Amazon Leo targets satellite-to-phone service in South Africa around 2028

Amazon Leo is planning a direct-to-device satellite service for South Africa from around 2028, according to comments made at the Conext 2026 conference this week. That would allow compatible ordinary mobile devices to connect via satellite in places where terrestrial coverage is weak or absent. It is a proposed timetable, not a confirmed South African commercial launch date.

The company already has a separate route into the local market. Its satellite broadband partnership with Herotel is expected to bring a fixed-connectivity service called evry to South African households in 2027. The newer direct-to-device proposal would add a mobile layer, subject to regulatory permissions and agreements with network operators.

What was said at Conext

ITWeb reported on 8 October that Mark Anderson, Amazon Leo’s head of consumer operations for Europe, the Middle East and Africa, outlined the mobile plans during a panel at Digital Council Africa’s Conext 2026 event. Anderson said the company had applied for licences and was working with operators, with a possible service window around 2028.

Direct-to-device is different from satellite home broadband. A fixed service generally requires a dedicated user terminal or antenna. Direct-to-device systems aim to reach compatible phones or connected devices using satellite infrastructure and arrangements with terrestrial mobile networks.

The technology could be useful in farming areas, remote roads and communities beyond economical fibre or tower coverage. But satellite-to-phone connectivity is not an automatic substitute for a 4G or 5G network. Capacity, spectrum rights, handset compatibility, service availability and commercial pricing all affect what customers can actually do.

A South African partner is already in place

Amazon announced its agreement with Herotel in July 2026. Herotel will distribute Amazon Leo connectivity through evry, with commercial residential service expected in 2027. Amazon says Herotel has more than 350,000 active customers across more than 550 towns and a network of 120 local offices.

That local footprint matters because satellite broadband still needs installation, billing, support and fault handling. A distributor with teams outside the biggest metros may reach households that a purely online sign-up model struggles to serve.

The fixed service and the direct-to-device proposal should not be confused. The Herotel announcement establishes a commercial distribution plan for broadband. It does not establish that a South African mobile operator has signed a retail direct-to-device contract or that the regulator has approved every element of the proposed mobile service.

The licensing question

South Africa’s satellite market sits inside the country’s electronic communications licensing framework. The Independent Communications Authority of South Africa, ICASA, controls relevant licensing and spectrum authorisations. The mobile service will depend on the outcome of the applications and any network partnerships, not simply on the number of satellites in orbit.

Amazon’s approach also differs from a direct-to-consumer model in that the Herotel arrangement uses a South African distribution partner. It would be premature to assume the same structure, pricing or customer service terms for direct-to-device.

For Vodacom, MTN, Telkom and Cell C, satellite coverage could eventually extend service into areas where new towers are costly. It also raises commercial questions: who owns the customer relationship, which spectrum is used, how roaming is charged and what performance can be guaranteed when demand rises.

What remains unknown

There is no verified South African retail price for Amazon Leo direct-to-device service, no published local handset compatibility list and no confirmed consumer launch date. Even the proposed 2028 window is conditional. Herotel’s 2027 fixed-broadband plan has its own outstanding details, including final retail packages and the timing of coverage by location.

The practical story for South African users is therefore a two-stage prospect: fixed satellite internet via evry first, followed by a possible mobile satellite layer. The next meaningful milestones are regulatory decisions, signed mobile-operator agreements and actual local service terms.

Why “around 2028” is not a product page

Conference language is designed to signal intent without locking Amazon into a hard launch. “Around 2028” leaves room for licensing delays, constellation capacity, handset certification and operator commercial talks. South African readers should treat it as a planning horizon, not a booking window.

ICASA’s process will also set the tempo. Even if Amazon’s applications move cleanly, spectrum and service authorisations rarely land on a marketing calendar. Operators will want clarity on liability, emergency calling behaviour, messaging versus data capacity, and how satellite sessions appear on a customer’s bill.

Handset compatibility is another quiet gate. Direct-to-device only matters if popular devices sold in South Africa can attach to the system without obscure firmware paths. Until Amazon and partners publish a local device list, the service remains a concept for most consumers.

Competition and the coverage story

Amazon is not the only company talking about satellite-to-phone. Global rivals have floated similar ambitions, and South African operators already sell coverage products that mix fibre, microwave and roaming. The commercial fight will be less about “does space work” and more about whether satellite minutes are cheap enough, reliable enough and simple enough for rural customers and travellers.

Capacity under load is the hard engineering question. A sparsely used remote link can look impressive in a demo. A crowded festival weekend or disaster surge can expose limits. Buyers should ask for service descriptions that separate emergency messaging from broadband-like expectations.

Wire Room takeaway

Two clocks are running. Herotel’s 2027 fixed Amazon Leo broadband plan is the nearer, more concrete South African product path. Direct-to-device around 2028 is the longer bet — useful if licensing, operators and handsets line up; irrelevant if they do not.

Until those artefacts appear, the story is intent plus a distributor for fixed service. Intent is news. A live retail SKU with coverage maps and prices is the scoreboard.