Summary

  • WIOCC's front page, as retained on 18 October 2026, states a position rendered as "Ownership on Equiano and 2Africa" and places it in South Africa, Nigeria and other markets, without a percentage, a capacity quantum, a fibre pair, a segment, a landing point or an as-of date.
  • The same block gives three counters that do carry units — more than 115,000km of terrestrial fibre, 13 OADC core data centres in Nigeria, the Democratic Republic of the Congo and South Africa, and more than 30 OADC edge data centres in South Africa — which makes the unquantified ownership line conspicuous by contrast.
  • The one non-company measurement retained here reports 1,429 observed BGP peers for AS37662. That is a statement about routing adjacency, and it cannot be read as a statement about fibre, equity or physical diversity.
  • No capacity instrument, landing-party schedule or consortium membership list naming the company was obtained as text. Until one is, a buyer's second path on a named system rests on a commercial assertion rather than on a document.

A wholesale capacity business sells a property that is hard to photograph. What a customer actually purchases is the expectation that the path it holds will still exist when the other path fails, and that expectation is only as good as the document behind it. On 18 October 2026, WIOCC's own front page published a fragment that speaks directly to that expectation and, in the same breath, declines to supply the document.

The fragment is short and worth reading as it stands. It states, in the company's own words, "Ownership on Equiano and 2Africa," joined by a plus sign to "in South Africa, Nigeria and other markets" (WIOCC, retained 2026-10-18). Beneath it sit three counters: more than 115,000km of terrestrial fibre, 13 OADC core data centres in Nigeria, the Democratic Republic of the Congo and South Africa, and more than 30 OADC edge data centres in South Africa. The retrieval is dated; the counters are not. Nothing in the retained text says when the 13 became 13, or whether the edge estate was counted on the same day as the fibre.

That asymmetry matters because of what the four statements are made of. Three of them are denominated: kilometres, sites, sites again. They can be compared against a later snapshot and they can be wrong in a way that is detectable. The fourth — ownership — is a legal and commercial relation that is denominated in nothing at all as published. It does not say what share of either system the company holds, whether that holding is equity in a consortium vehicle or capacity purchased under an instrument, which segment the position attaches to, or which landing station on that segment the company is entitled to use.

"Ownership" is doing a great deal of work in a four-word line, and the retained text does not tell the reader which of those senses is meant.

This is a narrower point than saying the company has published nothing. It has published something new. A BTW assessment earlier the same day recorded that the documented surface around AS37662 stated no design capacity, no fibre-pair count, no landing point, no licence reference and no repair interval, and concluded that deliverable capacity is produced by parties who keep separate books (BTW, 18 October 2026). A claim has now appeared where the earlier assessment recorded silence. What has not appeared is any accompanying instrument. The claim is therefore an advance in disclosure and not yet an advance in verifiability, and those are different things.

The second retained source shows why the substitute is tempting. A routing record for AS37662 reports "BGP Peers Observed (all): 1,429" and lists adjacent autonomous systems including AS1299, AS5511, AS9498, AS9002, AS20485, AS8220, AS31133, AS3216, AS174 and AS6939 (Hurricane Electric BGP record for AS37662, retained 2026-10-18). Four figures and a peer list look like an inventory. They are not one. A peer count aggregates observed adjacencies across whatever collection points the observer used; it records that networks exchange routes with this autonomous system at those points. It does not record which cable a packet crosses, who holds title to the glass, where the traffic lands, or whether two paths sold to one customer share a beach segment. The number is real, and it answers a different question from the one a redundancy buyer is asking.

That distinction has a commercial consequence rather than merely a semantic one. When a buyer asks for diversity, the useful question is not how many peers a provider has but which failure domains the two paths share. Landing stations, beach segments, consortium repair vessels and power feeds are shared by construction in many cases; a wholesale provider can hold genuinely separate capacity on two systems and still present two paths that converge at one landing point. Nothing in the retained material establishes that WIOCC does this, and nothing in it rules the possibility out.

The point is that the published record does not let a customer tell, and the peer count does not help.

Three of the remaining retained sources are summaries rather than text, and the difference is worth marking because it limits what can be said. A summary describes 2Africa as a consortium-led system and states explicitly that the specific landing-party list was not retrieved (2Africa project site, retained 2026-10-18). A second describes EASSy as owned by a consortium of telecommunications operators and characterises WIOCC as founded as an investment vehicle for several African telcos to acquire and manage capacity in that system (submarine cable reference, retained 2026-10-18). A third records a USD 20 million loan from the International Finance Corporation to WIOCC in 2020 (Telecompaper, retained 2026-10-18). A fourth describes the company as a wholesaler of capacity to international telecoms, cloud operators, content providers and internet service providers (aggregated company record, retained 2026-10-18).

Read together, those summaries sketch a business model in which the company's position on a system has historically been an investment-vehicle stake in consortium capacity rather than a landing-party franchise. That reading is consistent with the ownership line and does not confirm it. The EASSy characterisation arrives in a summary with no excerpt attached; the 2Africa landing-party list was not obtained; the 2020 loan is recorded without its terms, tranches or covenants. Each of those is a retention boundary in this assessment, not a finding about the company.

None of them is evidence that WIOCC is absent from any list, and the article does not treat it as such.

What the counters do establish is that the facility and terrestrial layers are being asserted at a scale where they would be the natural subject of a separate schedule. Thirteen core facilities across three countries and more than thirty edge sites in one country is a portfolio claim, and the retained fragment does not say whether those sites are built by the group, leased, or operated by partners under the OADC brand. A branded facility estate assembled from partner sites is a legitimate commercial fact and a poor predictor of physical redundancy, because the failure domains of a partner-operated site are not the group's to describe.

The same absence applies to the fibre figure: more than 115,000km is a length, and length is not lit capacity, not route diversity and not a maintenance obligation.

None of this is a criticism of a company for publishing less than a buyer might want. Submarine capacity is sold under instruments that are routinely confidential, and a consortium stake is often held through a vehicle whose terms are not public by design. The observation is narrower and more useful: the disclosed surface has a shape. Unit-bearing counters and a routing metric sit on one side; the asset relation sits on the other; and only the first side is checkable from outside.

A buyer signing a diversity requirement against the second side is relying on the seller's characterisation, which is exactly what the word "ownership" asks it to do.

Four documents would change that, and naming them in advance is the point of doing so. A capacity or indefeasible-right-of-use instrument naming the consortium and the quantum would convert the ownership verb into a measurable position. A landing-party schedule or consortium membership list naming the company would attach the claim to a place. A facility schedule with an as-of date, distinguishing built from leased capacity, would make the 13 and the 30 comparable across periods rather than a one-time assertion.

An operational notice naming a cut and the affected segment would test any diversity claim against an event rather than against a brochure. Any one of those would move this subject from a described position to a checkable one; their continued absence would show that reach is being sold on narrative.

What remains genuinely open, and worth stating plainly, is whether the ownership line refers to a stake acquired in the two systems' consortium vehicles, to capacity purchased on them, or to a landing and backhaul position that carries rights rather than title. The retained fragment supports no answer. The company's own earlier descriptions, as summarised here, suggest the investment-vehicle reading; that summary is not the instrument.

The honest conclusion at this dateline is that WIOCC has published a claim about the physical layer whose unit is undefined, beside three counters whose units are clear, and that the independent record retained alongside it measures interconnection and says nothing about the fibre.