Benefits at risk on a stalled planning-system rollout
Harbourline Components, a 640-person manufacturer, went live nine months ago with a planning and scheduling system replacing 14 years of spreadsheet planning. The business case promised a 12% reduction in excess inventory and a 6% cut in expedited freight. At the three-month review, planners used the system for 38% of scheduling decisions; at nine months it is 44%. Two of five plant planners keep personal spreadsheets they call 'the real plan'. Planner headcount was reduced by four at go-live on the assumption of efficiency. The operations director sponsors the programme but has attended two of nine steering meetings; plant managers say the system 'doesn't understand our constraints'. Union consultation about role changes was completed, but no follow-up sessions were held. You are the change lead with eight weeks before the annual planning cycle, when the CFO expects the benefits to appear.
1. Diagnose why adoption has stalled. Separate what is a system-fit problem from what is a change-leadership problem, and state what evidence would move your diagnosis.
Read the worked answer
44% at nine months after 38% at three months is a plateau, not a slow burn. That shape argues against 'more training' and towards a usefulness-and-trust problem.
Separate the two explanations:
- Fit gap (tool problem). Planners say the system does not understand their constraints. If real constraints - changeover times, tooling, minimum run lengths - were never configured, the system genuinely produces unusable plans and the planners are right. Evidence: take the two 'real plan' spreadsheets and code every deviation from the system plan, classifying each as a missing constraint, a data error, or planner preference. One week of that analysis settles most of the argument.
- Change-leadership problem. Headcount was cut by four at go-live, before any efficiency existed. Planners reasonably concluded that using the system faster would cost more jobs. A sponsor who attends two of nine meetings has told the organisation the programme is not a priority, whatever the comms say. The union consultation was a transaction rather than a conversation, and the absence of follow-up left the anxiety unresolved.
Evidence I would gather before acting: the spreadsheet deviation analysis, decision-level adoption rather than login counts, time lost reconciling two plans, and interviews with the two planners keeping 'the real plan' - they are the most informative users, not the least compliant.
The trap is treating resistance as the problem. Here much of the resistance is rational.
2. You have eight weeks. Set out a sequenced plan, including what you will deliberately not do.
Read the worked answer
Sequence matters: trust before tooling, tooling before targets.
Weeks 1–2 - facts and sponsor. Re-engage the operations director with a specific ask: chair steering, own one decision on headcount, and spend two hours on a plant floor. Run the workaround analysis. Hold one-to-ones with the two shadow planners and the plant managers, listening for what the system actually costs them in time and risk.
Weeks 3–4 - fix what is genuinely broken. Take the top three constraint gaps and get them configured. If the vendor cannot deliver, agree a documented interim manual step rather than a silent spreadsheet, and publish the decision and its reason so planners can see the system is being corrected.
Weeks 5–6 - honesty about roles. Re-open the union conversation with the real position on headcount, and separate 'efficiency' from 'redundancy' if that is true. Ambiguity is now doing more damage than bad news would.
Weeks 7–8 - the planning cycle. Agree with the CFO that the annual cycle runs on a single plan with a named exception route, and resource it. Set an adoption measure based on decisions made in the system, not logins.
What I will not do: launch a communications campaign, mandate the system, or run more generic training. All three would confirm that the programme's answer to every problem is persuasion rather than change.
3. What would you tell the CFO about the promised 12% inventory reduction?
Read the worked answer
The 12% is not a commitment the organisation can currently keep, and the CFO needs to hear that with a number and a mechanism attached.
The business case assumed plans run through the system. At 44% decision adoption, the mechanism that produces inventory reduction - fewer manual buffers, less expediting - is only partly operating. Two consequences follow. First, realised benefit is likely to be a fraction of 12%, and I would model it as a curve tied to adoption rather than a step at go-live. Second, four planner roles were removed ahead of the benefit, so part of the saving is already booked against an efficiency that has not been delivered. That is an accounting and credibility risk, not only a change risk.
I would give the CFO three things: the evidence from the workaround analysis, a re-based benefit profile with a decision gate at, say, 70% decision adoption, and the specific decisions needed from the executive - sponsor time, role clarity, and a single plan for the annual cycle.
I would also name the dis-benefit of stopping: the licence cost is sunk, the spreadsheets are a control risk, and a second failed planning implementation would make the next one considerably harder. The message is 'later and lower, with a plan', not 'failed'.