Phase 11 Secondary-Evidence Substitution Matrix¶
Purpose: Define exactly where current official/credible online evidence may close an entry assumption and where only primary pilot or external approval evidence is valid.
Evidence classes¶
- SECONDARY-OFFICIAL — current regulator, government, statistics, local-authority or official provider documentation.
- SECONDARY-CREDIBLE — credible institutional research used only when official data is incomplete.
- OWNER-DECISION — bounded product/operations decision made by the accountable owner from available evidence.
- EXTERNAL-APPROVAL — certificate, authorization, signed/provider/legal approval that must actually exist.
- PRIMARY-PILOT — consented real participant/field/operational evidence.
- SYSTEM-METRIC — measurement from the approved real pilot runtime.
What secondary evidence may legitimately close¶
| Decision/gate | Acceptable evidence | Phase 11A status |
|---|---|---|
| Zambia law/regulator applicability questions | current DPC/BoZ/CCPC/other official sources + conservative interpretation | Research complete; final qualified legal sign-off remains external |
| Controller filing direction | DPC categories + known operating structure + owner decision | Individual-controller filing path selected |
| Need for overseas storage/transfer authorization | DPC official registration/FAQ materials | Determined applicable to current overseas topology; authorization itself pending |
| Pilot geography | census/local-government/electoral/operational sources + owner decision | Kabwata + Chilenje approved for entry |
| Comparison geography | official local-government/constituency evidence | Matero evaluated as later stress/expansion candidate |
| Limited categories | official training/registry/licensing sources + implemented evidence architecture | Four categories approved for entry |
| Cohort cap/wave design | risk, support, verification and architecture capacity assumptions + owner decision | 24 providers/60 customers in 3 waves |
| Provider pathway mix | existing domain model + inclusion objective | target 2/2/2 per category |
| Support hours/escalation | operating capacity decision | Approved |
| Numeric pause/stop thresholds | conservative risk-based owner decision | Approved as pilot control limits |
| Device/network test matrix | Android support range + Zambia ICT heterogeneity evidence | Approved |
| Initial provider minimization | architecture/fallback capability + privacy/risk evidence | Maps/Sentry/payments/production comms disabled Wave 1 |
| Authentication provider direction | official Firebase support/privacy/security/region docs + existing GCP/Firebase boundary | Firebase phone OTP selected direction; activation pending |
| Consent/retention baseline | DPC principles + data-minimization architecture + owner decision | Product-approved baseline; legal/DPC review pending |
What secondary evidence can never substitute for¶
| Required evidence | Why it cannot be desk-researched into existence | Current status |
|---|---|---|
| DPC controller registration certificate/reference | regulator-issued evidence | Pending |
| DPC overseas transfer/storage authorization | regulator-issued authorization | Pending |
| Qualified Zambia legal sign-off | professional review of exact product/data/terms | Pending |
| Final processor contract/DPA approval | must bind the actual provider/account/topology | Pending for real pilot providers |
| Participant informed consent | individual act tied to exact notice/version | Not started |
| Provider onboarding completion/abandonment | real user behavior | Not started |
| Evidence rejection/resubmission rates | real evidence workflow | Not started |
| Trust-language comprehension | participant understanding cannot be inferred from copy alone | Not started |
| Search/no-result behavior | real cohort/system behavior | Not started |
| Enquiry response/unanswered rate | real provider/customer interaction | Not started |
| Contact-consent comprehension/revocation | real interaction behavior | Not started |
| Field verification time/cost/safety | real Zambia field operations | Not started; field claims disabled |
| Real device/network crash/upload/recovery rates | requires real hardware/network conditions | Not started |
| Support/moderation workload | requires real cases | Not started |
| Willingness to pay/price preference | requires participant evidence; desk pricing is not demand validation | Not started |
| Unit economics | requires measured pilot cost/time plus demand evidence | Not started |
| STOP/REPEAT/NARROW/PROCEED exit decision | must synthesize actual Phase 11 evidence | Pending |
Rules for progression¶
- Official online evidence may close an assumption or support an owner decision when the question is factual, regulatory, infrastructural or scope-setting.
- It may not be relabelled
PRIMARY-PILOTmerely because no participants are yet available. - Where official sources establish that an authorization is required, the gate becomes clearer but remains closed until the authorization is actually evidenced.
- A product decision may proceed conservatively by disabling an optional provider/feature; this can remove that provider from the entry critical path.
- The absence of Maps, Sentry, payments or production communications does not block Wave 1 because DIREKT has manual/fallback paths and those providers are explicitly disabled.
- Real participant processing cannot start until the remaining external-approval gate is satisfied.
- Phase 11 cannot close until required primary/system evidence is produced or the exit decision is
STOPbecause the pilot cannot lawfully/safely proceed.
Result¶
Online and prior research can materially advance Phase 11A by converting vague unknowns into explicit decisions and a very small set of external blockers.
It cannot truthfully make 11C–11H “complete” before a real pilot exists. Those stages remain evidence-producing stages rather than documentation exercises.