IT run-cost the CFO can allocate and change.

Anonymised composite. Identifying names, brands, and ledger codes are withheld. Headcount and cost bands are rounded.

The situation

Run-the-business IT was one of the larger standing costs in the group. Renewals arrived on the vendor's calendar. Entities consumed applications they did not own. Ownership sat in the centre; consumption sat in the countries. Nobody could answer, without a six-week extract, what IT cost by capability, by application, or by unit, including cost directly associated with units versus cost they consumed.

A cost-out target had been set from the top. Without a current-state, the default levers were a hiring freeze and a hope that SaaS true-ups would shrink. That does not distinguish unused entitlement from the systems the value streams actually run on.

The tax was heavier because unused seats were dollar-priced. When the rand moved, the same dark licence cost more and still bought nothing.

Questions the current-state had to answer

• What is the cost of running IT?
• What is the cost of my server portfolio?
• What is the cost of my application portfolio?
• What is the cost of my key business and technical capabilities?
• What is the cost of organisational units, including cost directly associated with units, application ownership, and application consumption?
• How did my cost develop over time?

What we did

Documentation and analysis of high-level cost across the organisation, tied into the same current-state as applications and capabilities. The CA(SA) and Advisor | CFO seats led allocation: ownership versus consumption, entity versus shared core, run versus change.

The CA(SA) and Advisor | CFO seats led allocation: ownership versus consumption, entity versus shared core, run versus change. Cost sat on the living picture of the software portfolio so the group could model saving scenarios and simulate the impact of a technology change before the next EXCO pack. A renewal was no longer an invoice. It was a line that could be switched off in a scenario: what happens to cost, risk, and process in 90 days. Architecture as a Service kept the model living, so a true-up or a cost-out ask could be tested against the portfolio as it stood.

What became visible

Track and allocate. IT cost could be read across business capabilities, applications, and teams, not only across vendors.

Time series. Cost development over recent periods showed where growth was new capability and where it was duplicate or dark spend.

Scenarios. Retire, consolidate, and retain paths carried a rand figure and a process impact, so cost-out was not a blunt cut.

Investment. Architecture and cost were one picture, so the next platform ask had to survive the same model as the last unused seat.

Outcome

The group moved from a single "IT run" number to allocated cost the CFO could take to EXCO. Unused seats and paid duplicates were large enough to fund a sequenced harvest at the next true-up. It was an internal list, not a published waste band.

Exact rands stay confidential. The shape of the result is the point: cost-saving scenarios with a process and risk consequence, linked to architecture, not to a spreadsheet that expires at month-end.

Time-to-value sat on the Architecture as a Service path: current-state, living picture, extract. Not an 18-month binder.

The problem · Cost management · Application rationalisation case study