The Journey of Adaeze Okonkwo: Building a Logistics Network Across Fourteen Markets

Adaeze Okonkwo does not describe the first year of Kestrel Logistics as a strategy. She describes it as a series of promises she had to keep.

"I had three trucks and a customer who needed forty pallets in Accra by Thursday," she says. "There was no operating model. There was only the Thursday."

Eleven years later Kestrel moves freight across fourteen markets in West Africa and the Gulf, employs just over six hundred people, and runs a routing platform that Okonkwo's team built in-house after two vendor systems failed to handle the exception rate her corridors produce. She remains, by her own description, an operator first.

Starting where the maps ran out

Okonkwo trained as an industrial engineer and spent six years inside a multinational shipping group before leaving to start Kestrel. The gap she saw was not in long-haul capacity. It was in what happens after the container is discharged.

The last two hundred kilometres are where the margin lives and where everybody stops paying attention. We decided to pay attention.

That decision shaped the company's cost structure in ways she says took years to justify to outside investors. Kestrel owns more of its fleet than a comparable asset-light competitor, and it staffs its own regional dispatch desks rather than routing everything through a central hub.

"An asset-light model assumes the subcontractor market is deep and reliable," she says. "In several of our corridors it simply is not. You can either wait for that market to mature or you can build the capability yourself. We built it."

The exception rate problem

The technical turning point came in 2021, when Kestrel's third-party routing software began producing delivery windows that its own dispatchers routinely overrode.

The vendor's model assumed a level of road, border and fuel predictability that Okonkwo's corridors did not offer. Her dispatchers were correcting roughly a third of all assignments by hand, which meant the software was, in practice, generating work rather than removing it.

Her team spent fourteen months building a replacement. The result is deliberately unglamorous: a routing engine that treats disruption as the baseline rather than the exception, and that asks a human to confirm anything it is not confident about.

"People hear 'we built our own software' and assume it was an ego project," Okonkwo says. "It was the opposite. We built the smallest thing that would stop my dispatchers from fighting the system."

What she still does herself

Every Monday morning Okonkwo reviews the previous week's routing exceptions. Not a summary. The list.

She is aware this does not scale, and she is unapologetic about it. "The exceptions are the only honest description of your business," she says. "Everything else is a number that has already been rounded."

Her leadership team has, at various points, tried to take the review off her calendar. She has let them redesign the format twice and has declined to give up the meeting.

On expansion, and on saying no

Kestrel has turned down two of the last five expansion opportunities its commercial team brought forward. In both cases, Okonkwo says, the corridor economics worked and the operating reality did not.

"We could have won the contract and then spent three years apologising for the service," she says. "I have done that before. It costs more than the revenue is worth, and it costs it in the currency you cannot replace, which is your operators believing that the company means what it says."

Asked what she would tell a founder building in a market where the infrastructure assumptions of standard playbooks do not hold, her answer is short.

"Stop trying to find the version of the playbook that fits," she says. "Write down what actually happens on your route on a bad Tuesday. Build for that. The playbook was written by people who never drove it."