Wholesale fibre and cloud adjacency in South Africa are no longer markets for raw bandwidth. They are markets for continuity. The buyer is not paying only for Mbps; it is paying to keep branches online, route transactions to cloud platforms without unpredictable public-internet exposure, absorb local power failures, satisfy procurement rules, and avoid the economic loss of a broken enterprise circuit. That is the correct lens for Liquid Telecommunications South Africa (Pty) Ltd: not a consumer fibre story, not just a legacy Neotel asset, and not simply a subsidiary name inside a pan-African telecoms group. Liquid South Africa is best understood as an infrastructure seller in a market where the physical cost of reliability remains high, while the price of bandwidth is under constant pressure from dense fibre competition.
The core economic question is whether Liquid can still sell reliability at a premium in a South African market with powerful rivals, improving national power availability, open-access fibre alternatives, and aggressive cloud/data-centre growth. The answer is selective. Liquid has real infrastructure credibility: licensed electronic communications rights, visible AFRINIC/BGP resources, a South African enterprise service catalogue, wholesale products, cloud interconnect products, data-centre adjacency through the wider Liquid/Africa Data Centres ecosystem, and a South Africa segment that is material in Liquid Telecommunications Holdings’ audited group accounts. But the evidence does not support treating Liquid South Africa as an unchallengeable toll road. For buyers, Liquid is strongest where the requirement is multi-site enterprise connectivity, wholesale carrier service, African reach, cloud private connectivity, and public-sector-grade contracting. It is weaker where the requirement is a simple single-metro access circuit that DFA, Openserve, Vodacom Business, MTN, Vumatel/Maziv partners, or other last-mile providers can price against. Liquid’s commercial problem is that reliability is valuable, but only if it can be proved link by link.
The business being priced is avoided failure
Enterprise fibre economics differ from residential fibre economics because the buyer is not optimizing entertainment bandwidth. It is pricing downtime. For a bank branch, retailer, call centre, hospital, public agency, school network, logistics depot, or cloud-dependent software company, a fibre line is a production input. A cheaper line that fails during trading hours is not cheaper if it stops card payments, call routing, VPN access, cloud applications, CCTV feeds, ERP systems, or public-facing services. The correct unit of analysis is not rand per Mbps; it is rand per available, supportable, contractually governed path.
That is the gap Liquid tries to monetize. Its South African Dedicated Internet Access product is positioned around guaranteed performance, symmetrical speeds, QoS, service levels, IPv4/IPv6 capability, and backbone connectivity into pan-African and global fibre routes connected to major subsea systems. Its own product page describes SLA tiers and both standard and 95th-percentile-style pricing options, which matters because these are enterprise-carrier pricing mechanics rather than mass-market broadband packaging. Liquid’s Business Internet Access launch in South Africa, announced in 2025, appears to move down-market from full dedicated access into a more cost-conscious business product with a dedicated line, SLA, 24/7/365 helpdesk, DDoS visibility reporting, and uptime claims “up to 99%.” The commercial point is clear: Liquid wants to segment buyers by required reliability and willingness to pay, not merely sell one undifferentiated internet line.
That segmentation is economically necessary in South Africa. The country has a sophisticated enterprise telecoms market, unusually dense data-centre and cloud infrastructure by African standards, and competitive wholesale fibre pricing, but also structural power-cost pressure and local operational risk. Xalam’s South Africa data-centre market brief describes South Africa as a regional economic and connectivity hub with mature digital infrastructure and deep cloud/enterprise demand, while also flagging intense competition, pricing pressure, fossil-heavy energy, high power costs, and possible oversupply risk in data-centre capacity. A fibre operator selling only commodity bandwidth gets dragged into price competition; an operator selling continuity, cloud adjacency, compliance-ready procurement, and multi-country reach has more room to defend margin.
Identity first: Liquid SA is a regulated, BGP-visible Neotel-lineage operator
The entity in scope is Liquid Telecommunications South Africa (Pty) Ltd. Public supplier and policy documents tie that legal name to the trading style Liquid Intelligent Technologies. A South African supplier profile lists “Liquid Telecommunications South Africa (Pty) Ltd t/a Liquid Intelligent Technologies,” with registration number 2004/004619/07 and a Gauteng location. Liquid’s South African Acceptable Use Policy uses the same legal identity and trading style, and identifies the relevant network/service environment as including products such as IP Transit, Dedicated Internet Access, Broadband Internet Access, voice, broadband, and related services. This is a useful identity anchor because “Liquid” is also a pan-African brand, a group of companies, a data-centre affiliate ecosystem, and a historical successor to Neotel assets; the legal entity must not be conflated with every Liquid-branded business on the continent.
The regulator trail supports the same identity. ICASA records show Liquid Telecommunications South Africa (Pty) Ltd in the context of Individual Electronic Communications Network Service and Individual Electronic Communications Service licence renewals. Another ICASA document says Liquid acquired licences through a transfer-of-control process approved by ICASA in 2016. This matters commercially because enterprise and wholesale fibre services depend on licensed network rights, wayleave execution, interconnection, numbering or service authority where relevant, and the regulatory legitimacy needed to sell to carriers and public-sector buyers. It does not by itself prove current market share, service quality, or route diversity.
The network-resource trail points to Neotel lineage. BGP.tools records AS36937 with the as-name “Neotel-AS,” the description “Liquid Telecommunications South Africa (Pty) Ltd,” AFRINIC organization reference ORG-NPL2-AFRINIC, country ZA, and a Midrand/Johannesburg address. That does not prove headquarters in a company-law sense, and the AFRINIC country field should be treated as a resource/member/service-area signal, not as a full corporate registration file. But it does prove that the South African Liquid/Neotel network identity is publicly visible in global routing records.
This distinction matters because a careless analysis can either overstate or understate Liquid South Africa. Overstatement happens when every claim about Liquid’s pan-African fibre system, Africa Data Centres, Cassava Technologies, or AS30844 is attributed directly to Liquid Telecommunications South Africa (Pty) Ltd as if all assets sit inside that exact company. Understatement happens when AS36937 is treated as a small isolated ASN without recognizing the wider Liquid group network, transit, cloud, and data-centre ecosystem in which South African enterprise services are sold. The right conclusion is narrower: the South African entity is real, regulated, BGP-visible, and historically linked to Neotel; the broader commercial proposition is strengthened by group assets, but each asset claim needs entity-level care.
What Liquid actually sells: enterprise access, wholesale routes, and cloud private paths
Liquid’s South African product set is not structured like a retail ISP’s catalogue. The official pages point to Dedicated Internet Access, enterprise Ethernet, IP VPN, CloudConnect, wholesale IP transit, metro access circuits, last-mile access, SD-WAN enablement, and international wholesale products. Liquid’s enterprise page describes Ethernet branch connectivity over MPLS and Ethernet private-line variants, CloudConnect using fibre and data centres, ExpressRoute connectivity to Azure regions including South African Azure, IP VPN products, and local/Africa/global reach. Its wholesale products page describes an open-access network of more than 116,000 km across the Liquid group, IP transit, CloudConnect, metro access, protected leased lines, Ethernet, IP VPN, last-mile, SD-WAN, and international wholesale.
Economically, those products sit in three overlapping markets. The first is enterprise connectivity: the buyer wants a site connected to the internet, a branch connected to head office, or a private network connected to cloud. The second is carrier/wholesale infrastructure: the buyer is an ISP, mobile operator, systems integrator, or service provider buying access, backhaul, transit, or capacity. The third is cloud adjacency: the buyer wants lower-latency, private, predictable paths to Microsoft Azure or other cloud platforms, often via carrier-neutral data centres. Liquid’s wholesale page explicitly names tier-one and tier-two carriers, service providers, systems integrators, and mobile operators as partner/customer categories.
This product architecture changes the economics. A commodity fibre access provider earns margin by passing more premises, filling ducts, and increasing take-up on a relatively standardized network. Liquid’s enterprise/wholesale model earns margin by combining access with routing, SLA, NOC support, cross-border reach, cloud on-ramps, security add-ons, and account-level contracting. The cost base is heavier: fibre capex, leased infrastructure, IRUs, data-network costs, skilled support, power resilience, route protection, interconnection, and sales engineering. But the revenue can be stickier when a customer uses Liquid as a production network rather than a cheap internet pipe.
The support posture is part of the product. Liquid’s customer operations material lists a South African enterprise service desk, escalation processes, root-cause-analysis handling on request, target ticket creation within 15 minutes, and recurring update intervals. These are not guarantees of excellent service in every incident; they are evidence that the service is sold as a managed enterprise dependency. In this market, the customer pays partly for the right to escalate and demand an RFO, not just for fibre strands.
The network evidence: AS36937 is local proof; AS30844 is group reach
Public routing evidence supports a two-layer reading of Liquid’s South African network. AS36937 is the local South African/Neotel-lineage ASN. PeeringDB lists AS36937 as Liquid Telecom SA under Liquid Intelligent Technologies, with regional scope, a network-service-provider classification, IPv4 and IPv6 prefix counts, and presence at NAPAfrica Johannesburg. PeeringDB also says AS36937 no longer accepts new peers and refers peering requests to AS30844. This is commercially meaningful: AS36937 appears to function as a legacy/local enterprise network identity, while the more open peering posture sits at the wider Liquid AS30844 level.
AS30844 is the larger Liquid Intelligent Technologies network identity. PeeringDB lists it with global scope, substantially larger prefix counts, 50–100 Gbps traffic scale, and an open peering policy across many internet exchanges. The page describes Liquid as a voice and data transit supplier servicing African ISPs and mobile GSM networks. That supports Liquid’s claim to be more than a local access seller. It also explains how a South African enterprise customer can be sold a proposition that includes African reach, transit diversity, and cloud-route optionality beyond the local metro.
NAPAfrica data reinforces the same picture. NAPAfrica’s member list shows Liquid Telecommunications Ltd AS30844 at Johannesburg, Durban, and Cape Town with open peering, while Liquid Telecom SA AS36937 appears at Johannesburg with closed peering. BGP.tools’ NAPAfrica Johannesburg page shows AS30844 with much larger visible capacity than AS36937 at that exchange. These figures should not be mistaken for full network capacity or all private interconnects, but they are useful indicators of role: the group ASN is the broader peering/transit instrument; the South African ASN is a visible local network identity.
The evidence proves public reachability and peering posture. It does not prove route diversity to any customer site, the physical path of a last-mile circuit, uptime performance, or whether a given link is self-built, leased, protected, or dependent on another fibre network operator. For procurement and due diligence, that distinction is decisive. A buyer should not accept “Liquid has a large African network” as proof that a specific Sandton, Midrand, Durban, Cape Town, or provincial government site has physically diverse paths. The evidence supports Liquid as an infrastructure-grade operator; the last-mile engineering still has to be verified at the circuit level.
Cloud adjacency: the margin moves toward the interconnect
The growth market is not just fibre to a building. It is private, governed connectivity between enterprise sites, data centres, and hyperscale cloud regions. Liquid’s CloudConnect page says the service uses ports at Africa Data Centres and more than 25 points of presence to connect customers directly to multiple hyperscalers, bypassing the public internet, with redundancy, BGP/routing, VLAN, and IP-addressing design elements. Liquid has also promoted Microsoft Azure ExpressRoute connectivity in South Africa, describing private, predictable, high-performance connections with SLA characteristics and nodes at more than 25 PoPs.
This is where Liquid’s economics can improve. A pure access circuit is exposed to price benchmarking. A cloud-adjacent circuit can be sold as part of a migration, compliance, performance, and architecture decision. Enterprises moving workloads to Azure, using hybrid cloud, or connecting multiple branches into private cloud paths are less interested in a single monthly fibre price and more interested in latency, packet loss, routing control, failover, cross-connect charges, security, and escalation. Liquid’s Microsoft Azure and Azure Stack material frames local cloud as a way to reduce latency, reduce bandwidth cost, and support regulatory or data-residency requirements. The commercial claim is not that Liquid owns the cloud; it is that Liquid can monetize the path into the cloud.
Africa Data Centres strengthens the adjacency proposition, though it must be treated as a related ecosystem asset rather than automatically as an asset of Liquid Telecommunications South Africa (Pty) Ltd. Africa Data Centres says it is part of the Liquid Intelligent Technologies group and describes itself as a carrier- and cloud-neutral pan-African data-centre platform. Its own pages cite more than 50 carriers and major internet exchanges such as JINX, CINX, and KIXP, and its 2018 expansion announcement tied Johannesburg and Cape Town facilities to cloud demand, Liquid CloudConnect, and ExpressRoute.
The Johannesburg/Samrand data-centre expansion is economically relevant because fibre demand tends to cluster around power-dense, cloud-dense campuses. Africa Data Centres announced expansion of its Samrand facility from 10 MW to 40 MW, with the wider platform above 100 MW, and framed South Africa as a de facto sub-Saharan technology and data-centre hub. Xalam’s market brief separately identifies Gauteng as Africa’s largest data-centre market, with South Africa hosting a large share of African live capacity and several cloud regions. In that environment, fibre providers earn better economics when they are present at the campus, in the meet-me room, at the cloud on-ramp, and in the metro route that connects enterprise offices to that campus.
The unresolved question is attach rate. Public sources show Liquid has the products and ecosystem position for cloud adjacency. They do not disclose how many South African enterprise customers buy CloudConnect, what percentage of South Africa segment revenue comes from cloud on-ramps, how much traffic is Azure-specific, or how profitable these circuits are after data-centre cross-connect, leased-fibre, support, and power costs. The direction of value is clear; the quantum is not public.
Power instability: the scarcity premium is changing, not disappearing
South Africa’s power story has shifted. In 2023 and early 2024, load-shedding made backup power a front-line telecom cost. By 2025 and 2026, national load-shedding had sharply improved. CSIR reported that load-shedding energy in the first half of 2025 was down 82% from the first half of 2024, while noting that electricity tariffs had risen far faster than inflation over the preceding decade. Eskom later reported a full year without load-shedding by May 2026, with major diesel-cost savings and improved operational availability, and in June 2026 said South Africa had experienced hundreds of days without interruptions since May 2025.
That improvement matters for Liquid’s pricing. When load-shedding is severe, resilience is a visible scarcity good. Customers can understand paying more for a provider with batteries, generators, NOC discipline, route protection, and data-centre adjacency. When load-shedding fades, the visible premium compresses. The buyer becomes less willing to pay for generic “resilience” language and more likely to demand hard evidence: battery runtime at the PoP, generator coverage, fuel logistics, dual utility feeds, route diversity, SLA credits, and incident-history disclosure.
But the cost does not disappear. Tariffs, local municipal outages, cable cuts, theft, weather, equipment failures, and backup-system maintenance remain. MTN South Africa’s public comments in 2026 are useful as an industry signal: even after the end of national load-shedding, MTN said backup systems remained critical because local grid instability, outages, weather, and infrastructure issues still affect power. Earlier industry reporting during load-shedding described Vodacom, Openserve, Vumatel, and DFA relying on batteries, generators, mobile generators, diesel logistics, and backup runtimes to keep networks online. This is not Liquid-specific proof, but it identifies the cost structure all serious South African network operators face.
Africa Data Centres’ solar PPA is another signal. The company announced a 20-year power-purchase agreement with Distributed Power Africa for 12 MW of solar power, initially for Cape Town and then Johannesburg, as a response to South Africa’s energy crisis. Reuters similarly framed the deal around energy-intensive data centres, power cuts, and backup power costs. For Liquid’s fibre proposition, this matters because data-centre power resilience and network resilience are complements: a private cloud path is only valuable if both the route and the destination facility remain available.
The correct conclusion is that power resilience is moving from emergency scarcity to operating discipline. Liquid can still sell reliability, but the argument must become more granular. In 2022–2024, “we are resilient during load-shedding” was itself valuable. In 2026, the question is whether Liquid can prove lower all-in risk than a cheaper alternative after local power faults, grid tariffs, SLA enforcement, and site-level redundancy are included.
Public-sector demand proves eligibility, not dominance
Public-sector procurement is commercially important because it can create sticky multi-year revenue, large site counts, and credibility in regulated enterprise markets. Public records show Liquid South Africa winning or participating in government work. South Africa’s Department of Justice awarded-tenders page lists Liquid Telecommunications South Africa (Pty) Ltd for PRI-line-related work, with published tender values and B-BBEE levels. A Gauteng Provincial Legislature award notice appoints Liquid Telecommunications South Africa (Pty) Ltd t/a Liquid Intelligent Technologies as internet service provider for a five-year, 60-month term.
The larger public-sector signal is the Eastern Cape/SITA broadband matter reported by MyBroadband. The article describes Liquid and SITA delivering broadband to the Eastern Cape Provincial Government through a contested piggybacking arrangement linked to an earlier Western Cape tender. It reports that a court did not suggest corruption and that the service might be reasonably cost-effective; it also reports competing claims around Telkom’s proposed 10 Mbps service and Liquid’s claimed 71% cheaper pricing for uncontended lines, with an initial plan to connect thousands of facilities at 100 Mbps increasing to 1 Gbps. This is not a clean procurement database entry, and it should not be treated as settled proof of universal price superiority. It is useful because it shows Liquid competing in high-site-count public connectivity where uncontended capacity, procurement mechanics, and price-per-site economics are decisive.
Public-sector work also has risk. Long payment cycles, political scrutiny, tender challenges, SITA framework issues, service-level disputes, and budget changes can turn nominally attractive contracts into working-capital and reputational risk. Liquid’s audited group accounts disclose that one major customer represented more than 10% of group revenue, without naming that customer. That does not prove the customer is South African or public-sector. It does, however, remind the analyst that large telecom infrastructure revenues can be concentrated, and that the loss, repricing, or delayed payment of one major account can matter.
The procurement evidence proves Liquid is eligible and active in public-sector connectivity. It does not prove dominant share, above-average service delivery, or superior economics on every tender. The correct interpretation is narrower and stronger: Liquid is a credible bidder for complex public connectivity, especially when the tender values uncontended bandwidth, multi-site rollout, and enterprise support.
Ownership and balance sheet: group strength, standalone opacity
Liquid South Africa sits inside the broader Liquid Intelligent Technologies/Cassava Technologies ecosystem. Liquid’s official pages say Liquid Intelligent Technologies is a business of Cassava Technologies, and the “Our Story” page describes Liquid as present in more than 20 countries, operating a fibre network above 116,000 km, and providing cloud and cyber services through partnerships. This supports brand/control context, but it is not a full standalone ownership chart for Liquid Telecommunications South Africa (Pty) Ltd.
The audited financials of Liquid Telecommunications Holdings Limited are more revealing economically, though still not the same as standalone statutory accounts for the South African entity. The 2025 audited financial statements describe the group as trading as Liquid Intelligent Technologies, operating in more than 25 countries, and serving carrier, enterprise, and retail customers. They report group revenue of USD 693.5 million, profit of USD 18.4 million, and note growth particularly in South Africa and Zimbabwe. They also report South Africa segment external revenue of USD 274.6 million and South Africa adjusted EBITDA of USD 91.0 million. That implies South Africa is not peripheral; it is one of the core economic engines of the group.
The same accounts show the asset-heavy nature of the model. Group expenses include large data and network-related costs, staff costs, depreciation, amortisation, finance costs, fibre infrastructure assets, and right-of-use assets related to fibre infrastructure and IRUs. The accounts also record intangible assets related to brand and spectrum from the acquisition of Liquid Telecommunications South Africa (Pty) Limited. In economic terms, Liquid is not a software-light reseller. It is a capital-intensive network and service platform with operating leverage: once fibre, PoPs, systems, and support are in place, incremental high-quality enterprise revenue can be attractive, but underutilized routes, expensive leases, power costs, and debt service can pressure returns.
Debt is part of the story. Liquid’s 2025 audited accounts showed significant borrowings and senior secured notes due in 2026, with going-concern discussion around refinancing. Cassava later announced that Liquid had fully repaid a ZAR term loan and USD revolving credit facility and agreed new ZAR/USD facilities, with Cassava injecting USD 195 million of fresh capital. Cassava also announced a USD 660 million debt financing package including a USD 300 million Eurobond. The updated evidence points to refinancing progress and sponsor support, but not to zero credit risk. Debt-financed fibre economics require steady utilization, disciplined capex, and protection against price compression.
The practical conclusion is that Liquid South Africa has group support and material segment economics, but public records do not provide a clean standalone income statement, balance sheet, customer list, or asset register for Liquid Telecommunications South Africa (Pty) Ltd. Any counterparty relying on Liquid for mission-critical service should evaluate both the local contracting entity and the group support arrangements behind it.
Unit economics: reliability is a bundle of fixed cost, utilization, and proof
The unit economics of enterprise fibre start with high fixed costs. Fibre routes require trenching or leasing, rights of way, ducts, poles or metro access arrangements, electronics, PoPs, spares, NOC systems, field teams, customer premises equipment, software platforms, IP resources, interconnects, and sales/support labour. Once built, the marginal cost of more traffic can be low over short intervals, but only if capacity is not congested, support is not overwhelmed, and power/security costs are controlled. This is why scale matters, but scale alone is not enough.
Liquid’s product documents show multiple monetization layers. Dedicated Internet Access monetizes guaranteed bandwidth, symmetrical service, QoS, SLA levels, and pricing structures such as standard committed pricing and 95th-percentile billing. Ethernet and IP VPN products monetize private branch connectivity. CloudConnect monetizes private hyperscaler paths and design work around BGP, VLANs, routing, and redundancy. Wholesale products monetize other operators’ need for IP transit, metro access, last-mile, protected lines, international capacity, and cross-border reach.
The cost side is equally important. Network-related costs in the group accounts are large. Fibre infrastructure and IRU/right-of-use assets show that Liquid’s economics include both owned and contracted infrastructure. Power resilience adds another layer: batteries, generators, maintenance, fuel, monitoring, security, and replacement cycles. In a severe load-shedding environment, these costs can be passed through as a reliability premium; in a calmer power environment, customers resist paying for them unless the service evidence is strong.
The margin is created when Liquid can combine scarce capabilities: a protected metro path, a private cloud on-ramp, a cross-border route, a credible NOC, a regulatory/procurement-ready contract, and support commitments that a smaller ISP cannot match. The margin is destroyed when the buyer sees the service as interchangeable with a DFA-based access circuit, an Openserve fibre line, a Vodacom Business package, an MTN enterprise circuit, or a Vumatel/Maziv last-mile service resold by another ISP. The economics therefore depend less on Liquid’s total kilometres of fibre and more on the buyer’s specific topology.
A useful way to price the service is to ask what failure the line prevents. If the buyer needs only office internet, Liquid competes in a crowded access market. If the buyer needs a dual-homed, SLA-backed, cloud-connected, multi-branch network with escalation and route engineering, Liquid’s group architecture matters. If the buyer needs public-sector coverage across many facilities, procurement ability and rollout governance matter. If the buyer needs carrier capacity into or across Africa, Liquid’s broader transit and backbone story matters. Same fibre, different willingness to pay.
Competition: the Mbps price is being compressed from every side
Liquid’s competitive environment is severe. DFA is a major open-access infrastructure competitor. DFA describes itself as a leading open-access fibre infrastructure provider with more than 20,000 km of dark fibre covering about 85% of metro areas in key South African cities. It leases infrastructure to public sector entities, mobile operators, fixed and wireless operators, ISPs, and enterprises, and it operates a build/lease/maintain model. For a buyer that mainly needs metro dark fibre or access to a route, DFA is a direct benchmark against Liquid.
Maziv adds another layer. Maziv combines Vumatel, DFA, and Herotel under CIVH-linked ownership, with Vuma passing more than 2 million homes, DFA serving mobile networks, systems integrators, hyperscalers and ISPs, and Herotel reaching smaller towns and cities. Remgro’s infrastructure disclosure reports DFA and Vumatel revenues and operating profit, and says Vumatel is an open-access FTTH market leader with about 32% market share, more than 2 million homes passed, and more than 864,000 subscribers. Although Vumatel is mainly a residential/open-access FTTH brand, the combined Maziv/DFA footprint affects enterprise economics because ducts, routes, wholesale terms, and cross-sell opportunities shape last-mile competition.
Vodacom’s Maziv transaction makes the competitive map more consequential. The Competition Tribunal initially blocked the proposed Vodacom/Maziv transaction, citing competition and public-interest concerns, while later Reuters reporting said the Competition Commission dropped opposition after revised terms and remedies. Vodacom’s FY2026 reporting says the acquisition of a 30% stake in Maziv was finalized in December 2025, with Maziv described as South Africa’s largest FTTH/FTTS player and a major share of homes passed. For Liquid, the risk is not just one more competitor. It is a mobile operator with enterprise customers, spectrum, backhaul demand, and balance-sheet scale becoming economically aligned with a major fibre infrastructure platform.
Openserve/Telkom remains a benchmark. Telkom’s reporting shows Openserve growing fibre data revenue, expanding homes passed and connected, and benefiting from lower diesel costs as load-shedding faded. Telkom also has a large enterprise arm through BCX, where fibre-related data revenue remains important despite broader revenue pressure. Openserve’s network depth means Liquid cannot assume incumbency weakness; the incumbent still has physical access, enterprise history, and wholesale relevance.
MTN and Vodacom Business compete from the mobile-network and enterprise-account side. MTN’s South African service revenue and data growth show a large operator with capital, enterprise relationships, and incentive to bundle fixed connectivity with mobile, IoT, security, and managed services. Vodacom Business advertises business fibre with symmetrical bandwidth, dynamic/static IP, and LTE while waiting or as failover, and explicitly uses both self-build and third-party networks including Vumatel, DFA, and Openserve. That hybrid model is dangerous for Liquid because it means large mobile operators do not need to own every strand to compete for the customer relationship.
Competition does not eliminate Liquid’s opportunity; it narrows where Liquid can price above commodity. Liquid’s advantage is strongest when a buyer needs African reach, wholesale transit, cloud on-ramp engineering, multi-site public-sector rollout, or a provider that can combine network and cloud-adjacent services. It is weakest when the buyer can run a reverse auction for a standard access circuit in a well-served metro.
Customer and market chatter: useful signals, weak proof
Unofficial signals are worth reading, but not over-reading. A MyBroadband forum thread from 2019 contains user discussion of a Liquid Telecom outage, including reports of an outage around Bryanston and one user switching traffic to another network. This is not statistically meaningful evidence of systemic unreliability; every network has incidents, and forum threads overrepresent unhappy moments. Its commercial value is different: it shows the buyer behaviour that matters in enterprise networking. Customers with alternatives route around failure; customers without alternatives suffer. Redundancy, RFO discipline, and transparent incident communication are therefore part of the economic product.
A WhichVoIP buyer review, last verified in May 2026, frames Liquid as a stronger fit for multi-country enterprise and complex connectivity than for South Africa-only SMEs, and flags quote-only pricing, commercial transparency, and multi-country contracting complexity as buyer diligence issues. This is not an engineering audit and should not be treated as authoritative performance evidence. It is still a useful market-signal source because it describes how buyers perceive the supplier: credible infrastructure, but diligence-heavy and not necessarily the simplest or cheapest path for smaller local buyers.
These soft signals align with the hard economics. Liquid is not likely to win every customer by being cheaper. It wins when the buyer’s cost of failure, route complexity, or cloud/inter-country requirement is high enough to justify a more engineered sale. When the buyer only wants a low-friction fibre line, Liquid’s sales complexity can become a disadvantage.
Where Liquid has an edge
Liquid’s first edge is credible infrastructure identity. The combination of ICASA licence records, legal/trading-name documents, AFRINIC/BGP identity, PeeringDB presence, and NAPAfrica visibility makes Liquid South Africa a real infrastructure operator, not a thin reseller brand.
The second edge is group reach. Liquid’s official network pages describe more than 116,000 km of fibre, cross-border routes, subsea cable connectivity, a Cape Town-to-Cairo terrestrial link, and hubs in London and Marseille. Its wholesale materials emphasize connections to major subsea systems and internet exchanges across Southern and Eastern Africa where it has PoPs. For a South African business operating into the rest of Africa, that reach is not cosmetic. It can simplify procurement, routing, escalation, and architecture compared with stitching together multiple national providers.
The third edge is cloud adjacency. Liquid’s CloudConnect, Azure ExpressRoute, Africa Data Centres, and NAPAfrica presence give it a credible path into the enterprise cloud migration cycle. This matters because cloud migration changes the buyer’s network budget from “internet access” to “application performance and risk control.” Private cloud paths are stickier than commodity broadband when properly embedded in architecture.
The fourth edge is procurement credibility. Public-sector awards and participation in large provincial connectivity projects show Liquid can operate in bureaucratic, multi-site, tender-driven environments. Smaller operators can undercut price, but they may struggle with scale, compliance, rollout governance, or public-sector contracting requirements.
Where Liquid is exposed
Liquid’s first exposure is price compression. South Africa’s wholesale and enterprise fibre market has too many capable alternatives for a generic premium. DFA, Openserve, MTN, Vodacom Business, and Maziv-linked infrastructure can all discipline access pricing. Xalam’s market brief explicitly identifies competitive pressure and low wholesale prices in South Africa’s data-centre/connectivity environment.
The second exposure is proof burden. Liquid’s public materials show large networks, SLAs, products, and support processes. They do not prove the physical diversity, power backup, historical uptime, or support quality of any individual customer circuit. Buyers should require route maps, PoP/power detail, failover design, SLA credits, escalation matrices, RFO samples, and references from comparable customers.
The third exposure is group-level financial leverage and opacity. The 2025 accounts showed significant borrowings and refinancing pressure, while 2026 Cassava announcements show refinancing and capital injection progress. That is materially better than an unresolved maturity wall, but the public record still does not give a standalone Liquid Telecommunications South Africa (Pty) Ltd financial picture. Counterparties should distinguish group support from local contracting-entity obligations.
The fourth exposure is power-cost pass-through. National load-shedding has improved sharply, but electricity tariffs, diesel logistics, municipal failures, and local outages remain. A network operator that invested heavily in backup resilience may struggle to recover those costs when customers perceive the crisis as over. The resilient operator deserves a premium only where the resilience is measurable.
The fifth exposure is consolidation around rivals. Vodacom’s Maziv stake aligns a major mobile and enterprise operator with a major fibre platform. Even if remedies limit foreclosure, the transaction changes incentives, route access, wholesale behaviour, and enterprise bundling. Liquid’s response must be differentiation by cross-border reach, cloud adjacency, and service quality, not commodity metro pricing.
Category recommendation
Liquid Telecommunications South Africa should be categorized as a selective infrastructure-grade enterprise and wholesale supplier, not as a default cheapest-access provider and not as an uncontested monopoly asset. For a large enterprise, carrier, systems integrator, cloud-heavy organization, or public-sector buyer, Liquid belongs on the shortlist when the requirement includes protected fibre, multiple sites, DIA, IP VPN, Ethernet, cloud private connectivity, African reach, wholesale transit, or formal support processes. For a South-Africa-only SME, a single office, or a commodity internet-access requirement, Liquid should be competed aggressively against DFA-based providers, Openserve, MTN, Vodacom Business, Maziv/Vumatel-based ISPs, and credible regional operators.
The procurement rule should be simple. Buy Liquid where its measurable sufficiency is higher: route diversity, cloud on-ramp proximity, NOC response, SLA enforceability, cross-border reach, public-sector rollout capacity, and power-continuity evidence. Do not pay a Liquid premium for brand, kilometres of group fibre, or generic “enterprise-grade” language. Require evidence at the circuit, PoP, and contract level.
For investors and strategic counterparties, Liquid South Africa is attractive because it sits in the densest enterprise and data-centre market in sub-Saharan Africa, contributes materially to group revenue and EBITDA, and participates in the structural migration from public internet access to private cloud-adjacent networks. The risk is that the same market is also the most competitive. The investment case is not “bandwidth demand rises, therefore margins rise.” The investment case is “complexity rises, and the operator that can prove reliable, cloud-adjacent, multi-site delivery captures the premium.” That is a narrower but more defensible thesis.
Evidence ledger
Source: ICASA, “Liquid Telecommunications” PDF. URL: https://www.icasa.org.za/uploads/files/Liquid-Telecommunications.pdf. Type: regulator document. Supports: Liquid Telecommunications South Africa (Pty) Ltd acquired licences through an ICASA-approved transfer-of-control process in 2016. Does not prove: current service quality, market share, or physical fibre footprint. Economic relevance: licensed status is the foundation for selling regulated enterprise and wholesale network services.
Source: ICASA, renewal notice for Individual ECNS/ECS licences. URL: https://www.icasa.org.za/uploads/files/Notice_Renewal-of-Individual-ECNS-ECS.pdf. Type: regulator notice. Supports: Liquid Telecommunications South Africa (Pty) Ltd appears in ICASA licence-renewal context. Does not prove: revenue, customer count, or route-level reliability. Economic relevance: regulatory continuity affects ability to serve enterprise, carrier, and public-sector buyers.
Source: Liquid South Africa Acceptable Use Policy. URL: https://za.liquid.tech/wp-content/uploads/sites/4/2025/06/Liquid-Intelligent-Technologies-South-Africa-Acceptable-Use-Policy.pdf. Type: company policy/legal document. Supports: legal/trading identity and service categories including IP Transit, DIA, BIA, voice and broadband. Does not prove: actual uptake or SLA performance. Economic relevance: anchors the correct legal entity and the product universe.
Source: BGP.tools AS36937. URL: https://bgp.tools/as/36937. Type: public routing/RIR-derived network record. Supports: AS36937 is associated with Liquid Telecommunications South Africa (Pty) Ltd, Neotel-AS, AFRINIC, ZA, and Midrand/Johannesburg address data. Does not prove: headquarters, customer numbers, or physical route ownership. Economic relevance: proves public internet routing identity and Neotel-lineage network continuity.
Source: PeeringDB AS36937 and AS30844. URLs: https://www.peeringdb.com/net/4841 and https://www.peeringdb.com/net/725. Type: peering/network database. Supports: Liquid SA’s local ASN posture and wider Liquid group transit/open-peering posture. Does not prove: private peering, uptime, or all physical capacity. Economic relevance: shows how local enterprise reach and group-scale routing interact.
Source: NAPAfrica member list. URL: https://www.napafrica.net/peering/peering-clients/. Type: internet-exchange member directory. Supports: Liquid-related ASNs at South African exchange locations and peering-policy differences. Does not prove: full traffic volume or private interconnects. Economic relevance: exchange presence lowers latency and transit cost, and improves cloud/content adjacency.
Source: Liquid Dedicated Internet Access and Wholesale Products pages. URLs: https://za.liquid.tech/solutions/dedicated-internet-access/ and https://za.liquid.tech/infrastructure-solutions/wholesale-products/. Type: company product pages. Supports: DIA, IP transit, metro access, CloudConnect, protected lines, Ethernet, IP VPN, last-mile and international wholesale offerings. Does not prove: actual contract prices or delivered service levels. Economic relevance: shows the revenue stack Liquid is trying to monetize beyond commodity broadband.
Source: Liquid CloudConnect and Azure ExpressRoute materials. URLs: https://liquid.tech/services/cloudconnect/ and https://www.liquid.tech/about-us/news/liquid-telecom-is-now-a-microsoft-azure-expressroute-partner/. Type: company/partner service materials. Supports: private cloud connectivity, hyperscaler interconnection, BGP/routing design, and Azure ExpressRoute positioning. Does not prove: customer attach rate or profitability. Economic relevance: cloud adjacency is a higher-margin defence against Mbps commoditization.
Source: Liquid “Our Network.” URL: https://liquid.tech/about-us/our-network/. Type: company infrastructure overview. Supports: group-level claims of more than 116,000 km of fibre, subsea connectivity, cross-border routes, data centres, and international hubs. Does not prove: all assets are held by Liquid Telecommunications South Africa (Pty) Ltd. Economic relevance: group reach is part of the enterprise and wholesale sales proposition.
Source: Liquid Telecommunications Holdings Limited Annual Financial Statements 2025. URL: https://liquid.tech/wp-content/uploads/2025/10/Liquid-Telecommunications-Holdings-Ltd-AFS-2025.pdf. Type: audited group financial statements. Supports: group revenue, South Africa segment revenue and EBITDA, debt, network costs, fibre assets, IRUs, and customer concentration disclosure. Does not prove: standalone statutory financials for Liquid Telecommunications South Africa (Pty) Ltd. Economic relevance: shows scale, operating leverage, capital intensity, and financial dependency.
Source: Cassava Technologies debt announcements. URLs: https://www.cassavatechnologies.com/liquid-intelligent-technologies-announces-debt-repayment-and-agrees-new-credit-facilities/ and https://www.cassavatechnologies.com/demand-for-300-million-bond/. Type: company financing announcements. Supports: refinancing, debt repayment, new credit facilities, bond financing, and sponsor capital injection. Does not prove: absence of future credit risk. Economic relevance: debt maturity and cost of capital directly affect infrastructure pricing and capex flexibility.
Source: Xalam/D4DHub South Africa Data Center Market Brief. URL: https://cms.d4dhub.eu/assets/Initiatives/Data-Governance-in-Africa/Digital-Investment-Facility/2507_Country-Market-Briefs/Data-Center-Market-Brief-South-Africa.pdf. Type: market research brief. Supports: South Africa’s data-centre capacity, cloud demand, competitive pricing pressure, power-cost issues, and wholesale fibre conditions. Does not prove: Liquid-specific market share. Economic relevance: defines the demand and pricing environment in which Liquid sells cloud-adjacent fibre.
Source: CSIR H1 2025 power-system analysis. URL: https://www.csir.co.za/media-hub/media-room/loadshedding-down-82-in-h1-2025-as-demand-eases-and-eaf-edges-up-but-tariffs. Type: public research institution analysis. Supports: sharp load-shedding reduction and continuing tariff pressure. Does not prove: local municipal power reliability at Liquid PoPs. Economic relevance: power-risk pricing is changing from crisis premium to measurable resilience premium.
Source: DFA official site and Remgro infrastructure disclosure. URLs: https://dfafrica.co.za/ and https://www.remgro.com/group-investments/infrastructure/. Type: competitor/company disclosures. Supports: DFA open-access footprint, Vumatel/Maziv scale, revenue, operating profit, homes passed and connected. Does not prove: exact enterprise prices against Liquid. Economic relevance: defines the infrastructure competition compressing access margins.
Source: Vodacom FY2026 SENS PDF and Competition Tribunal/Reuters coverage of Maziv. URLs: https://senspdf.jse.co.za/documents/2026/jse/isse/vod/FY26SENS.pdf, https://www.comptrib.co.za/info-library/case-press-releases/vodacom-maziv-merger-prohibited, and https://www.reuters.com/world/africa/safrican-competition-commission-drops-opposition-vodacom-maziv-deal-2025-07-01/. Type: listed-company disclosure, regulator release, press coverage. Supports: Vodacom’s Maziv stake, competition concerns, and market-structure shift. Does not prove: post-transaction wholesale pricing behaviour. Economic relevance: mobile-fibre consolidation changes bargaining power and competitive dynamics.
Source: Department of Justice awarded tenders and Gauteng Provincial Legislature award notice. URLs: https://www.justice.gov.za/cfo_tender/tenders-awarded.html and https://www.gpl.gov.za/wp-content/uploads/2022/03/AWARDED-TENDER-INTERNET-SERVICE-PROVIDER.pdf. Type: public procurement records. Supports: Liquid South Africa as a public-sector supplier. Does not prove: broad public-sector dominance or service quality. Economic relevance: public contracts can create sticky revenue and validate complex procurement capability.
Watchpoints
Track ICASA renewal outcomes and any licence-condition changes for Liquid Telecommunications South Africa’s I-ECNS and I-ECS rights.
Pull CIPC records and beneficial-ownership updates for Liquid Telecommunications South Africa (Pty) Ltd; separate the local contracting entity from Cassava/Liquid group-level claims.
Monitor AS36937 and AS30844 for routing changes, RPKI status, PeeringDB policy changes, NAPAfrica capacity changes, and new IX presence.
Watch Liquid group financial reporting for South Africa segment revenue, EBITDA, capex, network-cost inflation, refinancing cost, and the unnamed customer representing more than 10% of group revenue.
Benchmark Liquid’s DIA and BIA pricing against Openserve, DFA-based providers, MTN, Vodacom Business, and Maziv/Vumatel-based ISPs in the same buildings and metros; generic national price comparisons are inadequate.
Demand route-level proof on large deals: physical path maps, shared duct disclosures, upstream dependencies, restoration times, PoP backup power, generator runtime, and service-credit enforceability.
Track Africa Data Centres’ South African expansion, Samrand power delivery, solar PPA execution, and Johannesburg/Cape Town cross-connect pricing; data-centre adjacency is valuable only if power and interconnect economics hold.
Watch SITA, provincial broadband, education, health, justice, and legislature tenders for Liquid renewals, disputes, awards, cancellations, or pricing disclosures.
Collect RFOs, NOC notices, MyBroadband/forum complaints, and customer-reference signals; treat isolated complaints as weak evidence, but repeated incident patterns by metro or product as commercial risk.
Monitor Vodacom/Maziv integration, remedy compliance, wholesale access terms, and any complaints from rival ISPs or enterprise customers about foreclosure, route access, or pricing.
Track Microsoft, AWS, Google Cloud, Oracle, and local cloud-partner pages for Liquid on-ramp status, PoP count, and ExpressRoute/Direct Connect/Interconnect availability changes.
Watch Eskom and municipal power data separately; national load-shedding improvement does not remove local grid-failure risk at PoPs, street cabinets, enterprise premises, or data centres.