Why Africa’s cold-chain problem was never about technology, and why two recent deals might actually matter.
As a rule, I am sceptical of sectors that are described as “huge opportunities” for more than a decade without ever quite arriving. African cold chain is one of them.
For years, the data has barely shifted. Estimates from the FAO and World Bank consistently suggest that around 30–50% of food produced in sub-Saharan Africa is lost before it reaches consumers. For highly perishable products — fruits, vegetables, fish, and dairy — losses are frequently higher, closer to 40–60%, particularly where cold-chain infrastructure is weak or absent and temperature control collapses between farm and consumer.
The proposed solutions range from solar cold rooms, app-based logistics platforms, farmer aggregation schemes, and venture-funded “end-to-end” startups, each promising to reduce spoilage — often with software standing in for infrastructure.
Most failed.
Not because cold chain is unimportant, but because it was repeatedly misframed as a startup problem rather than an infrastructure problem.
Two recent investments suggest that framing may finally be changing:
- A European impact investor committing roughly $19m to a Kenyan cold-storage operator, and
- A West Africa–focused fund making its first deployment into a Nigerian food distributor that quietly built its own refrigerated fleet.
These are not flashy deals. That is precisely why they matter.
What Cold Chain Actually Is (And Why It Breaks)
Cold chain is not a warehouse. It is not a truck. It is not an app.
It is a continuous system that only works if every link functions:
- First mile: cooling immediately after harvest
- Aggregation: pooling volumes without temperature breaks
- Storage: reliable, monitored cold rooms
- Long-haul transport: refrigerated linehaul
- Distribution hubs: cross-docking without dwell time
- Last mile: predictable delivery to retailers, hotels, exporters
If any link fails, the entire chain collapses. And once temperature is lost, it cannot be “fixed” downstream.
This is why cold chain has defeated so many well-intentioned projects and startups. It is an unforgiving system. Partial solutions do not fail slowly; they fail all at once.
Owning the full chain is possible. But it is very expensive.
The Real Economics: CAPEX, Utilisation, Power
The cold-chain dilemma is not mysterious. It is arithmetic.
Cold infrastructure requires:
- high upfront CAPEX (refrigeration, insulation, backup power),
- continuous OPEX (electricity, diesel, maintenance),
- and, above all, utilisation.
A cold room that runs at 30% utilisation is not “inefficient”. It is bankrupt.
This is where most previous models broke:
- assets were built before demand existed,
- volumes were seasonal and volatile,
- and customers were price-sensitive wholesalers operating on razor-thin margins.
Add unreliable power, high fuel costs, and weak contract enforcement, and the economics deteriorate fast.
Cold chain does not fail because the technology is unavailable. It fails because idle cold assets bleed cash every hour they are empty.
Why These Two Models Are Different
The recent deals matter because they divide the problem and define it correctly.
1. The Midstream Utility Model (Cold Solutions Kenya)
The Kenyan operator is not chasing farmers.
It is not onboarding consumers.
It is not “digitising agriculture”.
It sells three things:
- aggregation capacity
- temperature-controlled storage
- reliable refrigerated transport
to exporters, processors, and institutional buyers who already have volume.
This is a midstream utility model.
The value proposition is not margin expansion — it is risk reduction:
- lower spoilage,
- predictable lead times,
- export compliance,
- fewer rejected consignments.
These customers pay because failure costs more than fees.
It treats cold chain like power or water — infrastructure that must work, not a growth hack to be optimised.
2. The Demand-Anchored Distributor Model (DeliFrost Nigeria)
The Nigerian business flipped the logic.
It did not start with cold assets.
It started with customers.
Supermarkets. Hotels. Caterers. Food service operators.
Predictable demand. Clear margins. Contractual volume.
Only after demand was anchored did it invest in:
- refrigerated trucks,
- storage,
- and fleet management.
This is not vertical integration for control.
It is vertical integration to protect utilisation.
Cold chain works when demand is guaranteed first. Everything else is secondary.
Why This Is Different From Previous Failures
Earlier cold-chain startups failed for three recurring reasons:
- They built assets before volume
- They tried to serve farmers who could not pay
- They relied on software to fix physical bottlenecks
Solar cold rooms without aggregation stayed empty.
Farmer-first models collapsed under weak purchasing power.
Marketplaces digitised transactions but not temperature.
These new models invert the sequence:
- volume first,
- assets second,
- optimisation last.
That sequencing matters more than innovation.
The Cold Reality: Why This Still Might Fail
It would be a mistake to declare victory.
Cold chain remains fragile, even under better models.
Failure modes remain obvious:
- Power risk: electricity and diesel costs can erase margins overnight.
- Utilisation risk: demand concentration creates customer dependency.
- Seasonality: food volumes fluctuate faster than fixed costs.
- Credit risk: buyers delay payment, while energy bills do not.
- Policy risk: export bans and price controls disrupt flows instantly.
Cold chain does not tolerate optimism. It punishes forecasting errors.
This is why most successful cold-chain businesses globally resemble utilities or distributors — not tech platforms.
The Broader Lesson
Africa’s cold-chain problem was never about awareness.
It was about misdiagnosis.
The constraint was not lack of innovation.
It was lack of boring, expensive, well-utilised infrastructure.
The recent investments do not solve the cold-chain dilemma.
They simply stop pretending it is something else.
Cold chain resists leapfrogging.
It punishes shortcuts and rewards sequence.
