West African telecoms infrastructure: thirty years on

In the early 1990s I spent the better part of three years on West African telecoms build-outs. The work was unlike anything I had done before. We were running copper into towns that had not had a working telephone exchange in living memory, training operators in switching equipment that had arrived a generation out of date elsewhere in the world, and negotiating right-of-way over ground whose ownership depended on a chain of authority that ran through a colonial registry, a post-independence ministry, a customary chief, and an extended family that had farmed the land for several generations. Each of those parties had a different idea of what they were entitled to, and each had a different mechanism for expressing that entitlement. The work was not principally technical. It was principally relational.

I have been asked, more times than I can count, what has changed in the thirty years since. The honest answer is that almost everything has changed in the technology and almost nothing has changed in the underlying realities of the work.

What has changed

The technology, first. The copper that I spent months installing across difficult terrain has been almost entirely superseded by mobile networks. The countries we worked in then had landline penetration measured in single digits per hundred. They now have mobile penetration that meets or exceeds many developed-economy comparators, and a digital-payments infrastructure that, in some respects, is more sophisticated than what is routinely available to a UK consumer. The leapfrog has been one of the genuinely remarkable infrastructure stories of the last several decades.

The capital structure has changed too. The build-outs of the 1990s were funded predominantly through bilateral aid and through the development banks, with sovereign guarantees underpinning supplier credit and equipment leasing. The build-outs of the 2010s and 2020s have been funded through a more sophisticated mix: sovereign wealth co-investment, private equity rolling up regional carriers, the major Chinese infrastructure programmes, and increasingly through the African development institutions themselves. The capital is more patient, more diverse in origin, and considerably more substantial in aggregate than it was when I first started.

The institutional capacity has changed in places. Some of the regulators we dealt with thirty years ago barely existed in any meaningful operational sense — the regulatory function was performed by whichever ministry had the closest political interest in the outcome of a given decision. Several of those regulators are now functioning institutions, with technical staff, defined procedures, and decisions that follow recognisable patterns. The pace of change has not been uniform across the region, but where it has happened, it has been substantive.

What has not changed

The underlying realities of getting infrastructure built, however, have changed less than the technology suggests. The work is still principally relational. The decisions that move a project forward are still made by individuals whose reasoning cannot be reduced to the documented institutional process, even where the institutional process is real. The right-of-way that runs over land with multiple overlapping claims of authority is still negotiated, ultimately, between the senior people on each side of the agreement, in a register that draws on personal credibility built over years of working in the region.

The capital still moves on terms that reflect the political-risk perception of the moment, and that perception is still set by people who, in many cases, have never spent meaningful time on the ground in the country they are pricing. The premium demanded for emerging-market infrastructure exposure is sometimes well-calibrated to the actual risk and sometimes not, and the projects that close successfully often do so because someone with credibility in both worlds — the deal world and the on-the-ground world — has been able to bridge the perception gap. That bridging role has not been automated away by any of the technology changes of the last thirty years, and there is no obvious reason to expect that it will be.

The deal cycle still tracks the political cycle. Election years produce a particular kind of slowdown, post-election years produce a particular kind of acceleration, and the relationship between the two is not always the relationship that the published country-risk frameworks would predict. The people who originate successfully in the region understand this rhythm and time their work around it. The people who do not understand it find themselves with mandates that go quiet for reasons they cannot diagnose and that come back to life for reasons they cannot replicate.

The lessons that still apply

Several lessons from the early work apply with as much force today as they did when I first learned them.

The first is that documentation matters less than the underlying agreement, but documentation that is loose makes future disputes more likely and more expensive. I have watched several deals fall apart for reasons that traced, in the end, to a letter of intent that was drafted in a register the parties did not share. The cost of getting the documentation tight at the start of an engagement is small. The cost of repairing it after a misunderstanding has hardened is sometimes the deal itself.

The second is that the people who introduce you to a counterparty are part of the deal in a way that the institutional framing of the transaction rarely makes explicit. A project that comes to you through a long relationship with a respected intermediary is a different kind of project from one that comes to you through a cold approach, even where the underlying opportunity looks identical. The credibility chain that runs through the intermediary is part of what makes the deal closeable, and it is part of what is at stake if the deal goes wrong.

The third is that the time horizon is longer than the deal sheet suggests. A project that takes nine months to close is, in the lived experience of the work, a project that has been worked on for several years before the nine-month phase began and that will continue to require attention for several years after the close. The people who do this work well treat the close as a milestone in a longer relationship, not as the conclusion of an engagement.

Where the region is going

The next decade in West African telecoms will, in my reading, be defined by three interrelated developments. The first is the consolidation of the regional carrier market, which has been ongoing for some years and which has further to run. The second is the integration of the digital-payments and connectivity stacks into a broader regional commerce infrastructure that will, for the first time, allow cross-border commerce within the region at the same level of friction as within comparable economies elsewhere. The third is the gradual maturation of the regulatory framework into something closer to the architecture of more developed regulatory systems, with the political-risk implications that follows from that.

The opportunities for capital and for infrastructure-origination work are substantial. The risks are real, and they are real for the same reasons they have always been real: the political environment is volatile in places, the rule-of-law expectations vary across jurisdictions, and the documentation that holds a deal together has to be designed with both the institutional and the relational dimensions in view. None of this is news to anyone who has worked in the region. It bears repeating because the people who have not worked in the region routinely underestimate it.

What this means for new entrants

I am asked, periodically, by people who are considering working in the region for the first time what they should attend to. The advice I would offer, distilled from a long acquaintance with the work, is in three parts.

The first is to take time on the ground before taking on commitments. The instinct to assemble engagements quickly, encouraged by capital partners who want to see deal flow, runs against the grain of how the region actually works. The engagements that close cleanly are the ones that have been built on relationships established over months at minimum and often over years. New entrants who try to compress this timeline find that they are accumulating commitments faster than they are accumulating the relational underpinning that those commitments require, and the resulting engagements break under their own weight.

The second is to identify, early, the small number of intermediaries whose judgement is worth depending on. Every regional market has a handful of people who have been working in it for decades, who know the principal parties, and who are willing to make introductions on the basis of a track record they have personally verified. Identifying these people, earning their willingness to make introductions, and treating their introductions with the care those introductions deserve is among the most valuable single investments a new entrant can make. The alternative — assembling a network through cold approaches at conferences and trade shows — produces a different kind of network, less reliable in operational terms.

The third is to expect the timelines to be longer than the published commercial wisdom suggests. The deal that the documentation describes as a six-month engagement is, in practice, a two- or three-year engagement when the relationship with the counterparty is included in the calculation. The new entrant who plans their cash flow and their attention allocation against the documentation timeline finds themselves under pressure when the engagement extends, as it almost always does. The new entrant who plans against the realistic timeline can absorb the extensions without distress.

A closing reflection

The young man who arrived in West Africa in the early 1990s with a copper-cable specification and a contract to build telecoms infrastructure could not have imagined the technology that would replace the network he was installing. He could, however, have recognised, in outline, the relational architecture of the work that produced the network and the relational architecture that produces the equivalent work today. The infrastructure changes faster than the institutions, and the institutions change faster than the underlying patterns of trust on which infrastructure origination depends. In a thirty-year career, that observation is among the most durable I have made.