Finance proposes charging teams for platform consumption to drive cost awareness. What is your view?

What the interviewer is testing

Whether you can reason about incentives, and whether you recognise that a funding mechanism shapes platform behaviour more than any technical decision.

The concern with chargeback

The goal — cost awareness — is legitimate. The mechanism produces predictable distortions:

Teams avoid the platform to avoid the charge, building shadow alternatives that are cheaper on their budget and more expensive for the organisation. This directly undermines the consolidation the platform exists to achieve.

Teams optimise the chargeback metric rather than engineering outcomes — fewer environments, less telemetry, skipped non-production testing. Observability is usually the first casualty, and its absence appears later as longer incidents.

The platform becomes reluctant to decommission anything that generates internal revenue, which inverts its purpose.

Cross-charging consumes real effort in metering, disputes and reconciliation, none of which produces engineering value.

The alternative

Showback: teams see exactly what they consume and what it costs, with no internal invoice.

This delivers nearly all of the cost awareness — the behaviour change comes from visibility, not from the transfer — without the avoidance incentive. Teams that can see one dashboard costing more annually than the analyst who built it act on that, whether or not they are billed.

Pair it with central funding for the platform, and measure the platform on adoption plus the delivery metrics of the teams it serves.

Where chargeback is defensible

Genuinely elastic, discretionary consumption where the team controls the volume and the cost is material — large analytics compute, long-lived preview environments, GPU capacity. Charging for that creates the right pressure. Charging for the core paved road does not.

What a strong answer adds

The funding model determines survival: platforms funded as projects decay into unowned legacy within two years, because a project is funded to build and then disbands, while a platform must be funded to serve continuously.

Common weak answers

Rejecting cost visibility. Accepting chargeback without naming the avoidance incentive.