Benefits not realised in Harborline's shared services SIAM model
Harborline is a public-sector shared services organisation providing IT to 14 agencies, using three external providers plus an internal infrastructure team under a SIAM model that has run for 12 months. The promised benefits have not materialised. Customer satisfaction has fallen eight points. The SIAM function costs 11% of the IT budget against a planned 6%. Forty per cent of SIAM staff time is spent on facilitation rather than improvement. Providers withhold root-cause data, citing commercial sensitivity. Four software-as-a-service tools have been procured directly by agencies outside the model, creating shadow IT. Five of the 22 SIAM staff have left in six months. Two provider account managers routinely bypass the SIAM process and go directly to agency chief information officers, who take their calls.
1. Categorise the main challenges and identify which are symptoms of a deeper design flaw rather than independent problems.
Read the worked answer
Sort the evidence into behavioural and structural challenges, then look for the root.
- Behavioural: provider reluctance to share root-cause data, account managers bypassing the process, agencies procuring outside the model, and customer satisfaction falling. These are rational responses to incentives: if visibility carries commercial risk and bypassing the process gets faster answers, people will bypass it.
- Structural: the SIAM function costs nearly double the plan, staff spend 40% of time on facilitation, and attrition is high. These suggest scope creep, unclear role boundaries and no career path.
The deeper design flaw is that Harborline established an integration function without establishing the governance, incentives and value proposition that make it worth using. Agencies bypass SIAM because it is not yet delivering a better experience than going direct; providers withhold data because no contract requires disclosure and no consequence follows; staff leave because the role is defined as coordination without authority or development. Cost and satisfaction are symptoms, not causes.
Treating each symptom separately will fail. Intervene on governance and incentives first, then role design, then measurement.
2. Propose a sequenced set of interventions and how you would measure whether they are working.
Read the worked answer
First 90 days - governance and visibility. Establish a service governance board with real authority: agency representation, provider executives with decision-making power, and the SIAM head. Publish a single end-to-end performance report, even if imperfect, so the conversation moves from anecdote to evidence. Open a route for direct escalations to be logged and routed through SIAM rather than refused, so bypassing becomes visible rather than forbidden.
Three to six months - contracts and incentives. Add data disclosure and process compliance to provider scorecards and to contract change at the next review, with consequences that matter. Agree with agencies a simple rule for procuring tools outside the model: a review step with a short turnaround, so the answer is not always no. Redesign SIAM roles with clear accountabilities, a development path and a reduction in low-value facilitation through automation and better process design.
Six to twelve months - value and experience. Reduce the SIAM cost ratio by eliminating duplicated coordination, and report cost alongside outcomes. Run a customer experience survey with the agencies and use it to prioritise improvements.
Measures: customer satisfaction trend, proportion of incidents where SIAM coordinated the end-to-end recovery, percentage of changes and incidents logged outside the process, provider data completeness, SIAM cost as a share of IT spend, staff attrition and internal mobility. Review quarterly; expect satisfaction and data completeness to move before cost.
3. The finance director asks why SIAM costs 11% of IT budget when the plan said 6%. How would you answer, and what would you do about it?
Read the worked answer
Answer honestly: some of the gap is genuine scope growth and some is inefficiency, and the two need different remedies.
Genuine scope growth typically appears when the integration function absorbs activities nobody costed - reporting for 14 agencies, coordination of audits, provider onboarding, or firefighting that should sit with providers. Inefficiency appears as 40% of time on facilitation, duplicated meetings, and manual reconciliation of provider data. The finance director needs the cost broken down by activity and by agency so the comparison with the 6% plan is like for like.
Actions: build an activity-based cost model for the SIAM function; identify activities that should be automated, moved to providers or stopped; renegotiate scope with agencies where their demands exceed the funded model; and set a trajectory rather than an immediate cut, for example a reduction to 8% over 12 months with a stated mechanism. Cutting the function to hit 6% without changing scope would simply push coordination back into the agencies and increase shadow IT.
The framing matters: SIAM is not overhead in the abstract, it is the cost of making a multi-provider model work. If Harborline does not want to pay it, the alternative is fewer providers or a managed service, and that trade-off should be put to the board explicitly.