The £2 Million Tooling Shortcut at Northvale Retail Bank
Northvale Retail Bank serves 4.2 million customers through a 24x7 mobile app. Delivery is split between a 60-person in-house core banking team, two infrastructure outsourcers (data centres and network), a SaaS payments provider and a 90-person internal service desk. Over nine months, three P1 outages were each traced to a handoff: monitoring alerts landed in the wrong queue, providers diagnosed in parallel without sharing findings, and the bank's change calendar was invisible to two suppliers. Average restoration was 3 hours 40 minutes against a 60-minute target. The CIO has approved £2 million for an AI service integration platform and expects lead time to fall within one quarter. The SIAM lead must write the business case. The regulator has asked for evidence of end-to-end accountability, and no provider is contracted to coordinate the others.
1. Why will the £2 million platform, on its own, fail to reduce restoration time? Use SIAM principles to justify your answer.
Read the worked answer
Diagnosis. The evidence points to an integration and accountability gap, not a data or tooling gap. In each outage the technology produced signals; the failure was that no one was obliged to act on them together. A platform would make existing fragmentation faster, not remove it.
Principles applied. End-to-end service focus means the bank must define the outcome it is buying: restoration of the mobile service, not three separate component recoveries. Collaboration across the ecosystem means providers must share findings during an incident, not after it. Transparency means one timeline and one severity model visible to all parties. 'Integrate, don't dictate' means the bank should not impose one supplier's tool and data definitions on the others and expect cooperation.
Why the platform alone fails. Tooling can route alerts and display a dashboard, but it cannot decide who declares a major incident, which provider must join the bridge, or what data each supplier is contractually required to publish. Without those decisions, each supplier will populate only what its contract demands. Restoration time is driven by time to engage the right people and reach a shared diagnosis, not by dashboard latency. A £2 million platform deployed first would create a reporting layer over unchanged behaviour, and could harden silos by making each provider's local performance more visible than the end-to-end outcome.
2. What should the SIAM lead propose for the first 90 days, and which three measures would show whether the approach is working?
Read the worked answer
First 90 days: a proof of value, not a platform procurement.
- Appoint an end-to-end major incident owner with written authority to convene all four delivery parties, backed by a CIO letter and a contract change. This role must not sit inside one of the infrastructure outsourcers.
- Agree a cross-provider severity matrix and a 15-minute engagement rule: all relevant providers join one bridge, maintain one timeline and accept one incident manager.
- Run two joint simulations based on the last two P1s. Publish a no-blame timeline and fix the top three handoff defects, such as alert routing and change visibility.
- Define a minimal data contract for alerts, configuration items, changes and status, and pilot it with one provider before committing platform spend. Hold back 60–70% of the £2 million until behaviour changes.
- Establish a weekly service integration forum with a decision log, so cross-provider issues have a route to resolution.
Three measures. Median time from first alert to all relevant providers engaged, target 15 minutes or less. Median time to a shared diagnosis, which tests collaboration rather than tool speed. Percentage of changes visible to all providers at least seven days ahead, target 95%, which tests whether the change calendar is truly integrated. Also track repeat handoff causes, not just mean time to restore, because a falling MTTR with the same handoff defects would be a false signal.
3. Which SIAM principle is most at risk if the platform is deployed before the operating model is agreed, and what is the likely consequence?
Read the worked answer
Principle at risk: collaboration and transparency, with a secondary risk to 'integrate, don't dictate'. If the platform is deployed first, the bank will tend to impose one supplier's tool, data model and workflow on the others. Providers will comply minimally, protect commercially sensitive information and optimise their own dashboards. Service credits may turn the shared timeline into evidence for blame rather than learning.
Likely consequence. The bank gets a technically integrated tool with behaviourally fragmented teams. Alerts route faster, but no one accepts end-to-end ownership, so parallel diagnosis, invisible changes and queue errors continue. Suppliers learn to feed the tool enough to avoid penalties, not enough to solve incidents. The £2 million becomes a cost with no measurable reduction in restoration time, and the SIAM lead loses credibility with the CIO and regulator.
Mitigation. Publish a data-sharing agreement and a no-blame review policy before go-live; make participation in joint incident and change processes a contractual obligation; and tie platform rollout to demonstrated behaviour change, such as the 15-minute engagement rule and change visibility, rather than to licence deployment. The platform should follow the operating model, not define it.