Web sQuad Connect and the South African ISP Margin Behind Local Peering

The contradiction at the centre The most interesting thing about South African fibre retail is that it can look both brutally competitive and quietly monopolistic at the same time. The brutality is at the retail layer: customers can often choose among dozens of ISPs on the same open-access fibre network, compare packages on network marketplaces, and switch when service, latency, or price disappoints. The monopolistic element sits one layer lower: once a fibre network operator has built the local last mile into a suburb or estate, that local access asset behaves like a geographically bounded monopoly or near-monopoly, and every retailer above it must rent the same basic access input. The Competition Commission’s data-services inquiry made the broader point years ago: fixed infrastructure is well-suited to large data volumes because costs are mostly fixed and sunk, but the same infrastructure can also lend itself to local monopoly positions if countervailing market discipline is weak.

That is why the key economic question around Web sQuad Connect is not whether it is “a fibre ISP” in the ordinary marketing sense. Plenty of firms can print that promise on a website. The real question is whether a mid-sized South African operator can convert a bundle of secondary advantages, local peering, local support, multi-network fibre resale, a little wireless reach, and some density in business customers, into durable margin even when the largest cost items are not under its control. The contradiction is that peering and engineering can genuinely lower the cost of an ISP’s network, while last-mile rentals, supplier outages, load-shedding resilience, theft, and national price competition can compress almost everything saved further up the stack.

The background conditions in South Africa make this contradiction sharper, not softer. Fixed internet at home remains a minority product nationally: Statistics South Africa reported that only 20.6% of households had fixed internet at home in 2025, although the figure was much higher in Gauteng at 31.1% and in metropolitan municipalities at 33.5%. Johannesburg itself reached 34.8%. That means the addressable premium market is still concentrated, urban and unequal: good news if an ISP can defend high-value metro and SME clusters, less good if it needs mass-market volume across the country.

The sector is also structurally alive rather than saturated in a way that would let margins settle quietly. ICASA’s 2026 sector report said annual investment in fixed, wired broadband increased by 11.9% in 2025 even as total telecommunications investment fell by 2.3%, which suggests that capital is still pursuing fixed infrastructure where operators think demand or strategic leverage remains available. More fibre investment means more capacity and more coverage, but it also means more pressure on the retail layer as each network seeks payback from homes passed and connected.

So the initial answer is not a clean yes or no. Web sQuad can plausibly earn margin, but only if that margin is created in the places where retail operators still matter: traffic engineering, business-grade packaging, supplier management, real support, and the internal economics of a group backbone. If it tries to live purely as a commodity residential reseller on other people’s fibre, the public record suggests the margin pool is too thin and too vulnerable.

What Web sQuad actually sells Web sQuad’s public pages show a much broader proposition than a single suburban fibre product. The retail front end spans home fibre, business internet, voice, hosting, domains and fixed wireless. The business page explicitly offers two different classes of service: “Dedicated Business Fibre” through “Corporate Connect,” described as guaranteed-speed, uncontended, synchronous bandwidth with SLAs, and “Business Broadband Fibre,” described as lower-contention business access with guaranteed uptime. The same page promises speeds from 10 Mbps to 10 Gbps, uptime guarantees up to 99.5%, and delivery across “15+ FNO and Wireless Networks” using “Licensed Wireless backbones.” That is not the language of a pure consumer-only ISP; it is the language of a firm trying to monetise the distance between commodity access and business-critical connectivity.

The residential side is designed to look simple, but its economic significance lies in the breadth beneath that simplicity. Web sQuad says its home fibre runs across a long list of South African fibre network operators, including Evotel, Frogfoot, MetroFibre, MTN, Netstream, Octotel, Openserve, Vumatel and many smaller footprints besides. That breadth matters because it turns the ISP layer into a reseller-aggregator of local fibre monopolies. The customer sees one helpdesk; Web sQuad sees a matrix of different installation processes, activation fees, VLAN behaviours, support procedures, packet-loss symptoms and wholesale price schedules. In South Africa’s open-access market, operational competence in that matrix can itself be a scarce asset.

The website also makes a fairly aggressive no-fuss retail pitch: uncapped, unshaped, unthrottled internet, no term contract, free activation, a free-to-use router, and IPv6-capable equipment on many packages. That tells you two things. First, Web sQuad is competing partly on quality signalling aimed at gamer, streamer and prosumer households rather than on the cheapest bare-bones package. Second, it is subsidising acquisition in the familiar ISP way, by absorbing part of the set-up cost and hoping the customer stays long enough for the subsidy to wash out.

Wireless is not an afterthought either, though it is not presented as a national mass platform. Web sQuad’s Air Fibre page offers 15, 25 and 50 Mbps packages, sold as uncapped, unthrottled and unshaped, subject to site survey and feasibility, with installation complexity explicitly called out. On the business page, wireless appears again as part of the delivery mix. Economically, wireless matters less because the advertised speeds are dramatic, and more because it offers reach where fibre economics are poor and can function as an alternative or backup path for businesses that cannot tolerate single-network dependence. That kind of product does not usually generate Vodacom-scale revenue, but it can meaningfully thicken margin in selected geographies and SME accounts.

The voice product points in the same direction. Web sQuad sells VoIP with per-second billing, geographic number porting, and a no-contract proposition. That matters because voice, domains, hosting and managed business access are exactly the sort of attachments that can turn a thin-margin access customer into a respectable account. Residential access is where churn lives; communications bundles are where stickiness often begins.

This is also where the public corporate record starts becoming more interesting than the front-end marketing. Web sQuad Connect’s PAIA manual identifies the operating company as Web Squad Connect (Pty) Ltd, registration number 2016/056869/07, at 4 Hans Schoeman Street, Malanshof, Randburg. The same address appears on the contact page. Yet PeeringDB lists the network organisation behind AS328137 as Wecom Holdings (Pty) Ltd, with “Web sQuad” as the public-facing brand. AFRINIC-derived WHOIS data shown by bgp.tools links the ASN to the same Randburg address and to contacts named Carl Fayolle and Clarissa Ferreira. Meanwhile, older commercial traces and third-party company directories show related entities at the same address, including Web Squad Business and other Web Squad-branded vehicles. Most tellingly, Web sQuad’s current voice terms define “WEB SQUAD” as Web Squad Telecom (Pty) Ltd, registration number 2005/037954/07, even though the contract header names Web Squad Connect.

That mishmash is not necessarily sinister. In telecoms, especially in firms that began in web hosting or IT services and then extended into connectivity, layered entities are common. But economically it matters a great deal. If retail access sits in one company, voice or telecom rights in another, and backbone or transit in a wholesale affiliate, then the margin story may not be visible at the storefront level at all. The real question becomes whether the group can internalise enough network cost and enough high-value customer spend to avoid becoming just another thin reseller of other people’s suburban fibre.

What the network record proves The strongest public evidence in Web sQuad’s favour is not in its advertisements but in its routing footprint. AS328137 is a real, active South African network with its own AFRINIC registration, its own prefix inventory, valid RPKI announcements, and a substantial peering posture. BGP tools shows AS328137 with around 555 peers and one direct upstream at the time of capture, while also showing a wide set of announced prefixes, including the 160.119.224.0/20 block and multiple location-labelled prefixes such as JNB, DB1, CP1 and JB1. PeeringDB lists the network as a cable/DSL/ISP operator with an open peering policy, an IRR route-set of AS-WEBSQUAD, and hundreds of IPv4 and IPv6 prefix limits. In other words: this is not just a billing and support front end glued onto someone else’s cheapest wholesale internet. It has enough network substance to matter.

The exchange presence is especially revealing. Public IX records show Web sQuad joined INX in 2018 and currently appears at Durban Internet Exchange with a 10 Gbit/s port, at Cape Town Internet Exchange with 1 Gbit/s, and at Johannesburg Internet Exchange with 10 Gbit/s. NAPAfrica’s member detail for the same ASN shows an even more assertive position: 20 Gbit/s in Johannesburg, 10 Gbit/s in Durban, and additional presence in Cape Town. BGP tools rounds this out with multiple ports: two 20 Gbit/s ports in NAPAfrica Johannesburg, two ports in NAPAfrica Cape Town, a 10 Gbit/s port in Durban, plus ports at DINX, JINX and CINX. For a South African mid-sized ISP, that is not decorative peering. It looks like deliberate investment in keeping local traffic local and in de-risking the cost of carrying popular content.

Why does that matter economically? Because internet traffic is not homogeneous. If a large share of customer demand is toward cached or peering-friendly destinations, Cloudflare-hosted content, Google platforms, AWS endpoints, Akamai properties, major SaaS and local-to-regional networks, then every byte exchanged locally through neutral IX fabric is a byte that does not need to traverse more expensive paid transit. NAPAfrica explicitly says it does not charge membership, port or cross-connect fees. That does not make exchange participation free in the real business sense, because colocation, transport to the exchange, routers, optics and staff still cost money. But it does mean that a competent ISP can disproportionately improve its gross network economics by doing the engineering work well. BGP tools’ peer list for AS328137 includes Cloudflare, Google, Amazon, Hurricane Electric and others, which is exactly the kind of local reach that reduces paid-transit dependence.

There is, however, an important catch hidden in the upstream graph. AS328137’s sole direct upstream in the sampled BGP data is AS37731, also identified as Web Squad Connect (Pty) Ltd on bgp.tools but tied publicly to the WECOM website and, on PeeringDB, to Wecom Holdings. AS37731 shows a different commercial profile: four upstreams, including Cogent, PCCW Global, Gateway Communications and Session Telecoms; roughly 556 peers; and 17 downstreams. That looks much more like a carrier or wholesale backbone role than a pure retail FTTH edge network.

This relationship is probably the single most important public clue in the entire business model. It suggests that the economically significant network is not just retail Web sQuad but a retail-plus-wholesale architecture in which the customer-facing ASN rides on a group-level or affiliated backbone. That has two consequences. The positive one is that Web sQuad may be able to buy transit, colo and backhaul more efficiently than a pure retail reseller because part of that cost stack is internalised or at least negotiated at group scale. The negative one is that the public record cannot show where the margin actually lands. If the wholesale layer earns the return and the retail layer mostly acquires and supports customers, then looking only at Web sQuad’s front-end pricing would understate the economics. The public network data proves the architecture exists; it does not prove how the profits are allocated inside it.

The founder-side narrative published in Carl Fayolle’s AFRINIC candidate profile fits this interpretation, though it must be treated carefully because it is self-authored and not an audited filing. In that profile he says he founded Web sQuad in 2011, secured AFRINIC membership for Web sQuad, built a metropolitan fibre broadband network with Arista switches, BGP peering and VXLAN overlays, and grew Web sQuad’s revenue above R10 million by 2015 and then to R25 million. He further claims that WECOM, co-founded in 2016, grew to roughly R70 million in annual revenue by 2024, offering transit, colocation and interconnects to more than 70 African and international networks. If even directionally correct, that tells you the group’s economics are not those of a simple consumer fibre helpdesk. But because the source is a candidate biography rather than a financial statement, it should be read as suggestive, not conclusive.

The network record also strengthens the case that Web sQuad has built a latency-and-support identity that matters in the South African context. Public forum discussions show users specifically associating Web sQuad with static IP friendliness, IPv6 support, gaming latency and direct engineering engagement. Those are not universal customer priorities, but they are exactly the priorities that create profitable subsegments inside a crowded FTTH market. A buyer who cares mostly about the last R40 on the monthly bill behaves differently from one who cares about stable routing, SIP trunks, on-call escalation and direct human support.

So the network evidence proves something quite specific. Web sQuad is not commercially interesting because its homepage promises “simple internet.” It is commercially interesting because the public routing footprint suggests real network investment, real exchange presence, and a plausible group-level backbone underneath the retail brand.

Where the margin might come from The blunt answer is that most of the obvious consumer-access margin probably does not come from the fibre line itself. Web sQuad’s own contracts make this clear in ways that glossy marketing pages never do. Home fibre is sold month to month, but with one full calendar month’s notice, a minimum of two months’ service, router return obligations, and clawback fees on subsidised activation or installation if the customer cancels early. The FTTH terms expressly contemplate the ISP recovering subsidised set-up costs from a customer who leaves within six months, with a general R999 clawback and a heavier R2500 charge for some Evotel new installations. That is textbook thin-margin retail behaviour: give away or subsidise set-up elements, then protect payback by making early exit more expensive.

The same terms contain an even sharper clue. Web sQuad states that it will not credit customers for fibre downtime caused by maintenance, line breaks, FNO network issues, load-shedding, or other third-party causes. That clause is not just legal boilerplate. It is an admission of economic reality. If your core gross margin on residential fibre were thick and comfortably under your control, you could afford more generous outage credits as a marketing tool. If your margin is narrow and much of the service risk sits with upstream access providers, you write the contract this way because you cannot insure the customer against every weakness in the underlying fibre chain.

That means the more credible sources of margin lie elsewhere.

One is traffic cost reduction through local peering and an affiliated backbone. If an ISP can keep common traffic on-net or on-exchange, avoid unnecessary international transit, and buy paid capacity only for the traffic that truly needs it, it can meaningfully improve contribution margin per customer. South Africa’s geography makes this more important than it might be in denser or more central markets: subsea path choices, east-versus-west-coast routing, and distance to global platforms still matter. Web sQuad’s status page has historical notices about steering traffic away from a WACS-affected path through other transit providers and about international traffic taking longer east-coast paths during upstream maintenance. Those notices show that transit selection is operationally real, not just theoretical. Peering and route control do not erase cost, but they can suppress it.

A second margin source is business packaging. On the business side Web sQuad is not selling “fibre” so much as it is selling risk transfer. Dedicated uncontended capacity, SLA terms, synchronous bandwidth, provisioning across multiple FNOs and licensed wireless backbones, voice trunks, and human support all allow the operator to charge more than a pure residential line multiple. This is where the group architecture around WECOM matters most. If the backbone, transit and interconnect economics are genuinely under partial group control, then the business unit can package access, transport, failover and voice into a product whose customer thinks in terms of uptime and response time, not just download speed.

A third margin source is operational arbitrage across fragmented fibre networks. Web sQuad’s home-fibre map spans numerous FNOs with different price points, support quality and service quirks. In South Africa, open access created retail competition, but it also created complexity. Many customers do not want to understand the difference between an Openserve authentication issue, a MetroFibre NNI outage, a Vumatel power failure and a misbehaving local router. They want one competent intermediary. The longer Web sQuad can remain credible as that intermediary, the more it can price above bargain-basement competitors without losing every rational customer. The value here is not proprietary last-mile infrastructure; it is proprietary know-how in navigating someone else’s infrastructure mess.

A fourth source is customer density in business-heavy metros, especially Gauteng. Statistics South Africa’s 2025 data show that fixed-home connectivity is much higher in Gauteng and the metros than in many other parts of the country. Web sQuad’s office, AFRINIC registry data and contact pages all centre on Randburg/Johannesburg, while the company also claims POP presence in Johannesburg, Durban and Cape Town. This is exactly the kind of geography that can support an operator built around service-heavy residential users plus SMEs: dense enough for efficient support and backhaul; affluent enough to buy fixed connectivity; fragmented enough that a mid-sized ISP with a good reputation can still win accounts.

A fifth source is attachment products. Voice, hosting, domains and static-IP friendliness are unlikely to transform group economics on their own, but they are efficient ARPU enhancers when sold into the same account. A customer already paying for home or business fibre can often be converted into a higher-margin bundle more easily than a cold prospect can be acquired. The voice page’s emphasis on number porting and saving money on calls makes the target segment obvious: households and SMEs considering a landline replacement or a hosted telephony layer.

Customer support is the softest and yet perhaps most monetisable of these assets. Public MyBroadband discussions over years show a visible Web sQuad representative responding directly to routing questions, FNO escalations, IPv6 issues, latency anomalies and provisioning snags. Users compliment after-hours support, activation assistance and direct troubleshooting. One user, frustrated with Vox and Telkom, said Web sQuad’s forum representative went the extra mile, after which the user cancelled existing accounts and moved to Web sQuad. These anecdotes are not scientific, but in fibre retail they matter because support quality functions as both churn defence and acquisition marketing. When many ISPs rent the same last mile, the human layer becomes economically real.

Still, margin from support is subtle. A local helpdesk is not itself a profit centre. It becomes margin only if it raises retention enough, increases cross-sell enough, or supports enough premium accounts to outweigh its cost. The public evidence suggests Web sQuad understands that game. It does not prove that the game is being won.

Where the model can break The strongest bearish case against Web sQuad is that much of its customer experience depends on infrastructure it does not own. The operator’s own status archive is full of the relevant reminder. There are resolved notices for MetroFibre general outages and major NNI incidents, WeFNO backhaul outages, Openserve authentication and throughput failures, Sibaya Connect network issues, and a long series of Vumatel fibre breaks, hardware failures and power failures in particular suburbs. An operator can be excellent at escalation and still lose goodwill if the underlying vendor fails often enough. That matters more in South Africa than in markets where fibre ownership and retail identity are more tightly integrated.

The company’s public posture on those incidents is rational but revealing. In the MyBroadband thread, Web sQuad’s representative often diagnoses problems as starting in the FNO’s aggregation network and pushes customers to raise tickets so that the ISP can escalate upstream. Support is real, but control is partial. This is the central limit of the retail-only part of the model: if the supplier layers own the physical path and much of the outage risk, the retailer can only monetise reassurance up to a point. Beyond that point, the customer either blames the brand on the invoice, or switches to another ISP on the same network and hopes the experience improves.

Load-shedding, backup power and theft deepen this dependence. ICASA’s 2026 report says telecommunications-theft costs in South Africa jumped from roughly R69.6 million to R201.5 million in 2025, a rise of about 189%, making theft the dominant cost driver in that category. The same reporting cycle shows continued spending on batteries and generators. In its own public forum explanations to customers angry about price increases, Web sQuad pointed directly at energy costs, Eskom instability, debt-funded fibre roll-outs, slower market maturation, and ongoing maintenance plus security costs. That explanation is self-serving in the normal way all supplier explanations are self-serving, but it is also consistent with the regulator’s sector data.

And the key point is not merely that these costs exist. It is that they sit awkwardly across the value chain. Some resilience spending happens at large FNOs. Some happens at colocation sites. Some happens in the ISP’s own POPs and core. Some happens at the customer premises, where a UPS may or may not exist. Electrification risk therefore creates both cost and blame diffusion. When customers complain about “the ISP” during or after load-shedding, the root cause may be a dead FNO battery, a damaged generator, a local neighbourhood fault, or an upstream routing shift. Economically this is nasty because the retailer bears much of the reputation cost without necessarily capturing the infrastructure return that would justify it.

Backhaul and international routing are another pressure point. Historical Web sQuad notices reference emergency work on an upstream provider’s WACS path and warn of increased latency if traffic must be steered via east-coast alternatives. A 2025 MyBroadband exchange about higher latency to Singapore produced a revealing answer from Web sQuad: it suspected odd return paths and noted that Akamai/Linode global routing did not include more expensive routes like SAFE, even though Web sQuad peers with Akamai locally. That is good engineering transparency. It is also an honest reminder that low latency is an output of a large chain of commercial and routing decisions, many of them not made by the end-customer’s retail ISP.

Then there is the competitive pressure that comes from South Africa’s extraordinary over-layering of retail brands. On Evotel’s public marketplace, Web sQuad is present but not consistently the cheapest. In one captured comparison, Web sQuad’s 30/30 FTTH product on Evotel was listed at R599 per month with a R1,500 installation fee, while various rivals on the same network advertised lower monthly prices or free installation. On 200/200, Web sQuad appeared at R1,079 per month with paid installation, alongside competitors at or below that monthly price and often with cheaper or free installation. This does not prove Web sQuad is overpriced everywhere. It does prove that it is not trying to win every sale on headline price alone.

That strategy can be sensible. The problem is that national players can also move upmarket while retaining stronger brand recognition, larger support operations, or bundle advantages from mobile, cloud and enterprise portfolios. Afrihost, VOX, RSAWEB, Cool Ideas, Vodacom and others all appear in the same network-level shopping environments as Web sQuad. MTN’s chief executive said in 2025 that South Africa has already built enough fibre and that MTN’s future fixed-market move will come via partnership or acquisition rather than fresh fibre build. Vodafone’s South African arm, through the Vodacom-Maziv process, has spent years trying to secure stronger fibre positioning. Openserve remains huge. Vumatel remains huge. In other words, the largest players are not leaving this margin pool to smaller specialists out of courtesy.

The supplier side is powerful for the same reason. Openserve’s 2025 reporting said its homes-passed footprint reached 1,378,930. Remgro said Vumatel had more than 2 million homes passed and more than 864,000 subscribers by March 2025. Once access providers are that large, their wholesale pricing and operational rules can dominate the economics of the smaller retail brands above them. It becomes hard for the retailer to keep a large share of incremental value unless it owns some other scarce capability.

There are also softer risks around trust and reputation. The customer-review record is thin and mixed. HelloPeter shows only a very small number of recent reviews, some strongly positive and some strongly negative, which means statistical confidence is low. But the visible complaints that do exist matter precisely because the sample is small and the SME market is word-of-mouth heavy. One complaint in 2024 centred on price increases and “zero consideration for loyal customers.” A 2025 complaint connected to Butylseal made more serious allegations and framed Web sQuad as reckless and dangerous. Those claims are unverified public complaints, not judicial findings, and they cannot prove systemic misconduct. But they do matter economically because disputes with business customers can damage trust disproportionately in a market where the service promise itself rests on trust, speed of response and technical credibility.

Even the positive market chatter contains a warning. Web sQuad is repeatedly praised for being responsive, for offering static IPs without hassle, for good latency, and for clear direct support. Those are all high-touch qualities. High-touch qualities are defensible, but they are not infinitely scalable. If the company grows too fast without widening engineering and support capacity, the very differentiators underlying any premium margin can be diluted. One 2022 exchange in the forum referenced capacity issues that had apparently been “cleared up” enough for a user to run three 4K streams again. That is not a condemning incident. It is a reminder that scale can stress the proposition.

A verdict with caveats So, can Web sQuad convert local peering, support, fibre resale, wireless reach and business density into margin despite South African power risk, backhaul cost, FNO dependence and aggressive national rivals?

The best answer from the public record is yes, but only in a narrow band of the market, and probably less through residential access alone than through a hybrid retail-wholesale-service model.

If one imagines Web sQuad as a pure mass-market FTTH reseller, the answer is much less favourable. The company’s own terms reveal activation subsidies, clawbacks, and broad disclaimers for third-party downtime. Public marketplaces show it is not always the lowest-price option on the same FNO. Outage archives show real dependency on MetroFibre, Vumatel, Openserve and other network owners. Large national brands crowd the same acquisition space. In that version of the story, margins are fragile and probably inferior to what a casual reading of “premium ISP” might imply.

If, however, one reads the public network evidence seriously, a different picture emerges. Web sQuad appears to sit on top of a more consequential network and group structure involving WECOM or Wecom Holdings. AS328137 has a meaningful peering footprint. AS37731 looks like a wholesale-capable backbone node with multiple upstreams and downstream customers. The business page points explicitly at dedicated uncontended services, licensed wireless, and SLA-backed delivery. The voice, hosting and domain products provide attachment opportunities. In that version of the story, the residential line is just the front door; the real margin comes from being the operator that can aggregate many FNOs, peer cheaply, route intelligently, support customers well, and monetise business accounts that need more than a home-fibre login.

That is why the right economic frame is not “Is Web sQuad a good ISP?” but “Where in the stack does Web sQuad capture value?” The public record suggests three candidate answer points.

The first is local traffic economics. Neutral IX participation across Johannesburg, Durban and Cape Town, plus open peering with major content and cloud networks, can materially lower unit network costs relative to an undifferentiated small reseller.

The second is SME and enterprise packaging. Selling uptime, uncontended access, voice portability and local support to businesses is a better path to margin than fighting for every household on price.

The third is group internalisation of wholesale functions. If WECOM is indeed doing a meaningful share of the heavy lifting in backbone, transit and interconnect, then the retail brand may enjoy economics that a standalone small reseller simply could not replicate. But that advantage is only partly visible to the outside world.

The caveat is important enough to state plainly: the public record still does not tell us the one thing investors and creditors would most want to know, namely how much margin Web sQuad or its group actually earns, by segment, after support cost, backhaul, supplier credits, churn, and bad debt. It also does not cleanly show which group entity holds which ICASA licences, whether the retail and wholesale layers transact on arm’s-length terms, how concentrated the business-customer base is, or how much of the founder’s semi-public revenue narrative survives scrutiny against audited numbers. The 2025 policy direction and ICASA inquiry into new individual ECNS licences underline why telecom permissions can matter economically in South Africa, especially when licence transfers themselves have become tradable assets; but the public sources gathered here do not let one map those permissions cleanly onto each Web sQuad or WECOM entity.

That unresolved opacity changes the conclusion only slightly. My commercial reading is that Web sQuad is more economically substantial than its modest retail branding first suggests, yet less protected than a true infrastructure owner. It can probably make money. It may make quite decent money in the right business mix. But it is unlikely to compound margin safely if it relies too heavily on residential fibre resale without continuing to thicken the parts of the model that customers cannot easily price-shop: routing, relationships, backup, voice, support and business-grade accountability.

Evidence ledger Web sQuad business page URL: https://websquad.co.za/business/ Source type: Company page. Supports: The company sells SLA-backed dedicated and business broadband services, advertises up to 10 Gbps, up to 99.5% SLA, and delivery across 15+ FNO and wireless networks. Does not prove: Actual take-up, realised uptime, margins, or the proportion of revenue from business services. Why it matters economically: It shows where higher-margin products are likely to sit if the consumer FTTH layer is thin.

Web sQuad home fibre page URL: https://websquad.co.za/home-internet/home-ftth/ Source type: Company page. Supports: Multi-FNO resale footprint across major South African networks; no-contract, uncapped positioning; acquisition subsidies such as activation and router inclusion. Does not prove: Whether Web sQuad has bargaining power over those FNOs or earns meaningful gross margin on each line. Why it matters economically: It shows that the company is an aggregator of access networks rather than a single-network retailer.

Web sQuad FTTH terms and conditions URL: https://my.websquad.co.za/index.php/knowledgebase/15/FTTH-Home-Fibre---Terms-and-Conditions.html Source type: Company legal terms. Supports: One-month notice, minimum-two-month service, early-cancellation clawbacks, router-return obligations, and explicit refusal to credit FNO-caused downtime. Does not prove: How often these rights are enforced or whether customers typically churn before subsidy payback. Why it matters economically: Few sources reveal retail-margin thinness more clearly than clawback clauses and downtime-credit exclusions.

AFRINIC-derived WHOIS and BGP record for AS328137 URL: https://bgp.tools/as/328137 Source type: Network-registry / routing observation. Supports: AS328137 is an active South African LIR-backed network with RPKI-valid prefixes, substantial peer count, multiple location-labelled prefixes, and registry contact ties to Randburg. Does not prove: Revenue, traffic volume, or whether the retail business is strongly profitable. Why it matters economically: It proves Web sQuad has real network substance, not just a billing relationship with a third-party wholesaler.

PeeringDB entry for AS328137 URL: https://www.peeringdb.com/net/14303 Source type: Industry registry. Supports: Open peering policy, Wecom Holdings organisational link, route-set, company website, and the presence of a formal peering posture. Does not prove: How much traffic is exchanged over those peerings or what savings result. Why it matters economically: Peering policy and organisational identity help distinguish a serious network operator from a commodity reseller.

INX and NAPAfrica member records URL: https://portal.inx.net.za/customer/detail/99 URL: https://ix.nap.africa/index.php/customer/detail/262 Source type: Internet-exchange membership records. Supports: Since-2018 membership and multi-city IX presence with 10G and 20G ports in Johannesburg, Durban and Cape Town. Does not prove: Port utilisation ratios or the exact percentage of total customer traffic served locally. Why it matters economically: Local peering is one of the few ways a mid-sized ISP can structurally lower network cost and improve latency without owning the last mile.

NAPAfrica policy statement URL: https://www.napafrica.net/ Source type: Exchange operator page. Supports: NAPAfrica does not charge membership, port, or cross-connect fees. Does not prove: Total cost of access once colo, transport and equipment are included. Why it matters economically: It explains why exchange presence can matter disproportionately for a traffic-aware ISP in Africa.

AS37731 WECOM routing record URL: https://bgp.tools/as/37731 Source type: Routing observation. Supports: A second Web sQuad/WECOM-linked ASN with multiple upstreams and multiple downstreams, consistent with a wholesale or backbone role. Does not prove: Legal ownership boundaries or internal transfer pricing between WECOM and Web sQuad. Why it matters economically: It suggests the retail margin story may be incomplete without the wholesale affiliate.

Carl Fayolle AFRINIC candidate profile URL: https://mybroadband.co.za/news/wp-content/uploads/2025/06/Candidate_Information_-_Region-Independent_Seats_7_and_8.pdf Source type: Semi-public personal profile / PDF. Supports: Founder-side narrative of Web sQuad’s origin, engineering build, and claimed revenue milestones, plus WECOM’s carrier-grade role. Does not prove: Audited turnover, EBITDA, or current segment mix. Why it matters economically: It is the clearest semi-public statement that the business may be deeper than consumer fibre resale. It must, however, be treated cautiously.

ICASA State of the ICT Sector Report 2026 URL: https://www.icasa.org.za/uploads/files/The-State-of-the-ICT-Sector-Report-of-South-Africa-31-March-2026.pdf Source type: Regulator report / PDF. Supports: Sector-wide theft, vandalism, battery and generator cost pressures, and continued fixed-broadband investment. Does not prove: Web sQuad’s own cost base or resilience spending. Why it matters economically: It sets the cost environment in which all South African ISPs, including Web sQuad, must try to preserve margin.

Stats SA General Household Survey 2025 URL: https://www.statssa.gov.za/publications/P0318/P03182025.pdf Source type: National statistics / PDF. Supports: Fixed-internet penetration remains a minority product nationally but is much stronger in Gauteng and metros. Does not prove: Web sQuad’s actual market share in those areas. Why it matters economically: It helps explain why metro density and SME clusters matter more than heroic national coverage for an operator like this.

Evotel public ISP marketplace and price lists URL: https://my.evotel.co.za/Shopfront/Packages URL: https://evotel.co.za/wp-content/uploads/2024/10/Price-list-template-Octrobe-2024.pdf Source type: Semi-public network marketplace / FNO price list. Supports: Web sQuad competes directly with national and regional ISPs on the same access network and is often not the cheapest visible option. Does not prove: Comparison quality across all networks or customer willingness to pay for better support. Why it matters economically: It shows the intensity of retail price competition above the same last-mile asset.

Web sQuad network-status archive URL: https://my.websquad.co.za/serverstatus.php?view=resolved Source type: Company operational archive. Supports: Regular dependency on FNO incidents, major NNI events, backhaul outages, and upstream transit maintenance. Does not prove: Outage frequency relative to competitors or internal MTTR performance after ticket escalation. Why it matters economically: It shows where service risk sits and why customer churn can be triggered by faults outside the ISP’s direct physical control.

MyBroadband forum threads and review chatter URL: https://mybroadband.co.za/forum/threads/web-squad-isp.1007232/ URL: https://mybroadband.co.za/forum/threads/web-squad-isp-feedback-thread-2.1246333/ URL: https://www.hellopeter.com/web-squad Source type: Informal market evidence. Supports: Real customer perceptions around support quality, pricing pain, latency, capacity issues and trust. Does not prove: Representative customer satisfaction or statistically valid churn drivers. Why it matters economically: In a service business selling largely interchangeable access, informal reputation can change acquisition cost and retention.

What would reprice the margin story The facts that would most sharply change the commercial view are not heroic technical details. They are boring, decisive financial ones.

If audited or credibly leaked evidence showed that Web sQuad’s retail FTTH base churns slowly, pays back acquisition subsidies within months, and reliably cross-buys voice or business add-ons, the margin case would strengthen fast. If, on the other hand, the customer book is mostly price-sensitive residential lines on third-party fibre, with high churn after promotions or after FNO outages, the equity in the model would look thinner.

If public records cleanly mapped which Web sQuad or WECOM entity holds which ICASA licences, and how retail traffic and wholesale backbone costs are allocated internally, one could finally judge whether the group’s network investment is a true structural advantage or just a technically impressive cost centre. The 2025 policy direction on new individual ECNS licences shows why permissions still matter in South Africa: the system has created a tradable layer of telecom rights even in a market already thick with licensees. But the public record here does not yet identify where those rights sit inside the Web sQuad/WECOM constellation.

And if the next two to three years show that South African fibre operators keep raising wholesale access prices while ever-larger national brands push harder into the same open-access footprints, the residential reseller layer will be repriced downward across the board. In that world, the winners will be the firms with true wholesale leverage, dense business accounts, and support customers are willing to pay not to lose. The rest will still sell internet. They will just sell it on someone else’s terms.