The Mandate That Almost Had No Sponsor
What happens when a national-scale banking program has executive approval, a business case, a $14.2M budget, and a 19-microservice architecture, but no one willing to own the outcome.
About This Series
This is a 22-post series that follows a single fictional program, Project Falcon at Atlas Bank, from inception through closing. Each post covers one decision point, one framework, one template, and one AI prompt used to navigate the moment. Across the series, the same technology stack, the same stakeholder cast, and the same financial baseline carry through, so the program evolves as a traceable whole rather than twenty-two disconnected lessons.
The goal of the series
To show what disciplined program management actually looks like in a hybrid delivery environment, where agile workstreams, waterfall workstreams, and regulated governance coexist. Standard program management frameworks are woven into the narrative. Tools in use include Jira for the agile workstreams, MS Project for the waterfall workstreams, Confluence for documentation, Slack for communication, and Miro for workshops. AI is treated as a working collaborator, not an accessory.
Who it is for
Mid-level and senior project and program managers navigating technology programs of meaningful scale. Technology leaders who want to see how their architectural decisions land inside a PM governance frame. Junior PMs building a mental model of how real programs actually unfold, rather than how textbooks say they should. Anyone considering hiring a senior practitioner or engaging one as a consultant.
What this is and is not
Project Falcon is a fictional composite created for teaching purposes. Atlas Bank does not exist. The named individuals are invented characters, not real people. The numbers, dates, and decisions described are illustrative. Any resemblance to actual programs, organizations, or individuals is coincidental.
The technology stack, the governance patterns, and the decision dynamics are drawn from general practice in regulated fintech and banking programs. These are the shapes of the work, not the specifics of any one engagement.
Project Falcon at Atlas Bank
Before the story of the missing sponsor, the story of the program itself. The decision described later in this post, to stop all planning work for two weeks at the very beginning, only makes sense against the shape of what Falcon was trying to do and what stood to be lost if it failed.
Atlas Bank is a mid-sized universal bank operating across three markets in a regional economic bloc. Retail, corporate, and SME divisions. Approximately four million customers at the time Falcon began. A legacy core banking system, siloed, branch-first. Regulated by a central bank that had recently issued a directive on digital access. A competitor had launched a comparable digital platform six months earlier and was already taking market share. The board had approved the capital expenditure publicly. The chief executive's next contract renewal was tied to delivery.
That is the pressure context, and it matters. Every decision the program touched carried a political shadow behind it. A program manager who does not read the political room is a program manager who gets blindsided in the steering committee.
The Mandate
Project Falcon was the delivery vehicle for a national digital banking platform. One platform, three markets, two channels. A smartphone application for customers with data access, and a USSD channel for customers on feature phones. The smartphone app was the future. The USSD channel was the present. One could not be built without the other. Losing the feature-phone segment meant losing financial inclusion, which meant losing the regulatory alignment that justified the program in the first place.
1. Enable core banking transactions (transfers, payments, balance inquiries, statements) across both channels.
2. Integrate a KYC and AML pipeline tied to the national identity authority.
3. Report to the central bank on a defined schedule, in defined formats, with defined reconciliation rules.
Everything else was secondary. Features, aesthetics, roadmap aspirations, none of them mattered if any one of those three failed. The shape of the commitments, the named cast, and the technology stack are summarized on the next page.
Program at a Glance
The numbers, phases, methodology, and named cast at a single view. The technology architecture follows on the next page. This is the reference every subsequent post in the series traces back to.
Headline Commitments
Phases
Methodology Mix
| Workstream | Approach | Cadence / Gate |
|---|---|---|
| Mobile app (iOS, Android) | Scrum | 2-week sprints, Jira |
| Microservices platform | Scrum + PI planning | Quarterly PI, Jira |
| Core banking integration | Waterfall | Stage gates, MS Project |
| Regulatory reporting | Waterfall | Milestone-based, MS Project |
| Infrastructure build-out | Waterfall | Milestone-based payments |
| Program governance | Hybrid | Fortnightly program review, monthly steering |
Named Cast
Nineteen Services, One Integration Surface
Falcon was built on standard, industry-available technology. Nothing on this page is exotic. The stack was selected deliberately to avoid being the first customer for any single layer. That choice cost some architectural elegance. It bought stability in delivery. Each layer is listed with its purpose, not just its components.
| Layer | Purpose | Technologies / Capabilities |
|---|---|---|
| User Channels | Customer interaction points | Mobile App (iOS, Android), USSD Gateway, SMSC |
| Notifications | Outbound engagement, no PII | Firebase Push (tokenized) |
| Edge and API Management | Security, policy, traffic control | Apigee, OAuth 2.0, REST APIs, mTLS |
| Security and Compliance | Cross-cutting data protection and regulatory enforcement | Tokenization at gateway, AES-256 at rest, TLS 1.3 in transit, Central Vault, PCI DSS scope isolation, PII residency per market |
| Business Services | Domain-aligned banking capabilities | Payments, Transfers, KYC, Reporting (19 microservices) |
| Event Streaming | Immutable audit trail, analytics, regulatory feeds | Apache Kafka |
| Async Messaging | Service-to-service processing and callbacks | RabbitMQ |
| Data and State Management | Persistence, caching, distributed consistency | PostgreSQL (system of record, ACID, backups, regulatory retention); MongoDB (document store, KYC artifacts); Redis (ephemeral cache, sessions, TTL < 24h); Hazelcast (distributed grid, locks, idempotency, rate limits) |
| Platform and Infrastructure | Runtime, scalability, availability | Kubernetes, Docker, HAProxy (active/active), Cloud (in-country, multi-region) |
| Observability (Run) | Production monitoring, logs, traces | Splunk (logs), Dynatrace (APM), Zipkin (tracing) |
| Quality (Build) | Code quality and delivery governance | SonarQube, CI/CD pipelines |
| Delivery and Governance | Program execution, tracking, collaboration | Jira, MS Project, Confluence, Slack, Miro |
Kafka and RabbitMQ coexist by design. Kafka carries event streams and audit trails (transaction events, central bank reporting feeds, analytics pipelines). RabbitMQ carries service-to-service async work (notifications, SMS dispatch, KYC callbacks). A unification debate will surface at month 4 (Post 12).
Redis and Hazelcast serve different needs. Redis is the application cache and session store (short-lived, single-node patterns). Hazelcast is the distributed in-memory grid for cross-service state that must be consistent across all 19 services (distributed locks, rate-limit counters, idempotency keys).
Approval Without Ownership
Everything summarized on the dashboard, the mandate, the commitments, the cast, the technology, was in place when the program manager walked into Atlas Bank on 15 January 2025. The brief received across the first forty-eight hours contained roughly what you have just read.
What it did not contain was a sponsor.
Atlas Bank had been talking about Falcon for fourteen months. The board had voted. The budget had been allocated. The PMO, under Ahmed Hassan, had been standing up the governance framework for six weeks. And yet, on the morning of Day 1, when the program manager asked Hassan for the program business case, a 47-page document arrived by email within the hour. It was well-structured. Strategic rationale, financial projections, risk summary, proposed timeline, technology stack overview. Page seven is where the problem surfaced. The sponsor field was blank.
Not missing. Not marked as to-be-confirmed. Blank. As if the field had been deliberately left empty, hoping no one would notice.
A blank sponsor field is rarely an oversight. Somebody, at some point, made a conscious decision to leave it empty. Usually because naming a sponsor would have forced a conversation the organization was not ready to have, about who exactly was on the hook if the program failed. Blank fields are where accountability goes to die.
Phone calls across the afternoon confirmed the shape of it. Two board members had separately told the PMO they supported the program. One had introduced it at the board meeting that approved the budget. The other had championed the business case through finance. But neither had been formally designated as program sponsor. Neither had agreed to own escalations. Neither had committed to sign the program charter.
By the end of Day 1, the picture was clear. Atlas Bank had approval without accountability. The program had a budget, a timeline, and a team. It did not have a sponsor.
Stop Everything Until Sponsorship is Named
The temptation in week one is to start building things. Stakeholder registers. Risk logs. Draft charters. Architecture decision records. It looks productive. It signals energy and competence to the PMO and the executive team. And most of all, it avoids the uncomfortable conversation about who is actually on the hook for this program.
The program manager chose not to start anything until the sponsor question was resolved.
That is a decision that looks obvious written down. In practice, it is hard, and every senior program manager has made the opposite decision at least once in their career and paid for it. When new in a role, the pressure to demonstrate motion is intense. The PMO wants to see a Gantt chart. Executives want an updated risk view. Holding the line and saying "I am not starting planning until we have a named sponsor" sounds bureaucratic, even defensive, to people who do not yet trust the new PM's judgment.
So the two weeks were framed differently. They were called "governance readiness". A daily fifteen-minute check-in was scheduled with Ahmed Hassan so the rhythm of the program stayed visible. And the time was used for the conversations that actually needed to happen.
A program sponsor is not the same thing as a program champion. Champions advocate. They speak well of the program in meetings, secure budget, introduce the PM to people. Sponsors decide. They make resource allocation calls, absorb political risk, unblock escalations that cannot be resolved at program level, and publicly own the outcome when things go wrong.
A program with a champion and no sponsor will stall the first time it needs a decision someone has to stake their reputation on. That stall is usually invisible until the moment it matters, which is always the worst possible moment.
What Makes a Viable Sponsor
The Standard for Program Management is precise on this point. The program sponsor is the individual or group who authorizes the program, provides resources and support, and champions the program within the organization. This is an active role, not an honorary one. It requires authority, alignment, and appetite for accountability, in that order.
Twelve senior stakeholders at Atlas Bank had some level of proximity to Falcon. Over the next two weeks, the task was to map which of them actually had all three of the following:
- Organizational authority to make resourcing decisions without seeking further approval from above
- Strategic alignment between their own mandate and the program's stated outcome
- Appetite for accountability, meaning a genuine willingness to own failure as well as success
Most of the twelve had one or two of these. Very few had all three. That was the filter.
The program manager did not ask an AI to pick the sponsor. That decision sits with people, in rooms, with consequences. The AI was used as a structured-thinking aid: to classify observations from seven one-on-one conversations against a defined framework, surface the pattern in the data, and flag risks in any recommendation. The output was advisory input to the PM's judgment, not a substitute for it. Every classification was reviewed by the PM before acting on it.
To structure the classification and make it defensible, the prompt on the next page was used.
Stakeholder Power Mapping
After one-on-one conversations with seven of the twelve stakeholders over Days 2 to 6, the following prompt was used to structure the classification and surface the clearest sponsor candidates. The prompt is reusable. Any program manager walking into a program without a named sponsor can adapt it.
The prompt below follows the CO-STAR structure: Context (who is asking and why), Objective (what decision the output informs), Style (advisor, not oracle), Tone (neutral, evidence-weighted), Audience (the PM, not the stakeholders themselves), and Response format (a table plus a short narrative). This framework suits classification-and-recommendation tasks because it forces the prompt author to specify not just the question but the shape of the answer, which is what separates useful AI output from generic advice.
What the prompt produced when the seven stakeholder profiles were fed into it is shown on the next page.
Classification and Recommendation
Based on the seven stakeholder profiles provided, here is the classification table followed by the recommendation narrative.
| Stakeholder | Authority | Alignment | Accountability | Recommended Role | Rationale |
|---|---|---|---|---|---|
| Fatima Idris CDO |
HIGH | HIGH | HIGH | Program Sponsor | Digital transformation is her mandate; previous program owner with public accountability. |
| Abebe Alemu CTO |
HIGH | HIGH | MEDIUM | Steering Committee | Strong technology authority but historically defers accountability to business owners. |
| Samuel Osei Head of Retail |
MEDIUM | HIGH | HIGH | Steering Committee | Strong business ownership but authority constrained to retail P&L. |
| Nadia Benali CRO |
HIGH | MEDIUM | HIGH | Steering Committee | Risk authority is central but alignment is regulatory, not transformational. |
| Amara Okonkwo Head of Compliance |
MEDIUM | MEDIUM | HIGH | Informed Stakeholder | Critical for central bank sign-off, not structurally positioned as sponsor. |
| Ahmed Hassan PMO Lead |
LOW | HIGH | HIGH | Informed Stakeholder | Operational ownership; does not hold executive authority for sponsorship. |
| CFO (unnamed) |
HIGH | MEDIUM | MEDIUM | Steering Committee | Budget gatekeeper; alignment is fiduciary, accountability appetite uncertain. |
Recommendation and Risk
The output gave a structured view of the landscape in under ten minutes. It also did what a good advisory output should do: it named the risk in its own recommendation. The first conversation with Fatima Idris was not about whether she would accept the sponsor role. It was about what sponsorship would actually cost her in time and political capital, and whether she was willing to pay it.
The answer came on 28 January 2025 (Day 14). Idris accepted the role on the terms documented, including the four-hour critical-escalation response commitment and the monthly steering cadence. That acceptance unlocked the full governance stack and made charter approval possible.
Program Business Case, Version 1.0
The full Program Business Case submitted for charter approval on 1 February 2025 (Day 17), after the two-week governance readiness exercise was complete. The sponsorship section that was previously blank is now populated. Every figure, date, and name below carries forward into the rest of the series.
Project Falcon will deliver a national digital banking platform across Atlas Bank's three operating markets. The platform consists of a smartphone application and a USSD channel, both integrated with the existing core banking system. Scope includes KYC and AML pipeline integration, central bank regulatory reporting, and parallel operation with legacy branch channels during a controlled rollout. Total investment: USD 14.2M over 18 months. Target: 100,000 active users within 90 days of go-live on 14 July 2026.
A direct competitor launched a comparable platform in August 2024. Market share loss is accelerating in the under-35 demographic. The central bank directive on digital financial inclusion requires all licensed banks to provide a compliant digital channel by Q4 2026. Without Falcon, Atlas Bank faces regulatory non-compliance and continued competitive erosion.
Hybrid delivery model. Agile Scrum (2-week sprints) for mobile application and microservices workstreams, coordinated through Jira and quarterly PI planning. Waterfall with stage gates for core banking integration, regulatory reporting, and infrastructure build-out, coordinated through MS Project. Program-level governance operates on a fortnightly program review and monthly steering rhythm. Program documentation is held in Confluence. Program-wide communication runs through Slack, with dedicated channels per workstream and a cross-cutting steering channel. Workshops and stakeholder mapping sessions are run in Miro.
| Category | Allocation | Amount (USD) |
|---|---|---|
| Vendor contracts (platform integrator, design, audit) | 40% | 5,680,000 |
| Internal headcount (18-month allocation) | 25% | 3,550,000 |
| Infrastructure and licensing | 20% | 2,840,000 |
| Contingency and regulatory | 15% | 2,130,000 |
| Total program budget | 100% | 14,200,000 |
Three-year benefits projection (post go-live): net new revenue of USD 18M, cost-to-serve reduction of USD 9M across retail channels, regulatory compliance value (not quantified). Break-even month: 28 post go-live.
| # | Risk | Likelihood | Impact | Response strategy |
|---|---|---|---|---|
| R01 | Central bank regulatory requirements finalize mid-build | High | High | Continuous engagement via Compliance lead (Okonkwo) |
| R02 | Prime vendor under-delivers at a milestone | Medium | High | Penalty clause in contract, schedule buffer in Phase 2 |
| R03 | Mobile operator API instability in at least one market | High | Medium | Early integration testing, per-market fallback plan |
| R04 | KYC pipeline depends on national ID authority SLA | Medium | High | MOU with SLA commitments, reconciliation queue design |
| R05 | Scope expansion pressure from retail business | High | Medium | Change control through sponsor and steering |
- 100,000 active users within 90 days of go-live (14 Jul 2026)
- Zero critical production defects in the first 30 days
- Central bank certification achieved before launch, formal sign-off letter received
- 99.9 percent transaction availability in the first quarter after launch
- Sub-three-second transaction response time at the 95th percentile
The business case above was submitted for charter approval on 1 February 2025. Every figure, every named stakeholder, every commitment becomes a reference point across the remainder of the program. When the EVM numbers are read in a later post, the $14.2M anchor on this page is what the earned-value math traces back to. When the go-live readiness review is described, the success criteria listed here are the specific metrics reviewed against.
A program without a populated business case is a program without a benchmark. Everything that happens later is measured against what was promised here.
What the Two Weeks Bought
Fatima Idris agreed to the sponsor role on 28 January 2025 (Day 14). She countersigned the program charter on 2 February 2025 (Day 19). The formal kick-off meeting took place on 7 February 2025 (Day 24).
The two-week delay at the start was real, and it was visible. The PMO flagged it on the weekly status report. One board member asked, during an informal conversation, why planning had not started. The explanation held: the governance foundation was being built, and commencing planning without a named sponsor would have created a program that could not be governed. Running a $14.2M program where nobody was willing to sign for it was not an option the new program manager was prepared to accept.
That explanation held. Nobody pushed back further.
At month seven, Falcon hit a critical vendor dispute. Dmitri Volkov, the Platform Integrator Delivery Lead, had missed two consecutive milestones and was claiming scope ambiguity as the cause. His delivery team was pushing for a three-month schedule extension and a 12 percent contract increase, approximately USD 682,000 on the integrator's $5.68M portion of the program budget. The dispute required a decision on whether to enforce the contract penalty clause, renegotiate the schedule with financial consequences, or accept the delay without penalty.
Fatima Idris made that call in a single steering meeting on 18 August 2025. Because her authority was documented in the business case and the charter, her decision was unambiguous and binding. The penalty clause was invoked, with a revised schedule offered as a commercial path forward. The vendor accepted within 48 hours. Delivery resumed within a week.
That moment would have taken four to six weeks to resolve without a named sponsor, because every escalation path would have led to a committee that had never agreed on who owned it. The two weeks spent at the start of the program paid back in full on that single decision, and there were other decisions like it.
What Changed in the Business Case
Version 1.0 of the business case differed from the draft received on Day 1 in three important ways beyond the sponsor field. These are the details that get skipped in most governance templates and cost the most when they are missing.
- The escalation path was documented explicitly with named roles and response-time commitments, not titles and aspirational availability. The sponsor would respond to critical escalations within four hours. That number was in the document.
- The sponsor's authority ceiling was stated in specific numbers, so the team knew exactly what decisions could be made at program level and what required board escalation. Clarity on authority ceilings prevents about half the political friction on a large program.
- The sponsoring body was named as the Executive Digital Transformation Committee, with a defined meeting rhythm and a defined decision log, rather than a generic reference to "executive leadership".
None of these required renegotiating scope or budget. They required sitting in rooms with people and asking direct questions about accountability that most programs avoid asking until it is too late.
A fictional case study for teaching purposes. Atlas Bank, Project Falcon and all named individuals are invented. Technologies are industry-standard and publicly available.