the vendor.
• True-up cost: expected over-deployment cost based on growth assumptions and historical patterns.

The common error in direct-cost modelling is assuming flat or low uplift; the contractual default is closer to 7–10%, and the compounding effect over five years is material. Use the contractual uplift cap, not a hopeful assumption.

Indirect cost — the hidden 25%.

Indirect cost is internal cost to operate and consume the software. The components:

• Internal labour: administrators, DBAs, integration engineers, training staff dedicated to the platform.
• Integration cost: point-to-point integrations, middleware, API consumption, ongoing integration maintenance.
• Training and change management: initial training, ongoing certification, change communications.
• Compliance and audit support: internal cost of responding to vendor audits, maintaining licensing position documentation, and SAM tooling operation.
• Productivity friction: the cost of unfavourable UX, system downtime, integration failures — typically modelled as an FTE-equivalent figure.

Indirect cost is often 25–40% of total TCO for major vendor relationships. It is the layer that most differs across vendor alternatives (a vendor with weaker UX, harder integrations, or a less mature SAM model can have materially higher indirect cost even at parity direct cost).

Exit cost — the lock-in tax.

Exit cost is what it would cost to leave the platform at end of term and migrate to an alternative. Most TCO models ignore exit cost entirely, which biases the analysis toward staying. A defensible exit cost model includes:

• Data migration: the cost of extracting, transforming, and loading data into an alternative platform.
• Integration rebuild: the cost of recreating integrations against the new platform.
• Retraining: the cost of retraining users, administrators, and operators.
• Transition operating cost: running both platforms in parallel during cutover (typically 6–18 months).
• Knowledge loss: the implicit cost of losing accumulated platform-specific institutional knowledge.

Exit cost is the "lock-in tax" — the difference between the cost of the relationship if it remains profitable on its own terms and the cost of changing course. Vendors price into the relationship knowing exit cost is high. Quantifying it explicitly is a precondition to using it as a negotiation lever (and to deciding when, despite the lock-in, the alternative is the right call).

Risk-adjusted cost — probability-weighted exposure.

Risk-adjusted cost is the expected cost of unfavourable events over the contract term, probability-weighted. The major components:

• Audit exposure: the probability of a vendor audit during the term multiplied by the expected non-compliance back-bill. For Oracle, SAP and IBM, base audit probability across a 5-year window is 60–80%; expected back-bill ranges from low six figures to mid-eight figures depending on estate complexity.
• Scope expansion: the expected cost of forced scope expansion (additional users, modules, geographies) above contracted baseline.
• Price uplift overrun: the probability-weighted cost of uplift exceeding contracted cap or hopeful assumption.
• Vendor failure: the probability-weighted cost of vendor product abandonment, vendor M&A disruption, or vendor business failure.

Risk-adjusted cost is the layer most affected by contract negotiation. The audit clause, uplift cap, true-up structure, and change-of-control terms each move the probability or magnitude of unfavourable events. A TCO model that captures risk-adjusted cost makes the link between contract negotiation and forward economics explicit.

Internal next steps

Three steps make TCO modelling pay back. First, build a standard four-layer TCO template that the procurement and CFO teams agree on. Second, require a TCO update at every major renewal as a precondition to commercial sign-off. Third, build risk-adjusted cost explicitly into the model and link it to specific contract clauses — this is where the negotiation strategy and the financial forecast connect.