The Sponsor's Pet Epic and the Compliance Mandate
A life insurer runs six squads on an annual funding cycle. The product owner has 260 story points of capacity across six squads for the next two sprints and three candidate epics:
- Epic A - regulatory reporting change. Required by the regulator in 8 weeks; estimated 120 points; failure to deliver triggers a remediation order and a public finding.
- Epic B - broker self-service quotes. Expected £1.8m of annual new revenue; estimated 90 points; delayed revenue costs roughly £35k per week.
- Epic C - data warehouse migration. 140 points, sponsored personally by the CFO, no named end user, and no other epic depends on it.
In addition, unplanned production support has consumed 60 points in each of the last three sprints and shows no sign of falling. The CFO has told the product owner to "just do all three" and to absorb the support work inside the existing capacity. Two squads carry 15 points of carry-over every sprint, and the release manager has flagged that quality checks are being shortened to hit sprint commitments.
1. Sequence the three epics and justify your order. Your justification must explain how you treat separately scored and non-negotiable work, rather than applying one ranking technique to everything.
Read the worked answer
Do not score all three on the same scale - the first step is classifying the demand.
Epic A is not a value trade-off. It has a hard external date and a penalty, so ranking it by cost of delay alongside revenue work produces a number that means nothing. Sequence it first and fully, but plan to finish by week 6 of the 8-week window to leave soak time and a remediation buffer - a regulatory change delivered on the deadline with no recovery room is a schedule that has not been risk-managed.
For the remaining discretionary work, use cost of delay divided by job size:
- Epic B: £35k per week ÷ 90 points ≈ £390 of delay cost per point per week. That is a real, quantifiable figure.
- Epic C: no named user and no dependent epic means the cost of delay is unquantified, not zero. A 140-point item with an unknown cost of delay and the largest job size will always rank last on any codified method, and that ranking should be shown to the CFO rather than applied silently.
The arithmetic that governs everything else: 260 points available, 120 consumed by support, leaves 140. Epic A takes 120 of that. That leaves 20 points for 230 points of B and C. The real question is not ordering but scope, and the product owner should present it that way.
Recommendation: A protected and finished early; B funded as a vertically sliced MVP within residual capacity; C formally deferred with a written rationale and a request that the CFO supply a cost-of-delay figure - if it cannot be articulated, that is itself the argument.
2. The CFO says "just do all three" and to absorb the 60 points of support work inside current capacity. What do you do, and what exactly do you take to the CFO?
Read the worked answer
Accept the instruction as a request, answer it with arithmetic, and force an explicit choice rather than a silent quality cut.
The sum is not a matter of opinion: 260 points of capacity; 120 points of support over two sprints; 120 points for Epic A; 230 points of Epic B and C. That is a 190-point gap. "All three" is arithmetically impossible in the stated window, and the release manager has already told you how the gap is currently being closed - by shortening quality checks. That is a leading indicator of a defect spike and, in an insurer, a conduct-risk issue, not a delivery inconvenience.
Take three costed options, not a complaint:
- Reduce scope of B to an MVP slice (quote-and-refer only), deferring the remainder - loses part of the £1.8m but starts earning from the earlier slice.
- Defer C entirely to the next funding cycle with the trade-off stated in revenue and risk terms.
- Add capacity - a dedicated support squad or contractor pair - costed against the £35k weekly delay cost of B.
Also put the support load on the table as a first-class line item: ring-fence a fixed 20% of capacity (or one squad) for it. Invisible operational work that is absorbed by hope is the single most common cause of missed sprint commitments.
If the CFO still insists, record the decision and the risk acceptance in writing. The product owner's job is to make the trade-off visible to the person accountable for it, not to absorb it.