Northern cotton activity
The programme annex attributes northern cotton-sector activity to SODECOTON. That supports considering a northern sourcing option.
Does not establishA site rating, a distance or any access for GFT.
GFT instruments / Location decisions
The right site
depends on the brief.
Compare the Douala-Dibamba corridor with the northern cotton belt. Make each priority explicit, challenge the ratings and find the weight that changes the answer. Illustrative inputs throughout.
Operate the instrumentProcedural visualisation. Schematic, not a facility.
Live, illustrative
CorridorNorthCorridor ahead by 13.35
Result unavailable. Correct the inputs.Controls ↑An instrument for the decision
Douala-Dibamba corridor or the northern cotton belt?
Change the inputs. Read the result. Keep the assumptions visible.
Illustrative workshop
These are editable preference scores, not measured site conditions. The default ranks the corridor at 66.80 against the north at 53.45 points out of 100. A higher score is a preference within this model, never permission to invest.
B / Current scenario
Douala-Dibamba corridorpreference points / 100
Northern cotton beltpreference points / 100
Corridor preferred13.35 points ahead
Read the decision
The corridor leads under these priorities and benefit ratings. Check whether port, customer and hub benefits have been counted independently, then test the cotton trade-off.
Schematic, not to scale. The profile shows ratings before weighting; the beam shows the weighted difference. A high score cannot compensate for a failed requirement such as land rights, water or safe working conditions.
Net gap carried down
Assumption register
These defaults are a workshop, not a survey.
No site ratings or weights were supplied in the programme evidence, so there is no sourced GFT preset. Every perspective and rating here is illustrative.
| Criterion | Share | C / N | Effect on corridor minus north |
|---|---|---|---|
| / | |||
| / | |||
| / | |||
| / | |||
| / | |||
| / | |||
| / |
Signed contributions reconcile exactly to corridor minus north, before display rounding. Positive favours the corridor; negative favours the north.
Read every basis and benefit anchor
- Power reliabilityIllustrative. Illustrative preference for continuity; replace using outage logs and the designed backup system.0: unacceptable interruptions. 100: the required continuity is achieved.
- Port accessIllustrative. Illustrative corridor preference; no measured travel time or freight quote is assumed.0: fails the dispatch window. 100: meets the dispatch window at the agreed cost.
- Logistics to populationIllustrative. Illustrative access to customers; validate against the actual sales and distribution plan.0: fails the delivery promise. 100: meets it at the agreed delivered cost.
- Cotton proximityIllustrative. Illustrative northern sourcing preference. R03 is context only, not a distance measurement.0: unacceptable inbound supply burden. 100: meets the agreed supply target.
- Wage levelIllustrative. Illustrative cost preference; no regional wage differential is established.0: unacceptable cost per good unit. 100: meets the fair-pay and productivity budget.
- Training baseIllustrative. Illustrative workforce readiness; no named training programme or capacity is assumed.0: cannot staff the planned ramp. 100: verified skills cover the ramp.
- Proximity to Camtext hubIllustrative. Illustrative corridor preference. R10 reports planned Dibamba and Garoua activity, not availability to GFT.0: no useful accessible interface. 100: verified useful access under agreed terms.
Balanced priorities are 25, 20, 15, 20, 8, 7 and 5 points in table order. The export-led illustration changes them to 25, 35, 15, 5, 8, 7 and 5. The cotton-first illustration uses 15, 5, 5, 55, 8, 7 and 5. These are design assumptions that stress dispatch or inbound sourcing. They are not benchmarks or GFT operating results.
A / B decision record
Keep the alternative in view.
A is a frozen snapshot of every input; B follows the controls. A opens pre-loaded with the export-led illustration, so the comparison is never empty. Drag across the radar to wipe from A to B, or save your own A.
No A saved. Use Save current as A.
Gap, corridor minus north
A
B
Saved A: corridor preferred. Current B: corridor preferred. The preference is unchanged.
Moving from A to B changes the gap by -18.75 points.
- A corridor
- 75.05
- A north
- 42.95
- B corridor
- 66.80
- B north
- 53.45
Priority share, A then B
- Power
- Port
- Customers
- Cotton
- Labour
- Training
- Hub
| Scenario | Corridor | North | Corridor minus north |
|---|---|---|---|
| A / Saved | 75.05 | 42.95 | 32.10 |
| B / Current | 66.80 | 53.45 | 13.35 |
The table compares the same calculation on both input sets. The dotted ghost profiles in the instrument show A's original ratings. Nothing is sent to a server.
Formula check
Method, no hidden factors
Normalise, multiply, reconcile, flip.
Four steps, worked on your current inputs. Adjust any control below or on the cards and the stage re-renders live. The method follows HM Treasury and DESNZ options-appraisal guidance, 24 June 2024 ↗. This is an editable illustration, not an accredited appraisal.
Live chart / your inputs
Priority points become shares. Total: 100 points. The arm is bare: seven empty hooks.
Share times score, for each site. The pans load: corridor 66.80, north 53.45.
Signed contributions from one zero line add up to the gap: +13.35 equals 66.80 - 53.45. The arm tilts by that total.
The tilt, unrolled: one share moves, the rest stay in proportion. Criterion: Cotton proximity. The ring is the tie.
x* appears only when a tie sits inside the share range. The criteria are location families for an early workshop, not shortlisted parcels, negotiated tariffs or engineering designs.
Result unavailable. Correct the inputs to see the worked steps.
- Step 01 · Normalisation
Normalise the priorities.
pᵢ = wᵢ / ΣwⱼEach priority becomes a share of the total. Zero excludes a criterion. Set the anchors before the weights: a priority is the value of moving from the worst acceptable anchor to the best one.
Your inputs:
- Step 02 · Weighted products
Multiply by the benefit score.
Sᴸ = Σ(pᵢ × sᵢᴸ)Scores sit on one preference scale. Zero means unacceptable against an agreed criterion. 100 means the agreed target is met. A score is not an observed percentage.
Your inputs:
- Step 03 · Signed waterfall
Reconcile to the gap.
Sᶜ − Sⁿ = Σ[pᵢ(sᵢᶜ − sᵢⁿ)]Every point of the gap has a cause. The signed contributions add up to corridor minus north, using unrounded values.
Your inputs:
Inspect criterion contribution on the register & mobile - Step 04 · Sensitivity line
Find where it flips.
Δ(x) = x dᵢ + (1 − x) bᵢ
x* = bᵢ / (bᵢ − dᵢ)Vary one share x and keep the rest proportional. The gap is a straight line that may cross zero at x*. Here dᵢ is the corridor-minus-north score for that criterion and bᵢ is the same difference averaged over the others.
Your inputs:
Select criterion for the sensitivity line
Nearest tie share
Where the preference changes
Find the weight that changes the answer.
Nearest flip
14.6 percentage points above the current share. The other priorities stay in proportion.
Preview, not applied. Drag along the line or use the slider. Illustrative, own arithmetic.
- Cotton proximity share
- 20.0%
- Gap at that share
- +13.35
- Winner at that share
- Corridor preferred
Δ(x) = x dᵢ + (1 − x) bᵢ with dᵢ = and bᵢ = .
Each track moves one normalised weight from zero to full priority. Circle: current share. Ochre tick: tie share. An interior tie separates the winners; a tie at an endpoint does not establish a reversal. This is a sensitivity test, not an uncertainty interval.
Gate rule
Boundary, what changes the decision
A site must pass its gates before it can win.
Active lens01 · Non-compensatory gates
Passes all seven gates. The preference score applies: points. Gate status is your assertion; this tool performs no diligence.
Caveat lens: the criteria each caveat touches, with their current shares
- Power
- Port
- Customers
- Cotton
- Labour
- Training
- Hub
Gates first. The seven posts are pass or fail, so a high mark elsewhere cannot buy back a failed gate. Gates first.
These four may share one cause. If they do, trim the duplicated weight before reading the gap. Power, Port, Customers, Hub together: 65.0% of the priority points.
Labour counts only as a benefit score against fair pay, productivity and retention, never as the lowest wage. Labour holds 8.0%; the Safety gate protects the workforce.
Wages, power and freight can move together. Save a pessimistic case as A and compare a whole case in B.
- Decision boundary 01
Strength in one place can hide weakness in another.
An additive score lets a high mark on one criterion offset a low one. Real constraints often cannot be traded, so gates come first.
PowerPortCustomersCotton+3 more100% totalTouched: Power, Port, Customers, Cotton, Labour, Training, Hub
- Decision boundary 02
One advantage can be counted twice.
Power and production losses may be linked. Port access, customer logistics and hub proximity may share a cause. Keep criteria distinct and trim duplicated weight.
PowerPortCustomersHubTouched: Power, Port, Customers, Hub
- Decision boundary 03
Cheap labour is not automatically better.
A wage preference must reflect lawful fair pay, productivity, recruitment and retention.
LabourTouched: Labour
- Decision boundary 04
One weight at a time is a partial test.
It does not capture wages, power and freight moving together, nor uncertainty in the ratings. Save a pessimistic case as A and compare a different complete case in B.
PowerPortLabourTouched: Power, Port, Labour
Preference points cannot be converted to financial return, probability of success or a capital budget. A planning aid, not investment, legal or tax advice.
Provenance beam
Provenance, read in context
The evidence frames the question. It does not score the sites.
Choose a source. The beam shows which criteria it gives context for, with no weight on the pans. Choose a reader to see whose evidence would replace which rating.
Reported Camtext plans
The annex reports Camtext plans involving Dibamba and Garoua, refreshed in 2026. Treat them as reported plans.
Does not establishVerified operating facilities, access agreements, a GFT relationship or a hub available to the proposed factory.
The weighted method
The arithmetic and sensitivity approach are informed by HM Treasury and DESNZ options-appraisal guidance, 24 June 2024 ↗.
Does not establishA facilitated or accredited appraisal. That needs agreed criteria, defensible scales and stakeholder validation.
Lit criteria Cotton proximity. Context only. No source here supplies a site rating, a weight or a distance.
Provenance receipt · 31 Jan 2024 deliverables · 2026 refresh flags
Read the source-open receipt and default-status register ↗GFT's known Cameroon textile programme combines feasibility, performance-management system design and digital strategy. Its deliverables were prepared on 31 January 2024, with source refreshes flagged in 2026. The citations are programme-annex references, not articles reopened for this instrument.
Record blend
Keep the record with the result.
Measure the difference that matters.
A weight-sensitive preference is a useful research question. Replace the most influential assumptions with parcel-specific evidence before committing to a location.
- Blend
- 100% B
- Corridor
- 66.80
- North
- 53.45
- Gap
- +13.35
Corridor preferred at this blend. The same location leads across the whole blend from A to B.
Blended read-outs are illustrative, own arithmetic: each priority and each rating is mixed in proportion t : 1 − t, then scored by the same model.
| Scenario | Corridor | North | Gap |
|---|---|---|---|
| A / Saved | 75.05 | 42.95 | +32.10 |
| B / Current | 66.80 | 53.45 | +13.35 |
A shared scenario preserves priorities, benefit scores and a saved comparator. It records a judgement under assumptions, not a measured site condition or an investment recommendation.
Closing resolution · Carry the record forward
Final balance state and next analytical bench
- Current B · Corridor
- 66.80
- Current B · North
- 53.45
- Current B · Net gap
- +13.35
- Saved A · Net gap
- +32.10
Related papers, programme entries and not released publications: Factories on a Weak Grid (wp07), The Supervisor Gap (wp09), Cotton to Garment Across Africa (wp19).