Public-data prototype — Public-data prototype. Built only from openly published sources. Not an official statistic, not investment advice and not a diagnosis of any business. About the prototype

Sector Observatory — method

Method and limits

This page states how records enter the terminal, how they are labelled, and what this Observatory will not claim. It is a public-data method under test, not an official statistic.

Method: Observatory method v0.1

Method version

Every value on every page carries this version.

Observatory method v0.1

SADC Region
Country
Industry
Subsector
Indicator
Evidence
Insight
Opportunity or Risk
Participant Effect
Decision

Content states

What a record is. These are never blended together in a view.

State	Meaning
Observed fact	Recorded exactly as a named public source states it.
Calculated result	Derived from recorded values using a stated formula.
Analyst interpretation	A professional reading of the evidence. Not approved.
Confirmed insight	An interpretation confirmed by an authorised human reviewer.
Opportunity signal	A possible opening indicated by evidence. Not a forecast.
Risk signal	A possible pressure indicated by evidence. Not a prediction.
Approved publication	Cleared by an authorised human approver for public release.

Release states

Nothing is published before human approval.

State	Label
DRAFT	Draft — human review required
IN_REVIEW	In review
APPROVED_PUBLICATION	Approved for publication
WITHDRAWN	Withdrawn

Any AI-assisted explanation stays marked Draft — human review required until a named reviewer approves it.

Confidence

How firmly a value is established.

Level	Meaning
HIGH	High confidence
MODERATE	Moderate confidence
LOW	Low confidence
UNVERIFIED	Unverified

Status scale

Missing evidence is never shown as zero, Green or Red.

GreenAmberRedInsufficient EvidenceNot ApplicableConflicting Evidence

Insufficient Evidence means no registered source supports a reading. Conflicting Evidence means two published sources disagree and the conflict is unresolved. Not Applicable means the measure does not apply to that country or industry.

Indicator dictionary

Every indicator configured in the terminal, with its unit and definition.

Indicator	Unit	Definition
Production volume	physical units	Output produced in a stated period as reported by a named source.
Export value	currency	Declared value of exports for the stated period.
Employment	people	Reported persons employed in the sector for the stated period.
Capacity utilisation	%	Share of installed or licensed capacity in use.
Input cost pressure	index or %	Movement in the cost of critical inputs.
Price index	index	Published price index relevant to the sector.
Installed capacity	MW	Installed generation capacity as published.
Supply reliability	hours or %	Published measure of interruption or availability.
Tariff level	currency per unit	Approved tariff for the stated period.
Project pipeline	count or value	Publicly announced projects at a stated stage.
Corridor throughput	tonnes or TEU	Volume moved through a stated corridor or terminal.
Transit time	days or hours	Published average time for a stated movement.
Credit growth	%	Change in credit extended to the private sector.
Policy rate	%	Central bank policy rate at the stated date.
Non-performing loans	%	Published NPL ratio.
Connectivity coverage	%	Published population or area coverage.
Visitor arrivals	people	Published arrivals for the stated period.
Occupancy	%	Published accommodation occupancy rate.
Retail volume	index	Published retail trade volume measure.
Import dependence	%	Share of supply met by imports.
Enrolment	people	Published enrolment for the stated period.
Skills gap	narrative or %	Published shortage measure for stated occupations.

From source to revision

The chain every published value follows. A value that has not completed it is shown with its release state, never as approved.

1Source registered
A named public source is registered with its publisher, URL, licence and retrieval date.
2Evidence captured
The value is recorded exactly as the source states it, with period, unit and coverage.
3Method checked
Definitions, comparability and any calculation are checked against the current method version.
4Interpretation drafted
An analyst (or an AI-assisted draft) states what the evidence may mean. It is marked Draft.
5Human review
A named reviewer checks source, method, wording and limits.
6Approved publication
An authorised approver releases it with approver, date and method version.
7Scheduled verification
A review date is set; the source is rechecked on schedule.
8Revision or withdrawal
Changed or contradicted values are revised or withdrawn with a visible record.

Approved source types

Categories of source a value may be drawn from. Individual sources are named on the record, not here.

National statistics offices
Central banks
Ministries and government departments
Regulators
Regional bodies and secretariats
Multilateral development banks
International organisations
Development finance institutions
Sector associations
Listed-company disclosures
Tender and procurement portals
Regulatory filings and gazettes
Trade and customs publications
Company registries
Published research and working papers
Reputable news reports (recorded as reports, not facts)
Published project documents
Published survey results
Published labour and skills data

Regional statistics rules

How a regional figure may appear, and the fields it must carry. Currently no regional figure is approved; the register is empty by design.

1.A regional figure is shown only when it is computed from approved country records with a stated aggregation method, or quoted directly from a named source.
2.Every value carries indicator, value, unit, currency, period, country coverage, source, source URL, retrieval date, last review date, confidence, method version, aggregation method, release state, comparability flag, missing-country count and analyst note.
3.A value with missing member states states how many are missing; it is not presented as a regional total.
4.Values from sources that define the indicator differently are flagged not comparable and are not combined.
5.Missing evidence is displayed as Insufficient Evidence. It is never displayed as zero, Green or Red.
6.Nothing is forecast. Nothing is estimated to fill a gap.

Regional indicators configured

Indicator	Unit	Kind	Definition
Regional GDP	USD, current prices	Sourced	Sum of member-state GDP for one reference year, in one currency basis, with the aggregation method stated.
Population	persons	Sourced	Sum of member-state population estimates for one reference year.
GDP per capita	USD per person	Sourced	Regional GDP divided by regional population for the same reference year and source basis.
Real GDP growth	% year on year	Sourced	Weighted real growth across member states with weights and base year stated.
Intra-SADC trade share	% of total trade	Sourced	Trade between member states as a share of member states' total trade, one period, one source basis.
Total exports	USD	Sourced	Sum of member-state goods exports for one period, source and valuation basis stated.
Total imports	USD	Sourced	Sum of member-state goods imports for one period, source and valuation basis stated.
FDI inflows	USD	Sourced	Sum of recorded foreign direct investment inflows for one period.
Industry contribution to GDP	% of GDP	Sourced	Share of one configured industry in member-state GDP, per country, before any regional statement.
Employment and unemployment	% of labour force	Sourced	Recorded employment and unemployment rates, per country, with the labour-force definition stated. Not aggregated across differing definitions.
Inflation	% year on year	Sourced	Consumer price inflation per country and period. Regional figures require a stated weighting method.
Exchange rates	national currency per USD	Sourced	Official or market reference rate per country, date and source stated. Never averaged across currencies.
Electricity access	% of population	Sourced	Recorded population share with electricity access, per country and year.
Mobile and internet penetration	% of population	Sourced	Recorded mobile subscription and internet-use rates, per country and year, definition stated.
Transport corridor performance	as defined by source	Sourced	Recorded corridor transit time, dwell time or throughput from a named corridor authority or study.
Project pipeline value	USD	Sourced	Sum of published project values in the capital and project register, by status, once that register exists.
Capital raised	USD	Sourced	Sum of published capital raised for registered projects and companies, by instrument, once that register exists.
Jobs linked to projects	positions	Sourced	Published job figures attached to registered projects, stated as announced or as verified, never combined.
Opportunities under investigation	signals	Derived count (not a statistic)	Count of approved opportunity signals in the terminal's own register.
Material risks recorded	signals	Derived count (not a statistic)	Count of approved risk signals in the terminal's own register.

Regional Integration Monitor rules

Eleven themes. Each reads Insufficient Evidence until a specific, sourced, cross-border record is approved.

Theme	Question
Cross-border trade	Which goods and services move between which member states, and how has that changed?
Regional value chains	Where are stages of production split across more than one member state?
Transport corridors	Which corridors carry regional trade and where is capacity or delay recorded?
Energy interconnection	Which power and fuel links connect member states, with what recorded capacity?
Digital connectivity	Which cross-border digital links, payment rails and data corridors are recorded?
Investment flows	What intra-regional and inbound investment is recorded, by whom and where?
Labour and skills mobility	What recorded movement of workers and skills crosses member-state borders?
Industrial complementarity	Where do member states' industrial structures supply or compete with one another?
Development plan alignment	Where do national plans and regional strategies name the same priority?
Trade barriers	Which tariff and non-tariff measures are recorded between member states?
Implementation constraints	What recorded constraints slow agreed regional measures?

A record must name at least two member states and one public source. Membership of a bloc, shared geography or an announcement is not a record of a relationship.

Participant lens rules

Thirteen lenses read the same evidence. The lens changes wording, never the evidence, its confidence or its release state.

Investor
Lender or development fund
Business owner
Entrepreneur
Employee
Job seeker or student
Supplier
Exporter
Importer
Government or institution
Researcher or media
Consumer or public
Community or development stakeholder

Field	How it is filled
What is changing	Quoted from the reviewed signal statement. The same wording for every lens.
Why it matters to this participant	An analyst-written lens statement attached to the signal. Missing lens statements read Insufficient Evidence.
Possible effect	Taken from the reviewed effect insight for the same country and industry. Drafts and AI-assisted explanations are labelled and never shown as approved.
Opportunity requiring investigation	Shown only when the signal is registered as an opportunity. An opportunity is a prompt to investigate, not a recommendation.
Material risk	Shown only when the signal is registered as a risk. A risk is a recorded pressure with an evidence state, not a forecast.
Evidence confidence	The signal's confidence grade and the count of linked evidence records. Confidence never rises because a lens was chosen.
Decision requiring consideration	Not published on the public terminal. Decision context for a specific position is provided through subscription or advisory engagement.
What to monitor next	Quoted from the signal's monitoring note: the named source or release that would change the state.

Map boundaries

What the SADC map is, and what it is not.

Country outlines are generalised from Natural Earth public-domain data for on-screen pointing only. They are indicative, not an authority on borders, sovereignty or disputed territories. Small island states carry a marker because their land area is too small to be a reliable pointer target.

Map shading shows evidence coverage in this terminal only: approved records, records in review, unresolved conflicts, or none. It never encodes performance, rank or risk.

Decision questions and where they are answered

The landing page is organised around these questions. Each links to the section built to answer it.

What is happening?
What is happening across the region
Why is it happening?
Industry momentum and pressure
Who is affected?
Participant Impact Explorer
Which signals are verified?
Evidence gaps and conflicts
What are the material risks?
Material risks
Which projects or funding are relevant?
SADC Capital and Project Intelligence
What should be monitored next?
Opportunities requiring investigation

Boundaries we do not cross

Stated so they can be held against us.

No access to government, regulator or banking systems.
No official mandate, authority, endorsement or partnership.
No statistic published without a named, checkable source.
No forecast presented as a fact.
No plan progress, target achievement or delivery performance asserted.
No sector rating where the evidence is insufficient.
No client evidence from the Business Doctor used here.
No investment, legal, tax or professional advice.

How we handle evidence and claims across the platform