n.V7-3.08 | When Leadership Fails

01| A system that pushes authority outward has to be equally serious about what happens when the authority is misused or simply not up to the job. Otherwise decentralization is just a way of distributing failure.

02| Problems have to be visible early, which means monitoring is continuous rather than annual. Maveriq satisfaction, project delivery rates, Impact Buyer satisfaction with verified outcomes, transparency compliance, financial health, training completion, validation rates, escalation frequency, ecosystem partner feedback. The dashboards show the same numbers to local and global at the same time. Information asymmetry is a structural risk, not a privilege of rank.

03| Specific patterns start a review automatically: performance below threshold for two consecutive quarters, transparency violations, financial distress signals, a spike in Maveriq complaints, concerns from ecosystem partners. The triggers are defined in advance and applied the same way everywhere, so that a review is never a signal that someone has fallen out of favour.

04| The response runs on four levels.

05| Level 1 — Advisory. Performance is sliding but not critical. Global helps diagnose it, brings in what other CITIs did about the same problem, coaches the local leadership, and shares resources if the pressure is temporary. Ninety days to show a trajectory. Most dips end here, and most of them are not leadership problems at all.

06| Level 2 — Direct intervention. The gaps are serious and sustained. A global team takes temporary control of the specific functions that are failing while local leadership stays in place under oversight. Corrective plan, milestones, tighter reporting. Six months to show improvement that holds, not a good quarter.

07| Level 3 — Leadership change. The pattern points at leadership rather than at the CITI. The leadership is replaced and the CITI keeps operating. Interim management comes from Global or from another CITI's experienced people, recruitment for a permanent local replacement starts immediately, and the whole point of the design is that Maveriqs and ecosystem partners barely feel it. Target: three months.

08| Level 4 — Termination. The CITI cannot be repaired. The triggers are narrow: corruption, irreparable brand damage, sustained failure across multiple interventions, exploitation of Maveriqs. Due process first — the CITI answers the findings, the evidence is documented — and the Board approves. Wind-down runs six to twelve months with explicit transition support.

09| These are not four steps of punishment. They are four responses to four different kinds of problem, and the architecture exists so that we can act decisively when something is genuinely wrong without over-reacting to every bad quarter.

10| The body that runs this cannot be the body that approved the CITI in the first place. That is the standard failure of franchise systems: the parent enforces quality on franchises it selected, so admitting the failure means admitting its own earlier mistake, and the admission arrives late every time. Ours is independent by construction — five to seven members, Maveriq representatives drawn from several CITIs so no single local context can capture it, one GMT representative, one Board representative, and outside experts to break the echo chamber. It can halt new launches when quality is slipping across the network, mandate intervention levels, and recommend termination, with the Board approving that last step. It reports publicly on network health.

11| The same failure pattern has an internal version, and it is slower and harder to see. Professional staff — Pro.Crowd — accumulate power over mission by increments. Knowledge becomes gatekeeping. Process becomes a moat. Metrics drift from measuring outcomes to measuring staff activity. The endpoint is the classic NGO: seventy percent overhead, thirty percent delivery, and by the time it is obvious it is close to irreversible.

12| The drift has a recognizable grammar. We serve them becomes we know what is best for them, justified by they do not understand yet, so we will decide for now, justified by we are protecting them from mistakes. Each step increases complexity, complexity increases the professionals' importance, and importance makes the next step easier.

13| So: Pro.Crowd exists to serve the platform, and the platform exists to serve the Maveriqs. That direction does not reverse. Compensation ranges are public and the ratio to the Maveriq average is published, so status cannot be built quietly through pay. Maveriqs rate Pro.Crowd performance quarterly, and the ratings feed compensation and contract renewal. Pro.Crowd advise; they do not approve, and they cannot veto a Maveriq's decision about a Maveriq's project. There are no closed-door Pro.Crowd meetings. Roles rotate. Growth in permanent staff beyond what the work demands is treated as a failure signal, not an achievement.

14| And there is a floor beneath all of that: a two-thirds majority of Ikosi can trigger a full review of Pro.Crowd, which the Board investigates and can end in replacement or restructuring. The bar is deliberately high, because operational chaos serves nobody. Its existence is the point. People behave differently when the population they serve holds the deepest authority and everyone knows it.

15| When a CITI does close, the obligations do not close with it. Maveriqs get transition support: connections to CITIs that can absorb them, certification of the skills they built so the work transfers, and financial bridging. Data is preserved for learning. Partners, governments and Impact Buyers are told what is happening and why, in time to adjust their own plans. The licence is revoked and the name protected legally. The process is slow and expensive, and we accept both, because a franchise system that terminates carelessly does itself more damage than any failing franchise could.

16| Every failure gets analysed and the post-mortem is published. Why this one failed, which signals we missed, what could have been caught earlier. Selection criteria get tightened when a pattern shows in who fails. Leadership training gets rewritten when the root causes repeat. Warning thresholds get retuned when a signal turns out to have predicted the collapse months before our threshold noticed. The non-negotiables get extended when an analysis reveals a hole in the floor that should have been covered. Failure is expensive, and the only way to recover any of that cost is the learning. A failure that produces no learning has been paid for twice.

17| Last, and most important: a CITI closes for exactly two reasons. The barrio loses its urgency — residents stop showing up, stop pushing, stop wanting it — and the invitation we operate under has been withdrawn. Or the buying power collapses, sponsors and Impact Buyers stop funding, and the engine has no fuel however much the barrio still wants it. Nothing else is a legitimate cause. A tool going red is a broken tool, and we fix tools. Confusing a broken tool with a broken platform is how organizations abandon places that were still worth being in.

18| Everything in this book part — the levels, the triggers, the escalation right, the four interventions, the independent enforcement — exists to protect one promise that is easy to state and hard to keep. We are not a project that visits. Trust in a barrio is not built by impact; it is built by permanence. The machinery is what allows us to still be there next year and the decade after, and to be worth having there.