20 September–1 October · W1–2
C → M/H foundations: 9 dev, 6 data, 4 QA, within 10/8/4. H built once. Foundation and permissions evidence precedes later work.
Recommend C + M + H + F + S + X + K for Friday 29 October 2027, with the £1,200 W4 data specialist. Keep acknowledgements manual and the legacy archive read-only. This protects winter referral capacity and removes the unresolved active mandates from the legacy payment process.
Jo’s proposed agreement, dated 20 September 2027. Department priorities are not agreed; no owner has signed this release and no future tests have passed. 09-R6 09-R7 09-R8
All target 29 October, conditional on acceptance and K. These are viable resource and funding plans, not stakeholder approvals. Costs are the specified temporary-service cash only; existing staff effort is shown separately. No salary rates or unprovided development prices are assumed. 09-R1 09-R3
| Decision | 1 · Service + finance Recommended | 2 · Finance + acknowledgements | 3 · Fundraising + consent |
|---|---|---|---|
| Scope beyond C/M/H/F/K | S + XW4 specialist | A + XNo specialist | A + YNo specialist |
| Temporary cash / reserve left | £1,200 / £600Data specialist only | £800 / £1,000Service administrator | £1,700 / £100Administrator £800 + finance bridge £900 |
| Whole-plan effort, Dev / data / QA | 21 / 18 / 13 days | 20 / 16 / 12 days | 20 / 15 / 12 days |
| Winter service capacity | 120/week; 20 above forecast upper bound | 110/week through 10 December; 10 above upper bound | 110/week through 10 December; 10 above upper bound |
| Mandates / finance acceptance | 620 active new; X results W5 before final sign-off | 620 active new; X results W5 before final sign-off | 584 active new; 36 active legacy for six weeks; daily dual-stream reconciliation, expressly accepted bridge |
| Contacts after selected repairs pass | 7,200 opt-in / 1,300 opt-out / 1,500 unknown | 7,200 opt-in / 1,300 opt-out / 1,500 unknown | 7,600 opt-in / 1,500 opt-out / 900 unknown |
| Business acceptance consequence | Service head accepts S in W5; fundraising head accepts manual acknowledgement process | F interface accepted W4 → A tests W5; service head accepts staffed inbox bridge | F interface accepted W4 → A and Y tests W5; data signs recovered evidence; finance and service sign both bridges |
| What waits | A, Y, L, R; no promise of complete cleanup | S, Y, L, R; service bridge needs December decision | S, X, L, R; both bridges need December decision |
| Inspect full allocation |
Counts, capacities, bridge costs and timing derive from 09-R2–09-R7. The preference for choice 1 is Jo’s proposed risk judgement: prioritise the winter service queue and fully linked active payment stream; A’s estimated 12 staff-hours/week saving can wait because manual staffing exists. Choice 3 buys 400 more evidenced opt-ins and 200 more opt-outs, never 600 extra marketing contacts.
Change scope to see the actual role/week load, record boundaries and temporary costs. A selected package is a request, not proof it can launch. Reload or reset restores the original recommendation; nothing here writes to a CRM or records an approval.
Estimates and dependencies: 09-R2; interaction and specialist: 09-R3. Automatic bridges below are necessary consequences of excluding S or X, not free capacity.
Each cell is demand / capacity in person-days, with its package contributions. Foundation W1–2 is deliberately pooled: C/M/H’s internal sequence is already validated. Later columns are hard week boundaries; spare days cannot move between roles or weeks. 09-R1 09-R2 09-R3
Allocation method: place fixed-week work first; allocate flexible work in earliest-deadline order to the first available allowed week. Any remaining demand is shown as excess in its last allowed week. An excess is a visible failed proposal, never extra capacity. Packages are selected whole; spreading their person-days does not partially implement them.
C → M/H foundations: 9 dev, 6 data, 4 QA, within 10/8/4. H built once. Foundation and permissions evidence precedes later work.
F builds and loads in W3; its interface QA and acceptance occupy W4. S/X/Y and L build/data stay in their allowed windows. F interface acceptance is not final finance sign-off.
K alone reserves 2 dev / 2 data / 2 QA. Selected packages must pass first. Demonstrate restoration of a reconciled baseline; the sponsor authorises cutover after all required owners sign. Friday 29 October remains a conditional target.
R can start no earlier than W6 if S passes W5. K already uses all W6 capacity; R cannot fit this release, even if earlier weeks have spare effort. Its 31 January reporting deadline needs a later plan; no post-launch delivery capacity or date is supplied. 09-R2 09-R3 09-R6 09-R7
Observed baseline: 10,000 contacts, including 1,500 without consent evidence; 640 mandates totalling £71,250, including 36 active unresolved links and 20 paused mandates. 09-R4 09-R5
M preserves history and accounts for all 10,000. Evidenced opt-outs remain non-marketable; unknown records remain in restricted quarantine and may be identifiable for non-marketing administration. Fundraising must never access service case notes. Counts do not confer permissions. 09-R4 09-R7
Proposed placement: keep the 20 paused mandates in the new register with processing disabled; preserve their £1,000 in reconciliation, never restart them. Finance must accept this placement and its controls. Without X, no guessed link enters the new export. 09-R5
“Accounted for” is a reconciliation claim. It does not mean every record is marketable, every mandate is on the new payment process, cleanup is complete, or a selected repair has passed. The fixture supplies no further pre-launch evidence for the remaining 900 unknown contacts. 09-R4
The responsibilities below are proposed allocations to supplied roles. Individual names and their acceptance were not provided. Jo must secure an explicitly named daily reconciliation owner if a finance bridge is selected; a role label alone does not close that condition. 09-R5 09-R7
The temporary administrator and legacy payment bridge both expire Friday 10 December. The proposed decision checkpoint is Friday 3 December, chosen here to precede expiry; it is not a sourced commitment. Specialist support is a one-off W4 purchase, not a six-week bridge. 09-R3 09-R5 09-R6 09-R7
Release authority: Jo recommends and assembles evidence; finance, data and affected business owners sign their criteria; the sponsor then authorises cutover. None of those future acts is recorded as done here. 09-R1 09-R7
If a selected package fails: stop this cutover proposal. Jo must re-scope and obtain affected-owner acceptance again. For example, failure of X in choice 1 would add a £900 bridge to the already committed £1,200 specialist, taking cash to £2,100: the existing reserve does not automatically fund that fallback. W6 has no spare role capacity. Do not silently substitute an unfunded bridge or an untested process. 09-R1 09-R3 09-R5
Fictional source records reproduced below. They remain unchanged when you alter the proposal. Source IDs beside the analysis link here; no external research was used.
Launch Friday 29 October, end W6; W1 starts Monday 20 September, five working days/week and no holidays. C,M,H,F,K below are mandatory. Development, data and QA are separate pools; person-days cannot move between roles or weeks. Unused early development cannot bring a dependency forward. £1,800 is available for the specified temporary services; excess needs sponsor approval. No additional developer or QA capacity exists. Jo recommends; sponsor authorises cutover after finance, data and affected business owners sign their criteria.
Figures are person-days. Packages cannot be partially implemented. Work may spread within permitted weeks. C/M/H have a validated internal sequence within W1–2; no hidden foundation task is needed. Later week boundaries are hard dependencies. Build and data work can proceed in parallel within an allowed week; acceptance requires the stated evidence. | ID | Package | Dev / data / QA | Allowed work and dependency | |---|---|---|---| | C | Contacts and role permissions |4 / 0 / 2|Foundation W1–2| | M | Migration spine, consent states, quarantine |2 / 5 / 1|Foundation W1–2, using C| | H | Shared event connector |3 / 1 / 1|Foundation W1–2, using C; built once for F/A/S| | F | Donation processing/consolidated finance export |4 / 2 / 2|Build/data W3; interface QA W4 after M/H| | S | Service referral queue |5 / 3 / 3|Build/data W3–4 after M/H; business QA W5| | A | Automated donation acknowledgements |4 / 1 / 2|Build/data/QA W5 after F's accepted W4 interface; incremental testing fits W5 if capacity does| | R | Service outcomes dashboard |3 / 2 / 2|Earliest build/test week: week after S's business acceptance| | L | Searchable legacy archive |2 / 1 / 1|Build/data W3–5 after M; QA W5| | X | Repair 36 active mandate links |1 / 5 / 2|Build/data W3–4; QA W5; financial acceptance uses these results| | Y | Recover 600 records' consent evidence |1 / 4 / 2|Build/data W3–4; QA W5; no claim for other unknown records| | K | Cutover, recovery rehearsal, controlled release |2 / 2 / 2|Reserved W6, after selected packages pass; includes rollback evidence| F's W4 interface acceptance permits A to start. Final finance sign-off follows X's W5 checks, or the explicitly accepted legacy bridge. Bridge reconciliation testing is included in F, not an extra unspecified task. Supplier marketing estimates count H within F,A,S separately; this table is the corrected non-duplicated basis.
Foundation capacity is pooled across W1–2; later columns are individual weeks. The optional screened data specialist supplies two extra data days in W4 for £1,200; indivisible, charged to the £1,800 reserve. It supplies no development, QA or final owner sign-off. | Pool | W1–2 combined | W3 | W4 | W5 | W6 | |---|---:|---:|---:|---:|---:| | Development |10|6|6|4|2| | Data |8|4|4|4|2| | QA |4|2|2|8|2| If A and S both launch, their shared routing needs an additional **2 development / 1 data / 2 QA days in W5**. It cannot move earlier: it needs F's accepted interface and S's completed configuration. H reuse does not remove it. No increment applies when only A or only S launches.
Source has 10,000 contacts:7,200 evidenced current opt-in; 1,300 evidenced opt-out; 1,500 without sufficient consent evidence. M accounts for all, preserves history and quarantines 1,500; it does not turn unknown into opt-in. Y recovers evidence for exactly 600:400 opt-in and 200 opt-out. No evidence exists for the other 900 in this fixture; no further pre-launch repair is available. Organisation policy prohibits marketing to opted-out or unknown contacts. Quarantined records may remain identifiable for restricted non-marketing administration.
There are 640 mandate records:584 active with verified contact links; 36 active with unresolved links; 20 paused. Corresponding opening amounts £66,000/£4,250/£1,000 total £71,250. All records/amounts must reconcile across new and retained systems; paused mandates must not restart. X verifies all 36 links, permitting all 620 active mandates to use the new process. Without X, 36 remain on a controlled legacy process for six weeks after launch; guessed links cannot enter the new export. The bridge costs £900 once and two finance staff-hours/week; three hours/week are available. A named finance owner must reconcile both streams daily. Finance accepts this limited route in principle if expressly selected and tested, but has not signed a release.
Service inbox capacity 80 referrals/week; current 72, winter range 80–100. S supports 120 and exposes owner, acknowledgement and overdue follow-up. Without S, a temporary administrator costs £800 for six post-launch weeks and raises capacity to 110; availability confirmed. A saves an estimated 12 staff-hours/week, but manual acknowledgements remain staffed until January. Acceptance requires exactly one acknowledgement per valid donation, suppression of disallowed contacts and an actionable failed-message queue. R supports a report due 31 January. Read-only legacy remains available through March at no extra charge, so L is convenient, not required for retrieval. No priorities have been agreed.
Finance signs mandate counts/states and £71,250 reconciliation including any bridge. Data owner signs consent, record accounting and permissions; fundraising must not access service case notes. Service head signs S or the inbox workaround; fundraising head signs A or manual acknowledgements. Six-week bridges expire Friday10 December: decide a permanent fix, renewal or reduced service before expiry. K needs demonstrated restoration of a reconciled baseline and a decision owner, not just a backup checkbox. These are future test requirements, not passed tests.
“Finance wants mandate repair, fundraising wants acknowledgements and better consent coverage, service wants its queue. Cleanup is not finished because the import ran. Show the compromise on 29 October and what must stop or wait. I will not accept a cheap-looking release relying on an unfunded bridge, extra people, or stakeholder assent nobody obtained.”
# BPB2-09 — Charity CRM replacement — choose the first release Suite BPB2 v0.3-pilot; Foray 247, third essai. ### Scenario Hearthside Support has six weeks to release its CRM before the winter appeal. Delivery lead Jo Mensah must reconcile fundraising, service and finance requests with scarce data capacity and an awkward acceptance sequence. Some migration records are usable, some can be repaired, and others lack evidence altogether. Shared components reduce development effort but create interaction tests when features combine. Jo needs a credible first-release agreement on 20 September 2027, including the operational bridges needed to make a deliberately incomplete migration usable. ### Exact task prompt Produce one usable in-browser answer page for the named practitioner, using the supplied fictional evidence. Keep observed facts, forecasts, assumptions and proposed decisions distinguishable, and cite source IDs beside material claims. No external research or real-world certification is required. Recommend Jo's first-release scope and acceptance route for 29 October. Account for shared work once, include interaction work where necessary, and demonstrate a feasible allocation by the supplied planning weeks and roles. State exactly which records and processes may move, what remains quarantined or on the legacy service, and who owns each temporary arrangement. Compare at least two viable stakeholder choices, with costs, dates and acceptance consequences. Distinguish a proposed agreement from stakeholder approval and completed testing. Do not promise complete migration merely because all source records have been accounted for. ### Supplied records **09-R1 — Sponsor's release rules, 17 September 2027.** Launch Friday 29 October, end W6; W1 starts Monday 20 September, five working days/week and no holidays. C,M,H,F,K below are mandatory. Development, data and QA are separate pools; person-days cannot move between roles or weeks. Unused early development cannot bring a dependency forward. £1,800 is available for the specified temporary services; excess needs sponsor approval. No additional developer or QA capacity exists. Jo recommends; sponsor authorises cutover after finance, data and affected business owners sign their criteria. **09-R2 — Remaining estimates and sequence, delivery team, 18 September.** Figures are person-days. Packages cannot be partially implemented. Work may spread within permitted weeks. C/M/H have a validated internal sequence within W1–2; no hidden foundation task is needed. Later week boundaries are hard dependencies. Build and data work can proceed in parallel within an allowed week; acceptance requires the stated evidence. | ID | Package | Dev / data / QA | Allowed work and dependency | |---|---|---|---| | C | Contacts and role permissions |4 / 0 / 2|Foundation W1–2| | M | Migration spine, consent states, quarantine |2 / 5 / 1|Foundation W1–2, using C| | H | Shared event connector |3 / 1 / 1|Foundation W1–2, using C; built once for F/A/S| | F | Donation processing/consolidated finance export |4 / 2 / 2|Build/data W3; interface QA W4 after M/H| | S | Service referral queue |5 / 3 / 3|Build/data W3–4 after M/H; business QA W5| | A | Automated donation acknowledgements |4 / 1 / 2|Build/data/QA W5 after F's accepted W4 interface; incremental testing fits W5 if capacity does| | R | Service outcomes dashboard |3 / 2 / 2|Earliest build/test week: week after S's business acceptance| | L | Searchable legacy archive |2 / 1 / 1|Build/data W3–5 after M; QA W5| | X | Repair 36 active mandate links |1 / 5 / 2|Build/data W3–4; QA W5; financial acceptance uses these results| | Y | Recover 600 records' consent evidence |1 / 4 / 2|Build/data W3–4; QA W5; no claim for other unknown records| | K | Cutover, recovery rehearsal, controlled release |2 / 2 / 2|Reserved W6, after selected packages pass; includes rollback evidence| F's W4 interface acceptance permits A to start. Final finance sign-off follows X's W5 checks, or the explicitly accepted legacy bridge. Bridge reconciliation testing is included in F, not an extra unspecified task. Supplier marketing estimates count H within F,A,S separately; this table is the corrected non-duplicated basis. **09-R3 — Capacity and interaction note, 18 September.** Foundation capacity is pooled across W1–2; later columns are individual weeks. The optional screened data specialist supplies two extra data days in W4 for £1,200; indivisible, charged to the £1,800 reserve. It supplies no development, QA or final owner sign-off. | Pool | W1–2 combined | W3 | W4 | W5 | W6 | |---|---:|---:|---:|---:|---:| | Development |10|6|6|4|2| | Data |8|4|4|4|2| | QA |4|2|2|8|2| If A and S both launch, their shared routing needs an additional **2 development / 1 data / 2 QA days in W5**. It cannot move earlier: it needs F's accepted interface and S's completed configuration. H reuse does not remove it. No increment applies when only A or only S launches. **09-R4 — Signed migration profile, data owner, 19 September.** Source has 10,000 contacts:7,200 evidenced current opt-in; 1,300 evidenced opt-out; 1,500 without sufficient consent evidence. M accounts for all, preserves history and quarantines 1,500; it does not turn unknown into opt-in. Y recovers evidence for exactly 600:400 opt-in and 200 opt-out. No evidence exists for the other 900 in this fixture; no further pre-launch repair is available. Organisation policy prohibits marketing to opted-out or unknown contacts. Quarantined records may remain identifiable for restricted non-marketing administration. **09-R5 — Finance reconciliation, finance lead, 19 September.** There are 640 mandate records:584 active with verified contact links; 36 active with unresolved links; 20 paused. Corresponding opening amounts £66,000/£4,250/£1,000 total £71,250. All records/amounts must reconcile across new and retained systems; paused mandates must not restart. X verifies all 36 links, permitting all 620 active mandates to use the new process. Without X, 36 remain on a controlled legacy process for six weeks after launch; guessed links cannot enter the new export. The bridge costs £900 once and two finance staff-hours/week; three hours/week are available. A named finance owner must reconcile both streams daily. Finance accepts this limited route in principle if expressly selected and tested, but has not signed a release. **09-R6 — Department requests, 19 September.** Service inbox capacity 80 referrals/week; current 72, winter range 80–100. S supports 120 and exposes owner, acknowledgement and overdue follow-up. Without S, a temporary administrator costs £800 for six post-launch weeks and raises capacity to 110; availability confirmed. A saves an estimated 12 staff-hours/week, but manual acknowledgements remain staffed until January. Acceptance requires exactly one acknowledgement per valid donation, suppression of disallowed contacts and an actionable failed-message queue. R supports a report due 31 January. Read-only legacy remains available through March at no extra charge, so L is convenient, not required for retrieval. No priorities have been agreed. **09-R7 — Acceptance and ownership, 20 September.** Finance signs mandate counts/states and £71,250 reconciliation including any bridge. Data owner signs consent, record accounting and permissions; fundraising must not access service case notes. Service head signs S or the inbox workaround; fundraising head signs A or manual acknowledgements. Six-week bridges expire Friday10 December: decide a permanent fix, renewal or reduced service before expiry. K needs demonstrated restoration of a reconciled baseline and a decision owner, not just a backup checkbox. These are future test requirements, not passed tests. **09-R8 — Jo's review, 20 September.** “Finance wants mandate repair, fundraising wants acknowledgements and better consent coverage, service wants its queue. Cleanup is not finished because the import ran. Show the compromise on 29 October and what must stop or wait. I will not accept a cheap-looking release relying on an unfunded bridge, extra people, or stakeholder assent nobody obtained.” ### Browser deliverable and completion conditions Provide scope selection, role/week capacity, dependencies, interaction work, migration coverage, acceptance owners and temporary arrangements. Selecting A with S must expose the W5 integration load; X/Y must update record claims and bridge costs. Compare two feasible release routes. Keep proposed sign-offs distinct from actual evidence; expose December expiry and its preceding decision. Original assumptions must remain recoverable. The page must work in the supplied browser/runtime; a download button or file preview alone is insufficient. Include a clear “Return to gallery” link using the shared shell route, or ../index.html if no route is supplied. Required controls must change the dependent results, not just their appearance. Original evidence and the initial dated position must remain recoverable; hypothetical changes do not create approvals or alter source records. Optional exports are secondary.