The real cost of a failed ICT procurement.
When a public-sector ICT contract fails, the headline cost is the value of the contract itself. The real cost is several multiples larger and is rarely captured anywhere in the entity's books. The hidden cost lives in the procurement re-run, the transition between suppliers, the irregular-expenditure provision raised by the auditor, and the personal exposure of the accounting officer whose name signed the original award letter. Each of these costs is preventable. Each is regularly paid.
1. The direct cost
Direct cost is the visible loss. If the contract was R20 million per year for three years and the supplier delivered nothing usable in year one, the entity has spent R20 million on a non-functional outcome. The procurement audit will quantify this number and table it for the audit committee. It is the smallest of the costs but the only one the entity is forced to write down formally.
In most failed procurements, however, the direct loss is partial rather than total. The supplier delivered something. The something is not what was specified, or it works but at a service level lower than the contract required, or it works for part of the user base but not others. Quantifying the direct loss in these partial-failure cases is harder than it looks. The auditor's number and the entity's number often diverge.
2. Procurement re-run cost
Re-running a major ICT procurement event is expensive in cash and in time. Cash cost: between R1 million and R3 million for a tender of meaningful size, accounting for internal time, legal review, supply chain officials, advertising, and the inevitable bid clarifications and protests. Time cost: 8 to 14 months from cancellation of the failed contract to award of the replacement.
During that 8 to 14 months, the entity is in a procurement limbo. The failed supplier is still operationally engaged in many cases, because turning them off without an alternative would create worse problems. Service quality during the limbo is rarely good. Many entities run an extension to the failed contract while the re-procurement happens, which then becomes an irregular-expenditure exposure of its own.
3. Transition cost
The transition from the failed supplier to the replacement supplier carries its own cost. Data migration. Service handover. Parallel running during the cutover. Knowledge transfer that the failed supplier has little incentive to provide thoroughly. Replacement of any client-side assets the failed supplier was responsible for maintaining.
Transition cost for a managed services contract typically ranges from 8 to 15 percent of the annual contract value. On a R30 million per year contract, that is R2.4 million to R4.5 million in one-off transition spend that the entity didn't budget for and the new supplier may or may not absorb.
4. Irregular expenditure provision
If the failed contract is found to have been irregularly procured (Section 32 of MFMA, Section 38 of PFMA), the entity raises an irregular expenditure provision in its annual financial statements. The provision sits as a contingent liability until the National Treasury, the audit committee, or the legislature condones the expenditure or directs it to be recovered.
The provision matters because the audit committee and the executing authority are required to act on it. Condonation often takes 12 to 24 months. During the period, the entity's audit opinion may be qualified, and the accounting officer's performance scorecard reflects the open finding. Some failed contracts produce irregular expenditure findings that persist on the books for years.
5. Accounting officer exposure
The accounting officer who signed the original award letter carries personal exposure under Section 38 of the PFMA or Section 60 of the MFMA. Personal exposure in this context rarely means prosecution, but it does mean the official's performance assessment reflects the failure, their next contract renewal is contested, and their political relationships within the entity are damaged.
For a municipal manager or a departmental DG, a single high-profile procurement failure can determine career trajectory for the following five years. The reputation cost does not appear on any financial statement but is felt acutely by the individual carrying it.
Three South African examples
Three publicly-documented examples illustrate the pattern. The Limpopo provincial education ICT modernisation in the late 2010s produced direct costs of around R200 million, transition and remediation costs estimated at 3-5x the headline, and a procurement re-run that took three years. The South African Police Service ePolice/CAS modernisation programme has had multiple cycles of vendor change since the early 2010s, each adding transition and re-procurement cost. Several large state-owned company ICT programmes have followed similar patterns where the visible contract failure is the smaller part of the eventual write-down.
The pattern is consistent across categories: the entity counts the direct cost, doesn't count the surrounding cost, and consequently under-invests in the procurement discipline that would have prevented the failure.
Five practices that prevent the failure
The patterns above are preventable with five disciplines applied during procurement: rigorous bid evaluation that doesn't reward paperwork compliance over delivery evidence; mandatory reference calls (the procurement officer phones two named contacts at previous clients before award); a discovery phase priced separately from the main contract, so the supplier's understanding of the work is tested before the bulk of the spend is committed; a documented exit clause in the contract that doesn't penalise the entity for early termination on demonstrated breach; and a supplier-management cadence that catches drift early enough to be corrected.
None of these practices is expensive. Each is missing on most failed procurements after the fact. Adding them costs less than 1 percent of the contract value and prevents 80 percent of the failure modes that produce the hidden costs.
The procurement failures referenced above are drawn from publicly-documented audit reports and from the Auditor-General's general reports on national, provincial, and local government. Specific numbers vary; the patterns are general and repeated.
