A €380,000 portal with 6% adoption and a sponsor who wants to move on
Four months ago a logistics company launched a customer self-service portal. It cost €380,000 and was delivered on time. Of 12,000 active customers, 6% have created an account; monthly support calls fell by 2%, against a target of 15%; customer satisfaction is unchanged. The team of six has €200,000 left in this year's budget. Only three of the nine outcomes named in the business case are instrumented. Sales has now submitted 14 new feature requests, led by a white-label version for two large accounts, and finance wants next year's roadmap in three weeks to set budgets. The original sponsor has moved to another division, and the new sponsor says the portal 'looks fine, so let's move on to the next thing'. The team is proud of the delivery and has already started designing the white-label feature, which the PO has not ordered.
1. With €200,000 remaining and a portal that is not delivering, how do you decide what to fund next?
Read the worked answer
Start from the outcome the business case promised, not from the feature list. The evidence is 6% adoption and a 2% call reduction against a 15% target. That means the hypothesis 'customers will self-serve' is unvalidated, not disproven - and the remaining money should buy learning before it buys more scope.
Discovery first. Segment adoption by customer size, region, and contract type. Interview non-adopters. Instrument the onboarding drop-off and task completion. Check whether the 2% call reduction is real or masked by seasonality. These are cheap and can run in weeks.
Then rank options by expected value and cost of delay. Fixing adoption - awareness, onboarding, usability - may return far more than white-label. White-label is a revenue play for two named accounts: quantify the revenue and the cost, and if it justifies the spend, build a thin slice for those accounts rather than the full feature.
Allocate explicitly. Most of the €200,000 to adoption and discovery, a ring-fenced slice to a white-label spike only if the accounts commit in writing, and a reserve. Do not fund 14 features.
The PO's job is to maximise value, and value here is realised customer behaviour, not delivered scope.
2. Only three of nine business-case outcomes are measured, and the new sponsor says the portal 'looks fine, let's move on'. How do you respond?
Read the worked answer
The measurement gap is why the sponsor can say that. Without baselines and targets, 'looks fine' is the only available judgement - and it is a judgement that will cost the organisation money again next year.
Fix the evidence. Define all nine outcomes with a baseline, a target, an owner, and a review date. Instrument the six missing ones. Then set value reviews at 6 and 12 months. Some benefits genuinely lag, so distinguish 'too early to tell' from 'not working' - that distinction is the PO's credibility.
Answer the sponsor with data and the cost of delay. €380,000 has been spent and value has not yet been realised. Moving on without a value review risks repeating the pattern. Propose a short stage gate: a value review before committing next year's budget, not a delay to the portfolio.
Be direct about the 'looks fine' claim. It is not a value judgement, it is an absence of measurement. If the sponsor still moves on, document the unmeasured benefits and the risk in writing. The PO's job is to make the value conversation possible and to keep it honest - not to win every argument. But silence here is complicity in spending money without evidence.
3. Sales has submitted 14 requests and finance wants next year's roadmap in three weeks. The team has already started designing white-label. What do you do?
Read the worked answer
Fourteen requests is a list, not a plan. Triage it: which are contractual commitments, which are revenue-linked for named accounts, which are speculative; and for each, revenue, cost, risk, and dependencies. Map them to outcomes. Then produce the roadmap as an outcome-based, capacity-constrained forecast with confidence ranges. Three weeks is enough for that if there is an ordering model - what takes longer is deciding the criteria, so start there.
Sales involvement. Give sales a channel and a cadence, not authorship of the backlog. Ask them to bring revenue evidence and customer commitments, not just feature names. White-label should be decided explicitly: if the two accounts commit in writing and the numbers justify it, do a thin slice; if not, it waits.
The team designing white-label before it is ordered is a signal that the PO's ordering is not visible or not trusted. Reinforce the backlog as the single source of truth, and make the reason for the order explicit. Also check whether the team is filling a vacuum - if the PO has been absent from refinement, that is the actual problem.
Finance. A roadmap is a forecast, not a promise. Present ranges, assumptions, and the decisions that would change the plan. A roadmap built on outcomes invites a better budget conversation than a list of 14 features.