← The Technology Ledger

Worked example / Illustrative only

A renewal decision.
A recovery question.
One board conversation.

How a Technology Ledger can turn incomplete evidence into a decision the board can own.

Fictional example. Not client work.

North Quay Services is an invented organisation. Every figure, document, finding, person-role and scenario below is fictional, created to show the form of the thinking. This is not a claim about a client, an achieved saving, an audit or a quote.

Situation A regional services business has a core platform renewal in 30 days.

Decision Renew, right-size or migrate — without accepting an unexplained recovery risk.

Scope A documentation and interview review. No production access, security testing or legal opinion.

01 / Current State Ledger / Fictional

The platform is running.
Recovery is not yet demonstrated.

The renewal should not be treated as a price-only decision. The supplied fictional records support a cost challenge, but do not establish that the business can recover its core service within the board’s tolerance.

What the evidence says

  • E1 — Contract extract: annual subscription S$240,000; renewal decision in 30 days. Notice and exit rights need specialist review.
  • E2 — Usage export: 15% of paid accounts show no activity in the sampled 90 days. That suggests a question about utilisation, not automatic removable spend.
  • E3 — Recovery record: backups are logged as successful. The latest supplied restoration record covers a single database, not a business service.
  • E4 — Management interview: the operating team expects recovery within four hours. No approved business tolerance or end-to-end timed exercise was supplied.
  • E5 — AI inventory: two staff tools are listed; interviews mention a third. Data flows and retention for that tool are undocumented.

What is not established

  • Whether inactive accounts can be removed under the agreement or are needed seasonally.
  • Whether the four-hour recovery expectation is achievable across data, identity, integrations and people.
  • The legal meaning of renewal, termination and data-export provisions.
  • The unlisted AI tool’s data handling or the completeness of the inventory.

Confidence: moderate in the recorded cost baseline; low in service recovery capability. Document presence is not evidence of operating effectiveness.

Review lens: cost, risk, capability and AI posture. A critical unknown remains visible; it is not averaged away by strengths elsewhere.

02 / Risk Heatmap / Fictional

Prioritise the decision exposure,
then the effort.

High impact / Evidence gap

Recovery capability is unproven

Impact
High — core-service interruption
Effort
Low to scope: 6–10 internal hours. Restoration testing and remediation effort remain unpriced.
Resulting priority
P1 — start in days 0–7, alongside renewal checks
Why this priority
High business impact and a modest initial evidence task justify immediate scoping. Unknown test/remediation effort does not reduce the priority.

Why it matters: a prolonged core-service outage could interrupt revenue and customer delivery. Likelihood cannot be inferred from the supplied records.

Next evidence: approved business tolerance, dependency map and a controlled end-to-end restoration exercise. Technology lead owns the evidence; operations validates business recovery. Testing requires its own authorised scope.

Time critical / Commercial

Renewal may close off options

Impact
High — a missed notice deadline may lock in cost and constrain exit
Effort
Low initial check: 4–8 internal hours. Specialist legal effort depends on the contract.
Resulting priority
P1 — verify dates in days 0–7; decide before the confirmed deadline
Why this priority
A time-sensitive, high-impact constraint can be clarified with relatively little initial effort. Run in parallel with recovery scoping; do not wait for a migration study.

Why it matters: a notice deadline could constrain negotiation or exit. A migration estimate is not an executable alternative within 30 days.

Next evidence: confirm dates and rights with procurement and legal advisers; request priced renewal options before choosing. CFO owns the decision.

Medium impact / Unquantified

AI use exceeds the documented inventory

Impact
Medium — incomplete visibility of information use; actual exposure unquantified
Effort
Low initial inventory check: 4–8 internal hours. Remediation effort is unknown until data flows are established.
Resulting priority
P2 — complete the inventory check in days 31–90; escalate earlier if sensitive use is found
Why this priority
A bounded evidence task is worthwhile, but no incident or imminent deadline is established. Prioritise the two P1 decisions while keeping a clear escalation trigger.

Why it matters: the board lacks a complete view of where information goes. The interview is a lead to examine, not proof of a privacy incident.

Next evidence: identify the tool, owner, data categories, supplier terms and current controls. Business owner and technology lead validate together.

Qualitative priorities only. No incident probability, compliance conclusion or certification is asserted. Resolve high-impact uncertainty before relying on a green status.

03 / Cost & Vendor Findings / Fictional

Three options. One consistent cost basis.

Illustrative planning estimates in SGD, excluding tax, on the same 12-month basis. None is a supplier quote. The baseline is S$240,000 annual subscription; business disruption, internal time and residual contract charges remain unpriced.

A / Renew unchanged

Preserve continuity

Annual subscription
S$240,000
One-off cost assumed
S$0
Modelled year-one cost
S$240,000
Change vs baseline
S$0

Least immediate change. Recovery evidence still needs attention. Do not infer that renewing transfers accountability to the vendor.

B / Right-size

Challenge the commitment

Annual subscription assumed
S$222,000
One-off work assumed
S$6,000
Modelled year-one cost
S$228,000
Potential net reduction
S$12,000 in year one

Assumes only part of apparent unused capacity is removable and the supplier agrees. Validate seasonal use, minimum commitments and the actual quote.

C / Migrate

Test a real alternative

Annual subscription assumed
S$192,000
Migration + exit assumed
S$20,000
Modelled year-one cost
S$212,000
Potential net reduction
S$28,000 in year one

Largest modelled reduction, highest execution uncertainty. Integration, export, overlap and recovery work are unvalidated; cannot yet be relied on before renewal.

Finding, not a savings promise

The 15% inactive-account observation does not establish a 15% saving. Option B models S$18,000 gross reduction less S$6,000 assumed one-off work. Option C models S$48,000 gross reduction less S$20,000 assumed one-off costs. These net figures are not achieved savings and do not include the unpriced items above. Timing and contract terms can change the result.

Judgement: validate B now and preserve C as an evidenced alternative. Do not allow an untested migration estimate to force a rushed platform change.

04 / Prioritised Action List / Fictional

Make the unknowns actionable.

  1. Days 0–7 / CFO + procurement

    Establish the decision window

    Verify renewal and notice dates, minimum commitments and exit obligations. Obtain legal review where needed. Dependency: the full signed contract. Output: confirmed dates and an option the business can actually exercise.

    Cost & effort / Illustrative only: Fictional planning allowance: 4–8 internal hours for the CFO/procurement check; internal time is not monetised. Legal review is unpriced: the CFO must obtain the full contract and an adviser’s scope/quote before approving it. No legal-review fee is included in Options A, B or C.

  2. Days 0–7 / Technology lead + operations

    Agree the recovery question

    Define the business service, tolerated outage, dependencies and evidence required. Scope a safe exercise separately. Dependency: operations agrees the business tolerance. Output: an authorised test plan, not an unsupported recovery claim.

    Cost & effort / Illustrative only: Fictional planning allowance: 6–10 internal hours from the technology lead and operations to define tolerance, dependencies and a test brief; internal time is not monetised. The restoration exercise and any remediation remain unpriced. The technology lead must obtain the dependency map, environment requirements and a specialist estimate before authorisation. These costs are excluded from all three option totals.

  3. Days 8–21 / CFO + technology lead

    Validate right-sizing

    Reconcile inactive accounts with seasonal and retained roles. Obtain a written quote and cost the change. Dependency: supplier terms and service-owner sign-off. Output: a comparable, decision-ready year-one cost.

    Cost & effort / Illustrative only: Fictional planning allowance: 8–12 internal hours to reconcile usage and obtain a quote; internal time is not monetised. Option B already carries a S$6,000 assumed one-off implementation provision, used only if B is selected; it is not an extra S$6,000 on top of the option model. The CFO must validate what that provision covers against the supplier quote and change plan before committing.

  4. Before renewal / Board + CFO

    Choose and record the residual risk

    Decide on the validated commercial options. Name who accepts any remaining recovery uncertainty, what evidence is still due and when it returns to the board. Dependency: confirmed contract rights and priced options.

    Cost & effort / Illustrative only: Fictional planning allowance: 2–3 internal hours for the board/CFO decision and risk record; internal time is not monetised. No additional cash spend is priced for this decision step. Any implementation, specialist advice or testing requires its own approved estimate; no new amount is silently added to the option totals.

  5. Days 31–90 / Technology lead + business owners

    Close the evidence gaps

    Complete the authorised recovery work, reconcile the AI inventory and assess the migration alternative if still relevant. Funding, specialist work and implementation remain separately scoped. Output: updated evidence and a revised decision, not a task list left open indefinitely.

    Cost & effort / Illustrative only: Fictional planning allowances: 4–8 internal hours for the AI inventory check and 8–16 internal hours for an initial migration assessment, if pursued; internal time is not monetised. Recovery testing, AI remediation and any external assessment remain unpriced: the technology lead must obtain a test scope, validated data flows and supplier estimates. Option C already includes a S$20,000 assumed migration-plus-exit provision, used only if C is selected; it excludes the unpriced work just listed and is not added a second time.

Cost reconciliation / Fictional, partial cost model

The internal-hour allowances above are additional effort disclosures, not cash charges and not monetised savings. No daily rate is assumed. Common diligence, recovery testing, AI remediation and specialist advice remain unpriced and are excluded from all three option totals; an owner and the evidence needed to price them are stated above. They may reduce or eliminate the modelled reductions.

A: S$240,000 + S$0 specified one-off cost = S$240,000. B: S$222,000 + the same S$6,000 provision = S$228,000, or S$12,000 below baseline before the exclusions. C: S$192,000 + the same S$20,000 provision = S$212,000, or S$28,000 below baseline before the exclusions. B and C are alternatives: do not add their provisions together. The action list does not approve either branch or create a second implementation budget.

05 / Board Summary / Fictional

The board paper, condensed.

Decision requested

Validate the renewal options now; keep recovery risk explicit.

Ask the CFO to return before the verified renewal deadline with priced options and confirmed contract rights. Ask the technology lead and operations owner for an agreed recovery tolerance and a separately authorised evidence plan. Do not approve migration on this review alone.

Rationale
Right-sizing could offer a lower-risk commercial improvement. Recovery uncertainty is material whichever supplier option is chosen.
What could change the recommendation?
A restrictive notice clause, seasonal licence requirements, an uneconomic quote, or recovery evidence showing that the assumed service cannot be restored.
Acceptance and escalation
The board names the accountable risk owner and return date. If the commercial window is shorter than expected, escalate the decision immediately rather than hide the missed dependency.
Not approved by this paper
Spend, production testing, a supplier change, certification or implementation. Those require explicit authority and an agreed scope.

Fictional illustration throughout. A real Ledger depends on the agreed scope and available evidence; it will not necessarily contain these findings, options, figures or recommendations.

Have a decision like this?

Start with the business question and the deadline. Keep confidential documents back until information-handling arrangements are agreed.

Discuss your situation →Review the three-week Ledger process →