Summary

  • AfNOG publishes extensive operational, financial and non-commercial requirements for an Africa Internet Summit host.
  • Its hosting page identifies three preferences: local impact, geographic rotation and prior familiarity with workshop requirements.
  • The same page assigns the final choice to AfNOG Convenors and AFRINIC Management and describes that decision as final.
  • The reviewed page does not publish weights, a tie rule, conflict handling, a reason practice or a factual-correction route; that does not prove those controls are absent internally.
  • A compact scorecard and decision record would make the final authority auditable without converting host selection into an appeal contest.

The application is more than a venue bid

AfNOG’s Hosting page does not treat a host as a logo above a conference entrance. It sets out a delivery role that reaches from the network operations centre to registration, classrooms, power, connectivity, equipment, publicity and local transport.

The prospective host may also be expected to support the event financially, in cash or in kind. Publicity, government participation, souvenirs and commercial uses require coordination. The page says the workshop and related meetings are non-commercial, technical and community-oriented, and it reserves approval over sponsor advertising that could undermine that atmosphere.

These are material obligations. A bid must connect local capacity with the operational needs of a multi-day technical event. A weak selection can impose costs on students, instructors, delegates and organisers long after the host announcement.

That explains why the final decision needs discretion. It also explains why the way that discretion is structured matters.

Three preferences can point in different directions

The page publishes three preferences for choosing a host.

One is impact: preference goes to countries where a local workshop and meeting would have the greatest effect. A second is geographic rotation: preference also goes to a region other than the immediately previous host location. A third is delivery familiarity: strong preference goes to organisations that know the workshop requirements through previous participation.

Each can be defensible. None necessarily identifies the same winner.

A place with the greatest capacity-building need may have less experience. An experienced organisation may sit in a region that hosted recently. A proposal with excellent facilities may offer less geographic reach than a more difficult bid from an underserved location.

The reviewed text does not assign weights or publish a method for reconciling those cases. That is not evidence that the convenors improvise. It is a limit on what bidders and the wider community can reconstruct from the public rule.

The final authority is named

AfNOG does publish who holds the last decision. The Hosting page says the host country or organisation will be selected by AfNOG Convenors and AFRINIC Management, whose decision is final.

Naming the authority is valuable. It is better than leaving applicants to infer whether the secretariat, a local partner, a programme committee or an informal group makes the choice.

The current AfNOG homepage also says the 2026 Africa Internet Summit is jointly organised by AFRINIC and AfNOG and hosted by TESPOK in Nairobi. That confirms a current joint-organiser and host relationship. It does not, by itself, disclose how the 2026 host was compared with any alternative.

The temporal distinction is important. The hosting call names proposal dates for 2021 to 2023 even though the page remains publicly linked and carries a 2026 footer. It should be read as a published selection architecture, not automatic proof that every clause governed the 2026 choice unchanged.

Finality is not the same as opacity

A host decision cannot remain open indefinitely. Venue holds expire. Budgets depend on dates. Instructors, sponsors and participants need a stable location. A final decision is therefore a reasonable design choice.

But finality answers when the process ends, not how the choice can be understood.

A public record does not need to reveal losing bids or commercially sensitive budgets. It can show the eligible decision body, the criteria used, whether a conflict required recusal, and a short reason explaining the chosen balance among impact, rotation, familiarity and feasibility.

That record protects the decision-makers too. Without it, a legitimate trade-off can look like an unexplained preference. With it, the community can disagree with the outcome while seeing that the published rule was applied.

A factual-correction window is not a second contest

The page expects an assessment before AfNOG and the selected host sign a memorandum of understanding defining their roles.

That creates a natural control point. Before the MoU is final, an applicant could have a short window to identify a material factual error: a facility marked unavailable when it was confirmed, a regional history stated incorrectly, or prior participation omitted from the record.

The review should not re-score the merits or let political pressure reopen the choice. It should ask only whether the decision relied on a verifiably incorrect input.

This narrow correction route preserves finality while reducing the risk that a fixable record error is carried into the memorandum of understanding.

Conflicts should be visible at the role level

Host selection can connect organisers, sponsors, venues, local associations and past collaborators. That does not imply a conflict in any particular decision. It makes conflict handling a predictable part of the architecture.

A public policy can state that a convenor or manager with a material relationship to a bidder declares it, does not score that proposal, and is replaced for the decision. The record need not publish private details. It can identify the role recused and confirm that the eligible decision body remained quorate.

The reviewed pages do not say whether AfNOG or AFRINIC already uses such a process internally. The governance recommendation is to make the control legible, not to allege its absence in practice.

A compact scorecard would be enough

The Hosting page already does most of the hard work. It tells applicants what the event needs and identifies the broad purposes behind host selection.

A compact public scorecard could complete the chain. It could list feasibility as a threshold, then show the relative assessment of local impact, geographic rotation, prior operational familiarity and delivery risk. It could identify the decision roles, note recusals, give the selected host’s principal strengths and confirm whether the factual-correction window changed the record.

The public output could avoid rankings and confidential proposal details. Its purpose would not be to prove that one city is objectively best. It would show how a final authority used the preferences it had already published.

AfNOG’s meeting history reflects a deliberate movement across African countries. The hosting page makes that ambition, and the work required to realise it, visible. Publishing the comparison behind the final choice would connect the same public rule to the authority that applies it.