The evening test for a regional South African ISP does not begin in a boardroom or a national speed ranking. It begins in a Pretoria North kitchen after the power has just returned. The kettle is back on. A child is trying to reconnect a school tablet. A parent is checking whether the inverter battery has enough charge left for another interruption. On a small shelf near the television, a router and a fibre terminal blink through their restart cycle, while a small backup unit decides whether the household's internet came through the cut or merely restarted after it. The family knows the ritual too well: wait for the power, wait for the router, wait for the line, then decide whether the internet provider has earned another debit order.
For Wan4u, that household scene is not sentimental background. It is the operating model. Wan4u presents itself as an ICASA-approved wireless internet service provider serving clients mainly in Pretoria and Thabazimbi, with wireless broadband, future and value-added fibre services, email and web hosting through a partner, and a long-running support culture aimed at homes and businesses (https://wan4u.co.za/ and https://wan4u.co.za/about). Its WAPA member profile is more direct: WAN4U is listed as a principal Gauteng member, phone 012 546 6100, with a description of a WISP serving mainly Pretoria and surrounding areas, using ICASA-approved products and supporting clients from home users to larger business environments (https://wapa.org.za/members-list?field_wapa_member_pa_region_value=All&field_wm_member_type_value=All&items_per_page=50&page=4&title=). The company is not claiming a national fibre footprint. It is claiming a relationship with places where the last mile is still a technical, power and support problem.
The hard public number that anchors the network evidence is AS328747. AFRINIC RDAP shows AS328747 as active, registered on 16 October 2020 and named ORG-WUC1-AFRINIC, with WAN 4 U cc contacts and Pretoria address data in the registration record (https://rdap.afrinic.net/rdap/autnum/328747). BGP.tools identifies the network as WAN 4 U cc, active under AFRINIC, with one originated IPv4 prefix, 165.73.224.0/21, and no originated IPv6 prefixes in its current public view (https://bgp.tools/as/328747). RIPEstat's announced-prefixes feed similarly showed 165.73.224.0/21 visible for AS328747 in the June 19 to July 3, 2026 observation window (https://stat.ripe.net/data/announced-prefixes/data.json?resource=AS328747). PeeringDB lists Wan4u as a regional Cable/DSL/ISP network, AS328747, 1-5Gbps traffic, mostly inbound traffic ratio, open peering policy, two exchange counts, four IPv4 prefixes in its self-maintained profile and no IPv6 prefixes (https://www.peeringdb.com/asn/328747 and https://www.peeringdb.com/api/net?asn=328747).
That is a modest public routing surface. It is also exactly why the household bill matters. A regional ISP cannot hide fixed costs inside tens of millions of mobile subscribers. It has to recover tower maintenance, customer equipment, line-of-sight inspections, installation labour, routers, batteries, support tickets, exchange and upstream capacity, open-access fibre wholesale charges and the repair bill from a local customer base. Wan4u's own fibre page makes the mechanism unusually visible: an Openserve FTTH 20-50Mbps uncapped fibre plan is advertised at R650 per month with a R1,750 installation fee and a R150 per month wireless backup link; MetroFibre FTTH packages run from 25Mbps at R520 to 1Gbps at R1,320, with a R2,050 installation fee and optional or included backup-link language depending on package class (https://wan4u.co.za/fibre/). A customer does not see an abstract margin model. They see a choice: pay for fibre, pay for backup, accept installation terms, and then judge whether a local provider answers when the line fails.
The choice is harsher because the broader retail market is teaching customers to benchmark every rand. At publication, Fibre Tiger's comparison surface listed 48 South African fibre deals sorted by price, including Afrihost over MTN Fibre 10/10 uncapped at R417 per month, Webafrica over Vumatel Vuma Reach 20/10 at R429, Vox over Frogfoot 25/10 Air at R435, Afrihost over Openserve 30/30 at R497 and Afrihost over MetroFibre 30/30 at R527 (https://www.fibretiger.co.za/). Those comparator prices are not address-level substitutes for every Wan4u customer, and they do not include the same backup or local-support assumptions. But they show the price pressure clearly: a regional provider asking R650 for entry Openserve fibre, R150 for wireless failover and installation fees measured in thousands must sell resilience and repair confidence, not just megabits.
Power cost makes that confidence expensive. ICASA's 2026 State of the ICT Sector report says telecom licensees' battery spending rose from R173.8 million in 2024 to R387.7 million in 2025, generator spending rose from R211.5 million to R426.8 million, and battery purchases rose from 44,708 to 84,829 in the 12 months ending 30 September 2025 (https://www.icasa.org.za/uploads/files/The-State-of-the-ICT-Sector-Report-of-South-Africa-31-March-2026.pdf). Those are sector figures, not Wan4u accounts. They are still the correct backdrop for the Pretoria kitchen. When a customer pays for backup, the hidden bill includes batteries that age, high sites that need power, routers that need replacement, staff who triage faults after outages, and field work when a pole route, fibre drop or wireless link does not return cleanly.
The central cost mechanism is therefore simple, but unforgiving. Wan4u sells ordinary broadband trust in a market where the consumer's tolerance is shaped by power cuts, fibre construction delays, mobile substitution, support fatigue and price comparison. The router has to survive the outage. The fibre or wireless path has to recover quickly enough after the outage. The upstream path has to be bought at a scale that does not punish small traffic volumes. The support queue has to distinguish a dead customer router from a failed FNO drop, a tower power problem, a congested evening sector, a DNS complaint, a payment problem and a true upstream event. The monthly tariff has to pay for all of that while open-access fibre and mobile data products teach customers to shop by headline speed.
Wan4u's own history explains why the company should be read as a regional engineering business rather than as a simple reseller. The about page says the network has been built since 2001, starting with one 64k Diginet internet line, then growing to use multiple fibre connections distributed wirelessly to clients (https://wan4u.co.za/about). The same page lists a 150 km link between Pretoria North and Thabazimbi through Kransberg, using only solar power at the remote mountain range, and describes event and temporary point-to-point wireless projects where fibre was unavailable or would take months to deploy. The exact legacy details matter less than the operating philosophy. Wan4u's public self-description is built around solving the gap between a customer who needs connectivity now and a fixed-line market that may not have reached that address, building, school, business site or farm road on acceptable terms.
Wireless remains the clearest expression of that philosophy. Wan4u's wireless page offers budget uncapped packages at 3Mbps for R350 per month, 5Mbps for R480 and 10Mbps for R650, all with upload speed at 25 percent of download speed and contract distinctions by package. Standard packages include 20Mbps at R550, 40Mbps at R1,050 and 50Mbps at R2,100, while premium packages show higher upload ratios at prices from R950 to R3,100 (https://wan4u.co.za/wireless/). The same page lists a broad coverage cluster across Pretoria North, Montana, Wonderboom, Silverton, Silver Lakes, Thabazimbi, Dwaalboom, Rooiberg, Leeupoort and other localities, and it makes line-of-sight central to the buying process. In its FAQ on the home page, Wan4u says connection begins with an application and documents, then a line-of-sight inspection, and that without line of sight to a tower the company cannot provide wireless internet (https://wan4u.co.za/).
That line-of-sight requirement is a commercial discipline as much as a radio requirement. A national ISP can sell first and discover complexity later because the address is already on an FNO coverage map. A WISP that installs a dish or radio on a customer's building has to know whether the path will hold in weather, whether a mast needs earthing, whether trees or terrain will become a future repair call, whether the customer expects fibre-like latency, and whether the installation distance changes the economics. Wan4u's wireless terms say installation distance longer than 15 km will be evaluated, minimum installation time is two hours, labour beyond that may be billed, contract budget covers up to R3,500 in total costs, qualified installers perform work, lightning earthing is the client's responsibility, and the company cannot guarantee traffic speed between 17:00 and 22:00 Monday to Friday and over weekends (https://wan4u.co.za/wireless/). Those caveats are not small print trivia. They are the economics of keeping a fixed wireless promise in a real geography.
The "open time" feature on the same page shows Wan4u trying to manage capacity with incentives rather than only with complaints. The company says open time disables queues for the current paid speed during specified off-peak periods: Monday to Friday from 10pm to 3pm and weekends from 1am to 6am. The page says speeds during that schedule are hardware dependent and can jump from 50Mbps to 200Mbps, and that the purpose is to encourage usage during non-peak hours and supply less load during peak hours (https://wan4u.co.za/wireless/). This is a rare public glimpse of a regional ISP telling customers the truth: capacity is a shared local resource, and evening demand is expensive. If subscribers shift large downloads away from 5pm to 10pm, the provider can defer some capacity cost and reduce contention. If they do not, either prices rise, support calls increase, or performance expectations break.
Fibre changes the same problem without removing it. Wan4u's fibre page advertises value-added services over Openserve and MetroFibre coverage. It offers Openserve FTTH tiers, Openserve FTTB tiers, MetroFibre FTTH tiers and MetroFibre FTTB tiers, and repeatedly includes a wireless backup link as a priced or package-related option (https://wan4u.co.za/fibre/). MetroFibre's own site says it is an open-access network with partnerships with more than 100 ISPs, and its visible partner carousel includes Wan4U (https://metrofibre.co.za/). That positions Wan4u as the customer-facing layer over infrastructure owned by a fibre network operator. In a clean installation, that can be good economics: the FNO absorbs civil works and access-network capex while Wan4u earns retail margin, support attachment and backup-link revenue. In a bad installation, the customer blames Wan4u for a physical fault another party must repair.
The fibre page's backup-link language is the key to the round lens. The home page says "Fibre internet down? Only R150/pm" and markets a wireless internet backup link that kicks in automatically when fibre goes down (https://wan4u.co.za/). The fibre package table repeats R150 per month for wireless backup links on multiple Openserve and business fibre plans and notes backup availability in selected areas for MetroFibre FTTH (https://wan4u.co.za/fibre/). This is not only an upsell. It is a recognition that South African broadband trust has become multi-path. The customer wants fibre because it is fast, stable and familiar. The customer wants wireless backup because the bill is judged during failure, not during a perfect speed test. Wan4u's opportunity is to be the local operator that can sell both the primary path and the emergency path. Its risk is that the backup path adds another thing to monitor, power and explain.
Power is the first cost layer behind that promise. South Africa's load-shedding crisis has eased sharply from its worst years, but the telecom sector's financial scar remains. The ICASA battery and generator figures above describe the environment in which a regional ISP's tower, exchange path, customer premises and support team operate (https://www.icasa.org.za/uploads/files/The-State-of-the-ICT-Sector-Report-of-South-Africa-31-March-2026.pdf). Backup power is no longer exceptional. It is embedded cost, and it keeps consuming management attention even when the national electricity headline improves.
Eskom's own 2026 update says South Africa had reached 301 consecutive days without interruption in supply, with only 26 hours of load shedding recorded in April and May 2025 during the financial year, and that diesel expenditure was far lower year on year (https://www.eskom.co.za/eskom-marks-300-days-without-loadshedding-as-sustained-generation-performance-maintains-grid-stability-and-energy-security/). The OECD's 2025 South Africa survey is more cautious: it says the suspension of load shedding since March 2024 was notable progress, but the system remained fragile, private solar rose from 1.2GW in 2021 to 6.1GW by 2024, and several power stations were operating below 60 percent capacity (https://www.oecd.org/en/publications/oecd-economic-surveys-south-africa-2025_7e6a132a-en/full-report/reforming-south-africa-s-electricity-sector_05fdccb6.html). The practical effect for Wan4u is that backup cost cannot simply be written off because the national headline improves. Batteries age. Customers remember. Municipal outages, load reduction, cable theft and local transformer failures can still create broadband failure even when national load shedding is not scheduled.
For a household, the power question becomes a device question. MyBroadband's load-shedding backup guide explains the consumer version clearly: fibre users need power for both the wireless router and the optical network terminal, and small UPS or battery systems can keep a home connection online during power cuts (https://mybroadband.co.za/news/energy/419360-cheapest-ways-to-keep-your-internet-connection-online-during-load-shedding.html). Wan4u's support surfaces have to handle that messy boundary. If the FNO is alive, the ISP is alive and the router is dead because the customer has no backup, the customer still experiences "the internet is down." If the customer's UPS is alive but a tower battery has failed, the complaint reverses. If a local pole route or fibre drop was damaged by a storm or repair work, the power story becomes a field-repair story. Trust depends on diagnosis that is faster than the customer's frustration.
That support burden is visible in Wan4u's channels. The support page says customers can use the mobile app to submit tickets, the customer portal to create support tickets, view usage and get account insights, call 012 546 6100, and use Telegram or WhatsApp; after-hours support runs from 5pm to 10pm only (https://wan4u.co.za/support/). The Google Play listing for the Wan4U app says it lets customers view usage, top up data bundles, receive network notifications, create support tickets, perform coverage checks and subscribe or unsubscribe from push-notification channels; the app was updated on 23 June 2026 (https://play.google.com/store/apps/details?hl=en_ZA&id=za.co.wan4u.wan4uapp). The Apple App Store listing describes similar functions and shows a tiny review base (https://apps.apple.com/za/app/wan4u/id1451883977). Support is not a soft brand promise here. It is an operating system for turning intermittent faults into retained customers.
Customer chatter should be handled carefully because it overrepresents people with strong feelings. Wan4u's own site publishes positive testimonials about installation, updates during downtime, long tenure, remote setup in Thabazimbi and technicians resolving intermittent connectivity (https://wan4u.co.za/). A Hellopeter search result for a 2023 review says a customer who teaches online had struggled with intermittent connectivity for months but praised Wan4u staff for working through the issue (https://www.hellopeter.com/wan4u/reviews/wan4u-service-4615895). Facebook's public page preview shows a 78 percent recommendation score from seven reviews (https://www.facebook.com/wan4uWISP/). Downdetector's Wan4u page showed no current problems at retrieval and frames comments as public reports that help others understand issues, not as a representative survey (https://downdetector.co.za/status/wan4u/). These are signals, not proof of system-wide quality. Their value is that they show what customers notice: installation, communication, intermittent connectivity, support responsiveness and outage anxiety.
The market chatter around Pretoria fibre broadens the point. In a Pretoria Reddit discussion, users did not treat "fibre" as a single product; they debated which ISP and FNO combination was available at an address, mentioned Afrihost, Cool Ideas, Openserve and Google reviews, and complained generally that support quality can change over time (https://www.reddit.com/r/Pretoria/comments/19d3uas/fibre_in_pta/). The Reddit thread is anecdotal and not Wan4u-specific, but it captures the South African buying logic. Customers know the retail brand and the physical network can be different. They know a faster package does not guarantee better support. They know the address matters. A local ISP has to win in that cognitive environment, not in a neat product brochure.
That environment is becoming more competitive because fibre has professionalized and overbuilt the older broadband categories. ICASA's 2026 report says total fixed internet and data revenue reached R40.612 billion in 2025, up from R34.966 billion in 2024, with fixed wired broadband revenue at R17.987 billion and other wireless-broadband services revenue at R6.698 billion (https://www.icasa.org.za/uploads/files/The-State-of-the-ICT-Sector-Report-of-South-Africa-31-March-2026.pdf). The same report lists 3.264 million fixed broadband subscriptions, including 3.006 million fibre-to-the-home/building subscriptions and 1.246 million terrestrial fixed wireless broadband subscriptions in its appendix table (https://www.icasa.org.za/uploads/files/The-State-of-the-ICT-Sector-Report-of-South-Africa-31-March-2026.pdf). Fibre is no longer an elite urban curiosity, but fixed wireless remains a large adjacent category. Wan4u sits at the overlap: wireless heritage, fibre resale and backup-path logic.
Market reporting reinforces the scale of the fibre shift. TeleGeography's South Africa FTTH overview reports Vumatel's network passing more than 2.040 million homes and supporting 864,208 end-user connections as of 31 March 2025, while Openserve had 1,453,810 homes passed and 756,409 homes connected in September 2025 (https://resources.telegeography.com/fiber-fever-meet-the-isps-driving-ftth-deployment-in-south-africa). MyBroadband Insights' Q1 2026 FNO ranking says Openserve was the best-scoring major fibre network operator, followed by Vumatel and MetroFibre, and notes that fibre network operators are competing for market share while 5G wireless providers such as Rain, Vodacom and MTN increase substitution pressure (https://mybroadband.co.za/news/fibre/637700-south-africas-top-fibre-network-openserve-beats-vumatel-and-metrofibre.html). WhichVoIP's consumer guide explains the open-access model plainly: in South Africa the company that trenched the cable is often not the company that bills the customer, and switching ISPs on the same FNO can be easier than changing physical networks (https://whichvoip.co.za/compare-fibre-providers/).
For Wan4u, that market structure cuts both ways. Open access lets a regional ISP sell fibre without funding all the trenching. It also commoditizes the retail layer because the customer can compare multiple ISPs over the same physical network. If Wan4u's package on MetroFibre or Openserve looks more expensive than a national brand, the customer asks what the extra rand buys. The answer has to be local coverage knowledge, backup wireless, responsive support, business-grade judgement, installation advice, and a willingness to handle edge cases. If the answer is only "same fibre, different invoice," the margin compresses. If the answer is "same fibre, plus a regional operator that understands your route and keeps a fallback path alive," the model can hold.
The same logic explains why Wan4u's prices should not be compared by speed alone. A 10Mbps budget uncapped wireless package at R650 per month looks expensive beside some entry fibre products, but it may reach an address where trenching is delayed, wayleaves are impossible, fibre has not been deployed, or the customer values quick activation and a known installer (https://wan4u.co.za/wireless/). A 25Mbps MetroFibre FTTH product at R520 per month with own-router language and optional backup link is closer to mass-market fibre pricing, but it carries installation and FNO dependency (https://wan4u.co.za/fibre/). A business fibre tier with R2,050 installation and R150 backup link sits in a different economic category, where downtime may cost more than the price difference. Wan4u's pricing ladder is less a neat segmentation table than a set of answers to address-level constraints.
The revenue stack underneath those packages is easy to underestimate because customers see one monthly price. A wireless customer paying R650 for a 10Mbps budget service is not buying only international bandwidth. The payment has to contribute to the original site survey, mast or bracket work, radio equipment, router setup, customer education, billing, tower lease or ownership cost, sector capacity, backhaul, upstream capacity, payment collection, support and future repair. A fibre customer paying R520, R650 or R800 per month may still create support effort even when the physical line belongs to Openserve or MetroFibre. The installation fee partly protects the provider from a customer who churns before hardware and labour are recovered. The contract term partly protects the same economics. The backup-link fee partly turns risk into recurring revenue. The problem is that every protection mechanism can feel like friction to a buyer trained by national retail brands to expect free installation, free routers and month-to-month flexibility.
That friction is where the household scene becomes a churn model. If the customer believes the provider prevented a bad week from becoming a lost workday, the installation fee and backup charge feel rational. If the customer spends two evenings repeating the same router reset to different support agents, the same charges feel punitive. Regional ISPs often win first on availability and lose later on expectation. A customer who originally accepted wireless because fibre was unavailable may later compare the same wireless bill against an FNO campaign at the end of the street. A customer who originally chose local support may later ask why a larger ISP offers a cheaper fibre package over the same network. Wan4u's defence is not nostalgia. It has to prove that the local bundle lowers the customer's total failure cost.
Support is therefore a cost centre and a retention engine at the same time. Wan4u's after-hours window from 5pm to 10pm lines up with the household period when broadband stress is highest: homework, streaming, gaming, remote work spillover and family messaging all hit the network after normal office hours (https://wan4u.co.za/support/). But those are also expensive hours to staff. If the company overstaffs, payroll damages margin. If it understaffs, tickets age and customers interpret delay as indifference. The app and customer portal help because they structure demand: ticket creation, usage views, notifications and coverage checks can reduce repetitive phone calls (https://play.google.com/store/apps/details?hl=en_ZA&id=za.co.wan4u.wan4uapp). Yet a portal does not climb a pole, realign a dish, swap a failed power supply or persuade an FNO to prioritize a repair. The saving from digitizing support is real only if the field organization remains fast enough.
The FNO handoff is particularly important because it changes who controls the root cause. On Wan4u-owned or Wan4u-operated wireless links, the company has more direct control over site survey quality, radio alignment, sector capacity and field response. On an Openserve or MetroFibre fibre service, Wan4u may control billing, router setup, customer communication and backup service, but the physical fibre fault, activation delay or outside plant repair may sit with the network owner. The customer rarely cares about that boundary. WhichVoIP's explanation of the FNO/ISP split is useful precisely because it describes a market in which the same physical line can be sold by many ISPs, while the FNO still owns and maintains the cable (https://whichvoip.co.za/compare-fibre-providers/). Wan4u has to translate that structure into plain accountability. "We are waiting for the FNO" may be true; it is not always satisfying.
Fixed wireless creates a different handoff, between physics and support. A line-of-sight inspection can reject a sale that would have produced short-term revenue but long-term complaints. That discipline is good economics because bad links consume support capacity and damage reputation. It is also hard sales culture because a national mobile router or satellite product may tell the customer yes where a WISP says no. Wan4u's public language around free line-of-sight inspection, longer installation distances, earthing responsibility and peak-hour limits is unusually candid (https://wan4u.co.za/wireless/). The commercial challenge is to make candour feel like professionalism rather than limitation. The best customer for that message is someone who has already learned that a cheap connection is expensive if it fails during the moments that matter.
High-site operations also turn weather and power into accounting. The 150 km Pretoria North to Thabazimbi link described on Wan4u's about page, using solar power at a remote mountain range, is the kind of story that makes regional WISPs valuable in difficult terrain (https://wan4u.co.za/about). It is also a reminder that remote resilience has operating cost. Solar equipment must be secured, batteries replaced, radios protected, masts maintained, paths monitored and spares stocked. A lightning event can become a customer outage, a truck roll and a margin hit. A tree, new building or changed rooftop installation can turn a once-good link into an intermittent problem. The customer only sees internet reliability. The provider sees the inventory of small risks that must be priced into an affordable local service.
Upstream and peering economics add a quieter form of scale pressure. PeeringDB's 1-5Gbps traffic band and mostly inbound ratio suggest a regional eyeball network that consumes content-heavy traffic from larger networks rather than exporting comparable volumes (https://www.peeringdb.com/api/net?asn=328747). That is normal for residential and small-business broadband. It also means the provider's traffic mix is shaped by video, cloud updates, gaming downloads, social apps and remote-work platforms. Direct peering at Johannesburg exchanges can improve path quality and reduce dependence on transit for popular traffic, but it does not eliminate the need for upstreams, monitoring and capacity planning. Wan4u's own peering policy says both parties bear the cost of reaching common physical locations and that direct peering gives more control over path preference and capacity management (https://wan4u.co.za/peering-policy/). For a regional ISP, the question is not whether peering is technically elegant. It is whether the traffic savings and quality gains justify the engineering time and cross-connect cost.
This is one reason IPv6 and routing transparency are not only technical footnotes. A small network can run well with a conservative public footprint, but the market around it is becoming less forgiving. More home devices, cameras, consoles, VPNs and cloud services increase pressure on address management and support scripts. More business customers ask about static addresses, remote access, firewall behaviour and SLA language. More sophisticated users know how to check BGP.tools, PeeringDB or Hurricane Electric before trusting a provider (https://bgp.tools/as/328747 and https://bgp.he.net/AS328747). Wan4u does not need to publish a national-carrier engineering manual. But public clarity on route security, IPv6 plans, maintenance windows and incident history would strengthen the trust premium that a regional ISP has to defend.
The buyer mix matters because not all broadband failures have the same cost. A household streaming video may tolerate a backup link that is slower but alive. A teacher, accountant, small manufacturer or medical office may treat a two-hour outage as lost income or service failure. A school or non-profit institution may care less about headline speed than predictable support and low total cost. Wan4u's first-party examples around events, schools, local organizations and remote Thabazimbi coverage point toward this mixed customer base (https://wan4u.co.za/about and https://wan4u.co.za/). That mix can stabilize revenue if the provider prices each use case honestly. It can also complicate operations if consumer-grade and business-critical expectations enter the same support queue.
The final pricing question is whether the customer understands the repair bill before the repair. If a provider waits until a failure to explain that lightning earthing was excluded, that the backup link is selected-area only, that a fibre repair depends on the FNO, or that evening wireless speeds are not guaranteed, the explanation arrives as an excuse. If those limits are explained at sale, they can become part of a resilience design: what is powered, what is backed up, who repairs which segment, what happens during peak hours, and what the customer should expect when the main path fails. Wan4u's pages contain many of those disclosures already (https://wan4u.co.za/fibre/ and https://wan4u.co.za/wireless/). The strategic question is whether they are used as living sales and support tools, not only as web-page caveats.
The cost of repair is the second layer behind the bill. South African telecoms infrastructure still carries theft and vandalism exposure. TechCentral, citing ICASA's latest sector report, reported that theft costs rose from R69.6 million in 2024 to R201.5 million in 2025, a 189 percent increase, while battery and generator spending also more than doubled from the prior year (https://techcentral.co.za/theft-and-power-cuts-hammer-sa-telecoms-operators/279869/). Engineering News reported the same ICASA-based theft-cost increase and tied it to changing security challenges for telecommunications operators (https://m.engineeringnews.co.za/article/theft-electricity-outages-still-weigh-on-telecommunications-operators-2026-05-01). These numbers are sector-wide, not a finding about Wan4u. But a regional ISP is more exposed to the psychology of repair because each outage lands in a narrower local reputation field. A customer who sees a technician at the pole or tower knows whether the repair happened quickly.
Backhaul and upstream dependence are the third layer. BGP.tools lists Wan4u's upstreams as Network Platforms (PTY) LTD and SEACOM Limited in its current view (https://bgp.tools/as/328747). Hurricane Electric's AS328747 page similarly shows one originated IPv4 prefix, two internet exchanges, 24 observed IPv4 peers at retrieval, Network Platforms and SEACOM as observed peer/upstream names, and exchange presence including JINX and NAPAfrica IX Johannesburg (https://bgp.he.net/AS328747). Wan4u's own peering policy says it has an open peering policy subject to technical, commercial and legal requirements, offers peering on AS328747, can peer at internet exchanges listed in PeeringDB, prefers direct BGP sessions for traffic exchange and generally prefers private peering above 1Gbps for networks with sufficient traffic (https://wan4u.co.za/peering-policy/). This is the public skeleton of an ISP that is trying to reduce avoidable middlemen while still depending on upstream and exchange reach it does not fully control.
The absence of public IPv6 origination in BGP.tools, PeeringDB and Hurricane Electric is worth watching, but not overstating. Some small and regional ISPs defer IPv6 because customer equipment, support processes, legacy applications and address scarcity trade-offs do not force immediate migration. But South Africa's fixed broadband market is becoming more professional, and customer expectations increasingly include gaming, streaming, remote work, VPNs, cloud cameras, smart-home devices and business services. The more Wan4u wants to sell business-grade trust, the more its public network posture matters: route hygiene, RPKI, IPv6 roadmap, exchange capacity, upstream diversity, DNS resilience and incident communication all shape the trust premium. The public data does not prove weakness. It identifies where more transparency would improve confidence.
Regulation adds another fixed cost and another trust signal. ICASA's 2020 class licence lists include WAN 4 U cc under both C-ECS and C-ECNS class licence categories (https://www.icasa.org.za/uploads/files/C-ECS-Licensees-May-2020.pdf and https://www.icasa.org.za/uploads/files/C-ECNS-Licensees-May-2020.pdf). Wan4u's own home page displays ICASA licence numbers 1863/CECNS/APR/2021, 1864/CECNS/APR/2021, 1863/CECS/APR/2021 and 1864/CECS/APR/2021, plus WAPA membership W085 / 2010 and a statement that it adheres to WAPA's code of conduct (https://wan4u.co.za/). Its about page separately states ECNS:0171/CECNS/NOV/10 and ECS:0171/CECS/NOV/10, also WAPA W085 / 2010 (https://wan4u.co.za/about). The numbering difference between older and current site references is a reminder that licence identifiers should be checked directly before legal reliance, but the broader point is firm: Wan4u operates in a licensed sector where compliance is part of the product.
The compliance burden is not abstract. ICASA and South African government notices have previously listed WAN 4 U CC among licensees required to submit outstanding compliance documents and information (https://www.gov.za/sites/default/files/gcis_document/202308/49076gen1938.pdf and https://www.icasa.org.za/uploads/files/Non-Responsive-Licensees.pdf). Those notices are historical administrative signals, not proof of current non-compliance. They matter because small ISPs often run close to operational load: customer faults, installations, spectrum coordination, licence filings, privacy requirements, billing and support all compete for management attention. For a buyer, the trust question is not only "does the internet work tonight?" It is also "is this provider sufficiently organized to remain available, legal and responsive over the next contract period?"
Wayleaves, poles and municipal permissions form the fourth layer, especially where fibre and wireless meet. Wan4u's own article list includes a June 2024 piece arguing that fibre deployment in South Africa depends on wayleaves for trenching roads and public spaces and for aerial infrastructure (https://wan4u.co.za/articles/). That is a company-authored opinion signal, not independent evidence of any municipality's conduct. But it reflects a real bottleneck: fixed broadband is not purely electronic. It needs permission to pass over, under and through a place. A regional ISP that understands a suburb, a business park, a tower site or a farm road can solve problems national process may miss. It can also be trapped by a municipal delay or an FNO repair queue outside its direct control.
This is why Wan4u's customer dependence is broader than residential entertainment. The about page says the company has supplied internet for LAN events, temporary business links, schools and non-profit institutions, and a 150 km solar-powered link into the Thabazimbi/Kransberg context (https://wan4u.co.za/about). The home page says Wan4u is currently supporting over 40 non-profit organizations across its region, including areas as far as Thabazimbi (https://wan4u.co.za/). Those claims should be read as first-party brand positioning, but they show the type of customer that can make a regional ISP sticky. A household can switch on price. A school, rural institution, small business, event organizer or online worker may value a provider that knows the site, the link and the backup plan.
The geography also changes the meaning of "ordinary" broadband. Pretoria North suburbs, business premises, smallholdings, peri-urban edges and Thabazimbi-area sites do not all buy connectivity in the same way. In a dense fibre suburb, the customer may be choosing between retail ISPs on a line that is already installed. In a more difficult wireless address, the real decision may be whether a mast, roof mount or high-site path can be made dependable at all. In a business setting, the relevant comparison may be the cost of one missed card-payment day, one failed online class or one payroll interruption, not the difference between two headline monthly tariffs. That gives Wan4u room to sell engineering judgement, but only where the buyer believes the company can diagnose the specific place better than a national call centre.
There is a cultural dimension to that sale as well. South African customers have learned to improvise around infrastructure: mini UPS units, inverter batteries, prepaid mobile data as fallback, WhatsApp support groups, neighbourhood outage rumours, FNO coverage checks and screenshots of speed tests. A regional ISP operates inside that informal operating room. It is judged not only by formal SLA language, but by whether the customer hears about planned downtime, whether a technician explains the real fault, whether the provider admits when an upstream issue is outside its hands, and whether the next invoice feels consistent with the inconvenience suffered. Wan4u's public testimonials emphasize exactly those human factors: being kept up to date, remote setup help, friendly technicians and work on intermittent connectivity (https://wan4u.co.za/). The risk is that these strengths are labour-intensive and hard to scale; the advantage is that larger competitors often struggle to make support feel local.
That creates a narrow but defensible position if Wan4u keeps the promise specific. It should not need to be the cheapest provider in every suburb. It needs to be the provider that knows when wireless is the right primary path, when fibre needs wireless backup, when a customer should buy more power protection, and when a sale should be refused because the path will not hold.
At the same time, trust can be fragile because the customer compares against alternatives that did not exist when WISPs were the only fast option. Fibre is one alternative. Fixed 5G is another. Mobile routers are another. A customer near a strong MTN, Vodacom, Rain or Telkom signal may treat wireless broadband as substitutable if support disappoints, even if performance differs. MyBroadband's fibre ranking explicitly notes growing competition from 5G wireless network providers such as Rain, Vodacom and MTN (https://mybroadband.co.za/news/fibre/637700-south-africas-top-fibre-network-openserve-beats-vumatel-and-metrofibre.html). The risk for Wan4u is that wireless becomes the backup product rather than the primary product in more suburbs. The opportunity is that Wan4u can own that transition by selling backup link, fibre advice and field support before a national mobile brand turns the customer into a purely price-led buyer.
The economics are therefore not a question of whether wireless or fibre is "better." They are a question of which failure mode the customer is buying protection against. Fibre protects against radio line-of-sight problems, sector contention and weather sensitivity. Wireless protects against fibre cuts, FNO repair queues, landlord delays and address gaps. Mobile protects against fixed-network failure, but can suffer congestion, coverage variation and data-pricing constraints. A good regional ISP can stitch those realities into a sensible household or business plan. A weak one sells headline speeds and leaves the customer to discover the gaps during the next outage.
The strongest positive reading of Wan4u is that it has built exactly that stitching capability. Its public evidence shows a long local history, a licensed WISP identity, a real AS number, exchange presence, a WAPA listing, official app and support channels, explicit fibre backup products, a broad wireless coverage list, open-access fibre partnerships and public language about education and support (https://wan4u.co.za/about, https://wan4u.co.za/wireless/, https://wan4u.co.za/fibre/, https://wan4u.co.za/support/ and https://wan4u.co.za/peering-policy/). Those are meaningful assets for a regional ISP. They allow Wan4u to say it is not just reselling bits. It is selling local diagnosis, local installation and a plan for when the primary path fails.
The strongest skeptical reading is that every one of those assets carries cost. A real AS needs upstream and routing competence. A wireless network needs tower access, high-site power, radios, spectrum discipline, line-of-sight inspections, lightning and earthing controls, and customer expectation management. Fibre resale needs FNO coordination, migration handling, billing clarity and repair escalation. Backup links need monitoring and power. Support channels create a queue that must be staffed. Positive testimonials help, but they also raise the expectation that someone will work "tirelessly" when connectivity is intermittent. In a market where entry fibre can be sold in the R300-R600 range by national brands or larger ISPs, a regional provider's trust premium is earned one incident at a time.
What facts would change the view? First, subscriber and churn data would show whether Wan4u's local trust is compounding or merely defending a legacy base. Second, uptime, mean-time-to-repair and ticket-resolution data by product would separate wireless faults, FNO faults, power faults and customer-equipment faults. Third, gross margin by product would reveal whether the R150 backup link is a profitable resilience layer or an underpriced support burden. Fourth, upstream contracts, exchange capacity, RPKI status, IPv6 plans and route monitoring would clarify network resilience. Fifth, customer-acquisition cost and installation-recovery periods would show whether new fibre and wireless customers pay back before they churn. Sixth, public clarity on licence references and current compliance status would reduce administrative uncertainty.
The base case is that Wan4u is a credible regional ISP with a real operating niche, not a scale winner. Its value is not that it can outspend Vumatel, Openserve, MetroFibre, MTN, Vodacom, Rain, Afrihost, Vox or Cool Ideas. It cannot and does not need to. Its value is that it can understand why one Pretoria household needs fibre with a wireless fallback, why one Thabazimbi customer needs a line-of-sight inspection rather than a coverage-map promise, why one small business will pay for an installation that a cheaper consumer package does not solve, and why one support call after an outage decides whether the customer keeps trusting the bill.
The risk is that the same intimacy becomes a ceiling. If fibre overbuild reaches more of Wan4u's best wireless suburbs, wireless ARPU may face pressure. If mobile fixed wireless keeps improving, customers may use cellular routers as their own backup rather than paying a local ISP for one. If support queues grow faster than staff and field capacity, local reputation can turn quickly. If upstream or power resilience is not transparent, sophisticated customers may migrate to larger ISPs with clearer engineering disclosure. If installation and repair costs rise faster than household budgets, the monthly bill starts to feel like a tax on South Africa's infrastructure failures rather than a trusted local service.
But the household after load shedding does not buy an industry forecast. It buys continuity. It wants the router and backup battery to behave, the pole route or tower path to hold, the FNO fault to be chased, the support ticket to be answered, the bill to be explainable, and the connection to be ready before school, work, streaming, payroll or a video call starts again. Wan4u's economics live inside that ordinary expectation. The company matters because it shows the repair bill behind broadband trust: power, last mile, upstream, support, fibre overbuild and wireless substitution all arrive at the same kitchen shelf, where a blinking router decides whether a regional ISP still earns its place.