MegaLink makes Bolivian internet trust travel through Miami routes

Summary

• MegaLink SRL is best read as a Bolivian access and business-connectivity operator, not as a generic web-hosting or cloud story. Its public pages market symmetric Internet Online by fiber, Layer 2 data transmission, routed fiber VPN, its own fiber network in La Paz and El Alto, 24/7 personalized technical support, BGP4 and multi-provider international redundancy. Sources: https://www.megalink.com/Servicios and https://www.megalink.com/
• The paid economic unit is a business or institution buying dedicated connectivity. MegaLink's public tariff table lists Internet Online plans in La Paz and El Alto from 10/10 Mbps at Bs 1,019 per month through 200/200 Mbps at Bs 14,849 per month, with zero installation charge for that Internet Online table and a note that peak-hour symmetric speed is guaranteed up to 90% on a best-effort basis. Source: https://www.megalink.com/
• The network evidence is strong enough for a Regional ISP economics and Network-resource evidence article. AS22541 is registered to MEGALINK S.R.L.; public routing observers show active IPv4 announcements, visible peers and IX participation; PeeringDB lists MegaLink as a Cable/DSL/ISP network with open peering policy, 50-100 Gbps traffic level and public exchange presence including PIT Bolivia, MLIXP, JumboIX Peru, 4b42 and UNM-Exch Canada-West. Sources: https://bgp.tools/as/22541, https://bgp.he.net/AS22541, https://www.peeringdb.com/net/11830 and https://stat.ripe.net/data/routing-status/data.json?resource=AS22541
• The Miami route thesis needs a careful reading. Historic routing-policy text in public IRR mirrors lists Telia/Arelion Miami, Cogent Miami and GTT Miami as upstream providers, while current live collectors surface Cogent, IPTP Networks S.A.C., Hivelocity and BreezeHost plus local Bolivian downstream or neighbour ASNs. That means the public record supports a cross-border route-surface analysis, not a claim that every current packet or commercial contract runs through Miami. Sources: https://bgp.he.net/AS22541, https://bgp.tools/as/22541 and https://stat.ripe.net/data/asn-neighbours/data.json?resource=AS22541
• Bolivia's wider market makes this route geography economically important. ATT's first-semester 2025 sector report says mobile internet lines reached 11.18 million, fixed internet connections stood near 1.443 million, fixed internet household penetration was 39.4%, and FTTX represented 95.35% of fixed internet connections. The same report says PIT Bolivia traffic averaged around 70 Gbps and peaked near 90 Gbps, showing that local exchange is growing but does not erase the need for international transit. Source: https://www.att.gob.bo/sites/default/files/archivos_listados_pdf/2026-03-11/Boletin%201er%20sem%202025%2026-01-26.pdf

The account starts in La Paz, but the risk does not stop there

Imagine a finance team in La Paz at the end of a month. Invoices have to be uploaded to a cloud accounting platform, tax records have to be checked, a WhatsApp support channel is open, a bank portal is waiting, video calls are stacked through the afternoon and a branch in El Alto needs the same file repository to behave as if it sits in the next room. The staff do not care whether the next hop is called an autonomous system, an internet exchange point or an upstream provider. They care whether the connection is stable enough for the workday to continue.

That is why MegaLink is more interesting than its size alone would suggest. It sits in a market where the local customer buys an access account in bolivianos, but part of the service quality depends on foreign transit, imported equipment, route policy, dollar liquidity, peering discipline and the willingness of larger networks to carry Bolivian traffic cleanly. The office sees a local sales and support contact. The packets may see La Paz, an exchange point, an upstream carrier, a route server, a cache, Miami, Peru, Tampa or a foreign cloud region. The bill is local. The trust boundary is not.

MegaLink's public offer is explicit enough to classify the company as a live connectivity provider. Its service page markets VPN by routed fiber, Internet Online, and fiber-optic data transmission. The Internet Online product is described as a symmetric dedicated channel over fiber, recommended for companies, organizations or people that need to upload large volumes of information to the cloud. It promises a secure and managed fiber connection, an exclusive private network, 1:1 symmetry, full duplex upload and download, fixed and permanent bandwidth, and BGP4 Tier-1 international redundancy. The same page says support is personalized, by phone and email, 24 hours a day, seven days a week. Source: https://www.megalink.com/Servicios

Those claims should not be read as audited performance data. They are still marketing claims. But they tell the reader what MegaLink is selling: not merely a household best-effort broadband line, but an access and transport account whose value rests on predictable upload, dedicated bandwidth, route diversity and support escalation. That is the right starting point for the economics of the company. MegaLink matters where a Bolivian customer wants a named operator to own the access path, explain faults and manage the route out of the country.

A narrow but real business-connectivity catalogue

The company's public catalogue is concentrated rather than sprawling. The first product is a routed fiber VPN. MegaLink presents it as an encrypted connection from one device or network to another, suitable for secure transfer of sensitive data and remote or branch-office work. The feature list names encrypted tunnels, trunk fiber redundancy, 10 Mbps or 100 Mbps transmission speeds and private sending and receiving of information. The second product, Internet Online, is the symmetric dedicated internet access offer. The third, fiber-optic data transmission, is a Layer 2 point-to-point service for connecting distant sites where physical distance or other conditions prevent simpler links. MegaLink says it has an extended fiber network across La Paz and El Alto, and that ring topology helps guarantee connection security. Source: https://www.megalink.com/Servicios

That product mix shapes the economic unit. The buyer is more likely to be an office, branch network, local institution, professional service firm, school, small operator, regulated business or software-dependent organization than a casual mobile-data user. The public pages do not prove customer mix, account count, revenue or churn. They do show a paid unit centered on connectivity rather than content, SaaS, managed hosting or resale of cloud products. That matters for topic selection. The strongest public categories are Regional ISP economics, Peering and transit, and Network-resource evidence. There is not enough public evidence to tag the company as a cloud service provider merely because the product page says some customers upload data to the cloud.

The contact and identity pages add local grounding. MegaLink lists Av. Sanchez Lima 2520, Edificio Anibal, Planta Baja, Oficina 2, Sopocachi, La Paz, along with sales email and phone contacts. Its "Conocenos" page describes the company as a consolidated internet business whose history has been associated with technological innovation and customer service in Bolivia. The same page says its mission is to connect national organizations through internet and communications networks to help solve their needs. It also says the company is regulated and supervised by ATT, Bolivia's telecom and transport regulator. Sources: https://www.megalink.com/Conocenos and https://www.megalink.com/Contactenos

Again, the important point is not corporate biography for its own sake. The point is operating accountability. A customer buying dedicated internet from a small operator does not only ask whether the brand exists. The customer asks whether there is an office, a technical contact, a regulator, a tariff, a service category and an autonomous system whose public routing can be observed. MegaLink checks those boxes with enough evidence to support a serious market note.

The tariff table says what kind of service this is

MegaLink's tariff table is one of the most useful public documents because it separates the company from the cheap consumer-fiber comparison. The Internet Online table lists postpaid plans in La Paz and El Alto. The 10 Mbps symmetric plan is Bs 1,019 per month; 20 Mbps is Bs 1,979; 30 Mbps is Bs 2,879; 40 Mbps is Bs 3,739; 50 Mbps is Bs 4,529; 60 Mbps is Bs 5,279; 70 Mbps is Bs 6,049; 80 Mbps is Bs 6,819; 90 Mbps is Bs 7,419; 100 Mbps is Bs 7,969; 120 Mbps is Bs 9,239; 140 Mbps is Bs 10,389; 160 Mbps is Bs 11,879; 180 Mbps is Bs 13,359; and 200 Mbps is Bs 14,849. The same table says the listed speeds apply to both download and upload, and a note says that up to 90% of the contracted symmetric speed is guaranteed in peak hour on a best-effort basis. Source: https://www.megalink.com/

That price curve is revealing. On a simple monthly tariff divided by nominal Mbps basis, the 10 Mbps plan is much more expensive per Mbps than the 200 Mbps plan. Scale lowers the unit price, but the service still sits far above household fiber tariffs sold by national operators. That is not automatically a weakness. It tells the customer and the analyst that the account is not intended to compete only on headline residential download speed. It is priced around symmetry, business use, route management and support. A company that only needs entertainment streaming or casual browsing can choose cheaper mass-market options. A company that needs an accountable dedicated path may still compare MegaLink against enterprise offers from Entel, Tigo Business, AXS, local fiber operators or wireless backup.

The tariff language also contains a useful tension. The service description speaks of fixed, guaranteed and permanent bandwidth, while the table note says 90% peak-hour speed under best-effort conditions. That does not make the offer invalid. It shows why business connectivity must be read through both marketing copy and regulatory tariff disclosure. The buyer is not purchasing a metaphysical guarantee that the internet never slows. The buyer is purchasing a managed service whose expected performance, support path and route diversity are stronger than a low-cost consumer account, but whose public table still contains operational limits.

MegaLink also lists a FibraVPN line at 10 Mbps down and 10 Mbps up, with Bs 800 installation, Bs 800 prepaid and Bs 1,200 postpaid, described as a local symmetric fiber channel in La Paz and El Alto without internet access and with 1:1 guarantee. Source: https://www.megalink.com/ The difference between FibraVPN and Internet Online is important. The former is closer to local private transport; the latter is internet access. Both depend on fiber economics, but only the internet product directly exposes the customer to upstream transit and wider route geography.

Why the ASN evidence matters, and where it stops

The strongest technical proof that MegaLink is more than a brochure is AS22541. Public routing records associate AS22541 with MEGALINK S.R.L. BGP.Tools shows it as active and allocated under LACNIC, registered in October 2001, and tied to address resources including 190.14.64.0/18, 200.75.160.0/20 and 2803:7680::/32 in the displayed LACNIC whois text. RIPEstat's routing-status API showed AS22541 with its first seen prefix in 2001, a last seen origin observation on 2026-07-10, 73 IPv4 prefixes and 18,688 IPv4 addresses in announced space, zero IPv6 prefixes in observed routing, and visibility through all 326 IPv4 RIS peers in that snapshot. Sources: https://bgp.tools/as/22541 and https://stat.ripe.net/data/routing-status/data.json?resource=AS22541

Hurricane Electric's BGP view is similar but not identical. It showed 74 originated IPv4 prefixes, zero originated IPv6 prefixes, 15 observed IPv4 peers, no observed IPv6 peers, 18,944 originated IPv4 addresses, all 74 originated routes RPKI-valid in its view, and an average AS path length around 4.7. Source: https://bgp.he.net/AS22541 The difference between 73 and 74 prefixes is not alarming; public collectors use different vantage points and update cycles. It is precisely why routing evidence should be treated as dynamic observation, not a statutory filing.

PeeringDB provides another layer. Its MegaLink SRL network page lists ASN 22541, network type Cable/DSL/ISP, an AS-set of AS22541:AS-ALL, traffic level of 50-100 Gbps, mostly inbound traffic ratio, South America geographic scope, open general peering policy, no multiple-location requirement, no ratio requirement and no contract requirement. It lists public peering points at 4b42 Internet Exchange Point, JumboIX Peru, MLIXP, PIT Bolivia and UNM-Exch Canada-West, with PIT Bolivia shown at 10G and several other ports at 1G. Source: https://www.peeringdb.com/net/11830

This is meaningful evidence. It shows a current routed surface, exchange presence, route-policy declarations and public visibility. It does not prove the number of access customers, realized traffic volume, uptime, route quality, latency from a specific customer site, ownership of every fiber segment, exact upstream contract terms, exact customer list or financial performance. It also does not prove that IPv6 service is available to end users. PeeringDB lists IPv6-related protocol and address information at some exchange points, while HE and RIPEstat showed no globally originated IPv6 prefixes from AS22541 in the snapshots reviewed. The safest conclusion is that the public record supports a live IPv4 network operator thesis, while IPv6 availability and deployment should be treated as an open question.

The Miami claim is a route-policy clue, not a packet-by-packet proof

The planned title matters because Miami has long been a gateway market for Latin American connectivity. But the evidence needs care. Public IRR text shown through Hurricane Electric includes AS22541 routing-policy remarks naming upstream providers and locations. It lists Telia/Arelion in Miami, Cogent Communications in Miami, GTT.net in Miami, IPTP Ltd in the United Kingdom, IPTP Networks S.A.C. in Peru, SoftButterfly S.A.C. in Peru, Grupo ZGH SPA in Chile, Hivelocity.net in Tampa and BreezeHost in the United States. It also lists IXP participation for PIT Bolivia, IXP.MegaLink.com, JumboIX Lima-Peru, 4b42, UNMETERED Exchange, EVIX and SoftButterfly SAC Peru. Source: https://bgp.he.net/AS22541

Current observed neighbours are narrower. RIPEstat's asn-neighbours view showed four left-side neighbours: Cogent AS174, IPTP Networks S.A.C. AS263681, Hivelocity AS29802 and BreezeHost AS400810. It also showed nine right-side neighbours, including WebLink and several Bolivian institutional or MegaLink-linked ASNs. BGP.Tools similarly listed Cogent, BreezeHost, Hivelocity, IPTP Networks S.A.C., WebLink, MegaLink-linked ASNs and Bolivian organizations in the peer and downstream tables. Sources: https://stat.ripe.net/data/asn-neighbours/data.json?resource=AS22541 and https://bgp.tools/as/22541

The difference is the story. A routing-policy record can list intended, historic or maintained import and export policy with providers that may not all be visible in every live collector snapshot. A live collector can miss links or classify them differently, and it captures observation rather than contract text. The article's thesis should therefore be framed as route-surface dependence, not as a claim that MegaLink's live traffic is always or mostly carried by Arelion, GTT or Cogent in Miami. The public record says MegaLink has documented Miami upstream policy in its route registry text and current visible external neighbours that include U.S. and Peruvian networks. That is enough to make Miami and cross-border transit part of the trust analysis.

For a Bolivian business customer, that distinction is more than technical caution. If a route to a bank portal, cloud service, software vendor or foreign email platform leaves Bolivia, service quality can depend on where the traffic exits, how the upstream provider handles congestion, whether return paths are symmetric, whether route filters are current, whether RPKI and IRR data are correct, whether a cache is local, and whether a fault can be escalated through the right commercial contact. The customer may call MegaLink. MegaLink may need to diagnose a local access issue, a domestic exchange issue, an upstream issue or a remote-platform issue. That is why trust travels through the route, even when the invoice is local.

Bolivia's local exchange layer reduces distance but does not remove cross-border dependence

Bolivia has spent more than a decade trying to keep more traffic local. The Internet Society's Bolivia IXP account says Law 164 in 2011 forced Bolivian ISPs to interconnect through an IXP, with regulations in 2012. It says PIT Bolivia was born on 13 November 2013, initially exchanging only 90 Mbps, and that trust among operators was limited because the exchange had been imposed rather than built organically. It later moved from Entel to a more neutral ATT location, gained legal status as a nonprofit entity in 2019, and expanded traffic and membership during and after the pandemic. Source: https://www.internetsociety.org/issues/ixps/success-stories/bolivia/

That history is directly relevant to MegaLink. PeeringDB lists MegaLink at PIT Bolivia with a 10G operational port, and PCH identifies PIT Bolivia as an active association in La Paz established on 13 November 2013. PCH also lists IXP MegaLink as an active commercial exchange in La Paz, managed by Javier Galvez under the Megalink organization and established on 1 November 2013. Sources: https://www.peeringdb.com/net/11830, https://www.pch.net/ixp/details/1789 and https://www.pch.net/ixp/details/1530

MegaLink's own IXP page says IXP MegaLink is the first 100% free IXP for internet providers, educational institutions, government and general business users operating under BGP-4. It says participants need a Bolivian LACNIC-assigned IP range, a fiber loop to connect, dynamic BGP-4 routing, a FastEthernet port and a significant local bandwidth requirement in La Paz. It also says the use of the IXP is free up to 10 Mb of connectivity, while participants may bear the cost of the fiber loop needed to reach the exchange. Source: https://ixp.megalink.com/

The economics are straightforward. Local exchange can reduce transit cost, reduce round-trip distance and make local services feel local. But it is not magic. It depends on who participates, what prefixes are announced, whether content networks and government services connect, how route servers are managed, and whether operators release enough traffic into the exchange. The Internet Society account says PIT Bolivia traffic grew from early low levels, and by the time of its story had reached peaks of 13 Gbps, with claims that local download latency had fallen sharply. ATT's first-semester 2025 report gives a later public benchmark, describing PIT Bolivia traffic in 2025 as sustained and stable, with average values around 70 Gbps and peaks near 90 Gbps. Sources: https://www.internetsociety.org/issues/ixps/success-stories/bolivia/ and https://www.att.gob.bo/sites/default/files/archivos_listados_pdf/2026-03-11/Boletin%201er%20sem%202025%2026-01-26.pdf

That means MegaLink's customer proposition is partly national and partly international. A route to another Bolivian network may improve when local exchange works. A route to foreign SaaS, overseas hosting, a foreign bank correspondent, a cloud region or international content still depends on transit and foreign interconnection. The better the local exchange layer becomes, the clearer the remaining cross-border problem becomes: what cannot be localized must still be carried reliably out of a landlocked market with limited foreign-exchange flexibility and heavy dependence on international carriers.

A crowded substitute set keeps MegaLink honest

MegaLink does not sell in a vacuum. Bolivia's customer can compare several kinds of substitute, even when those substitutes are imperfect. The cheapest visible substitute is mass-market fixed fiber. Entel's household fiber page says new plans entered force on 4 March 2026, including 120 Mbps at Bs 169, 180 Mbps at Bs 229, 600 Mbps at Bs 369 and 1,000 Mbps at Bs 450 per month. Source: https://www.entel.bo/HogarInternetFibra Tigo's consumer internet page lists household and combo plans such as 100 Mbps at Bs 249, 150 Mbps at Bs 289, 300 Mbps with TV and mobile at Bs 399, and 500 Mbps with TV and mobile at Bs 439. Source: https://www.tigo.com.bo/internet Viva's fiber page lists fiber plus mobile offers such as 150 Mbps with a WOW plan at Bs 229, 200 Mbps at Bs 279 and 200 Mbps at Bs 339, subject to coverage checks. Source: https://www.viva.com.bo/internet-fibra-optica/

Those offers are much cheaper per Mbps than MegaLink's Internet Online table. But they are not the same product. They are primarily household or bundled consumer accounts, often with asymmetric or mass-market performance assumptions, coverage checks, equipment conditions and a different support model. A business can still buy them as backup or for noncritical use, but a finance office, branch network, security-conscious institution or software firm may need stronger upload symmetry, public addressing, private transport, route escalation or a supplier that expects business-critical calls.

The more direct substitutes are enterprise and business products. Entel's business page lists Internet Fibra Empresa plans and Internet Corporativo plans, with corporate tiers from 10 Mbps at Bs 610 through 600 Mbps at Bs 26,690, plus installation and equipment-related charges. It also lists On Line products, with 10 Mbps at Bs 3,000, 20 Mbps at Bs 4,500, 40 Mbps at Bs 7,000, 70 Mbps at Bs 9,000, 100 Mbps at Bs 14,000, 150 Mbps at Bs 17,300 and 200 Mbps at Bs 23,000. Source: https://www.entel.bo/EmpresaInternet Tigo Business lists asymmetric fiber access for companies with national coverage, six redundant rings, 90% guaranteed bandwidth, 2:1 upload-download relation and 24/7 monitoring, with plans from 30 Mbps at Bs 4,550 through 75 Mbps at Bs 6,650. Source: https://www.tigo.com.bo/medianas/conectividad/internet-asimetrico

Viva creates another kind of substitute: fixed wireless and mobile. Its VIVA WIFI postpaid page describes LTE TDD fixed wireless service available across major Bolivian cities, with 4 Mbps, 6 Mbps, 16 Mbps and 22 Mbps plans and quotas from 200 GB through 800 GB, subject to technical feasibility and equipment rules. Source: https://www.viva.com.bo/personas/hogar/viva-wifi-postpago/ That is not a dedicated fiber replacement for a bandwidth-heavy office, but it is a credible backup or low-end substitute where speed and latency tolerance are lower.

The result is a price-pressure triangle. Low-cost fiber bundles set customer expectations for megabits. Enterprise offers from national operators set a benchmark for managed business connectivity. Mobile and fixed wireless provide backup and reach where fiber is absent or too slow to install. MegaLink's defendable niche is narrower: local business access, symmetric fiber, point-to-point transport, BGP-fluent routing and support in La Paz and El Alto. That niche can be profitable if customers value accountability more than headline Mbps. It becomes vulnerable if national operators can match the same accountability at lower price, or if customers decide that cheap fiber plus mobile backup is good enough.

The macro risk sits inside the router budget

Connectivity may feel digital, but the cost base is physical and foreign-currency exposed. A provider such as MegaLink must finance fiber loops, switches, routers, optical gear, power protection, customer-premises equipment, replacement parts, skilled network labour, office support and upstream transit. Its revenue is likely collected mostly in bolivianos. Many inputs are imported or priced directly or indirectly in dollars. When a country experiences foreign-exchange shortage, import scarcity or a widening gap between official and parallel exchange rates, the local telecom bill can become harder to price even if the service is technically stable.

The IMF's 2025 Article IV press release on Bolivia warned that depleted reserves and an untenable U.S. dollar peg called for a decisive shift in the monetary framework. It said directors called for exchange-rate realignment, greater flexibility, fiscal consolidation and restrictive monetary policy to address inflationary pressure, alleviate foreign-exchange shortages and eliminate restrictions. Source: https://www.imf.org/en/news/articles/2025/05/30/pr-25168-bolivia-imf-concludes-2025-art-iv-consult The World Bank's Bolivia country page similarly says the post-commodity-boom growth model led to rising debt and reduced international reserves and fiscal savings. Source: https://www.worldbank.org/ext/en/country/bolivia

Those macro facts do not tell us MegaLink's margin, debt, supplier contracts or cash position. They do tell us why a Bolivian ISP's economics cannot be assessed only from its Mbps table. If replacement routers, optics or software support have to be paid in scarce dollars, a stable boliviano-denominated tariff may compress margins. If upstream transit is priced in dollars, route diversity can become more expensive just when customers need reliability most. If fuel, power backup or import logistics deteriorate, truck rolls and repairs become less predictable. If inflation pressures wages for scarce network engineers, service labour gets more costly. The customer sees a monthly internet bill. The operator sees a local-currency revenue stream funding a global supply chain.

This is where MegaLink's multi-provider route posture matters. A business customer may not know whether a route leaves through Cogent, IPTP, Hivelocity, BreezeHost, a historical Miami policy record or a local exchange. But route diversity has an economic cost. One upstream can be cheaper than several. Several can be more resilient than one. A local exchange can reduce transit load but requires participation, equipment and policy discipline. A small operator must balance those costs against what the market will pay. The higher the macro stress, the more that balance becomes a strategic question rather than a network-engineering footnote.

Regulation gives structure, but not a full quality view

ATT records give MegaLink formal context. The public Registro Unico de Licencias page lists MEGALINK S.R.L. with a current departmental data-transmission service in La Paz. Source: https://plataformas.att.gob.bo/index.php/Rul/publicos An older ATT operators PDF also lists MEGALINK S.R.L. for data transmission in the La Paz local service area. Source: https://www.att.gob.bo/sites/default/files/archivos_listados_pdf/2021-07-26/OPERADORES%20DE%20SERVICIOS%20PUBLICOS%20DE%20TELECOMUNICACIONES.pdf MegaLink's own site repeatedly says the company is regulated and supervised by ATT. Sources: https://www.megalink.com/ and https://www.megalink.com/Conocenos

The tariff disclosure on MegaLink's site says it is published in compliance with Article 120 paragraph VIII of Bolivia's General Telecommunications Regulation, D.S. 1391. Source: https://www.megalink.com/ That matters because the customer-facing tariff is not merely a private quote page. It is part of a regulated service environment. The same ATT sector report shows why that environment is changing. It says projected 2025 telecom net income for the sector was Bs 10.759 billion, after earlier declines and with operators facing saturated markets, OTT substitution for traditional voice and messages, tighter margins and rising investment needs. It also says projected 2025 mobile internet revenue was Bs 4.485 billion and fixed internet revenue was Bs 2.523 billion, together representing 75% of final-service revenue to users. Source: https://www.att.gob.bo/sites/default/files/archivos_listados_pdf/2026-03-11/Boletin%201er%20sem%202025%2026-01-26.pdf

This regulatory context cuts both ways. On one side, it gives customers a tariff and a supervisory authority. On the other, it shows that fixed and mobile internet are the main economic pillar of a sector under revenue pressure. If national mobile and fixed operators push more aggressively into business connectivity, a regional operator must defend its accounts through route expertise, local support, specific building reach or specialized data transport. Regulation can set the field; it cannot by itself create trust in repair execution.

The ATT report's fixed-internet statistics also help scale the opportunity. At 30 June 2025 Bolivia had about 1.443 million fixed internet connections, down slightly from 2024. Fixed internet household penetration was 39.4%, below half of households. FTTX accounted for 95.35% of fixed internet connections. La Paz's fixed internet penetration was 40.36%, below Cochabamba, Tarija and Santa Cruz but close to the national middle. Source: https://www.att.gob.bo/sites/default/files/archivos_listados_pdf/2026-03-11/Boletin%201er%20sem%202025%2026-01-26.pdf The message for MegaLink is that fiber is no longer exotic in Bolivia; it is the dominant fixed access technology. The differentiated question is not whether fiber exists, but who can provide accountable business-grade service over it.

Repair accountability is the product customers remember

The operational test for MegaLink is not the moment the customer signs a tariff. It is the first serious fault. A symmetrical connection can look impressive in a table, but a business customer evaluates it when uploads stall, a remote branch cannot reach the central server, a video call fails, an application times out or a foreign platform becomes reachable through one path and unreachable through another. The difference between a commodity internet line and a trusted business account is the ability to tell the customer where the problem is and what is being done about it.

MegaLink's own language points in that direction. The Internet Online service page emphasizes "atencion personalizada" and support by phone and email 24/7. The data transmission product repeats the same promise. The contact page gives a physical address, fixed line, WhatsApp and sales email. Sources: https://www.megalink.com/Servicios and https://www.megalink.com/Contactenos Those details do not prove actual ticket closure speed. They do show that the public offer is built around named support channels, not only around an anonymous Mbps label.

For the customer, repair accountability has several layers. The first is local access. A fault can come from the customer's router, an optical connector, power at the premises, a fiber break, a switch port, a configuration error, an overloaded local segment or a problem in the building path. The public sources do not let us inspect MegaLink's field procedures, inventory or escalation rules. But the fact that MegaLink markets its own fiber network in La Paz and El Alto, and speaks of an extended ring topology for data transmission, means the local physical layer is a core part of the purchase rather than a hidden wholesale detail. Source: https://www.megalink.com/Servicios

The second layer is routing. A business customer may report that "the internet is down" when only a route to one destination is failing. If the target is local, the relevant question may be PIT Bolivia, MLIXP or a domestic peer. If the target is foreign, the relevant question may be Cogent, IPTP, Hivelocity, BreezeHost, a historic policy entry, a remote platform, DNS or return-path asymmetry. Public route evidence shows that MegaLink has enough BGP and IX surface for those distinctions to matter. Sources: https://www.peeringdb.com/net/11830 and https://stat.ripe.net/data/asn-neighbours/data.json?resource=AS22541

The third layer is proof. A customer can rarely prove whether a problem belongs to the access provider, upstream carrier or remote service without help. That is one reason local operators with BGP competence can hold value even when national bundles are cheaper. The support conversation becomes part of the product. A good answer narrows the fault domain, explains whether local routes are healthy, identifies whether only one destination is affected, gives a workaround if a backup path exists and tells the customer whether escalation has left the local network. A poor answer turns the customer back into a commodity buyer looking only at price.

The fourth layer is commercial substitution. If the failure is recurrent, the customer can test substitutes. Entel and Tigo sell business products; Viva and mobile providers offer backup paths; consumer fiber can carry noncritical traffic; foreign cloud platforms can be accessed through multiple local providers. MegaLink's retention problem is therefore not merely technical. It has to persuade customers that the support and route-control layer is worth the premium over a cheap fiber line plus mobile failover. The more a customer depends on upload, branch connectivity, fixed public addressing, private transport or clear fault ownership, the stronger MegaLink's case becomes. The more the account is simple browsing and entertainment, the weaker the case becomes.

There is also a regulatory and reputational layer. A public tariff makes expectations more concrete. An ATT-supervised service gives the customer a reference point. But a regulator cannot answer every fault call, and a tariff cannot diagnose an upstream route. In practice, customer trust is won or lost in the gap between the formal service description and the lived repair process. That is why the article's economic lens is route geography and accountability rather than scale alone. A small operator can be strategically important if it owns a hard-to-substitute repair path for a defined customer base. It can also become fragile if customers conclude that the support premium does not buy faster restoration.

What customers are really buying from MegaLink

The most useful way to think about a MegaLink account is as a bundle of five promises. The first is local access: a fiber path in La Paz or El Alto, with the physical feasibility and support constraints that follow. The second is symmetry: upload and download are both part of the product, which matters for cloud backup, file transfer, video, remote work and branch synchronization. The third is route control: BGP, multi-provider redundancy and exchange presence are part of the value proposition, even when the customer never sees a route table. The fourth is repair accountability: someone must answer when the line, route or customer premise fails. The fifth is regulatory legibility: the tariff and operator status are visible enough to make the service a regulated account rather than an informal connectivity arrangement.

Each promise has a corresponding weakness. Local access is valuable only where the fiber reaches and where installation can be done. Symmetry costs more than mass-market asymmetry and is harder to defend if customers compare only headline download Mbps. Route control is only as good as current upstream diversity, route hygiene and fault escalation. Repair accountability costs money and can be damaged by labour shortages, equipment scarcity or weak communication. Regulatory legibility does not prove service quality. The customer's trust therefore rests on execution, not on any single public fact.

That is why route geography matters. If MegaLink's public routing policy points to Miami carriers and live collectors point to U.S. and Peruvian neighbours, the operational accountability chain crosses borders. A fault might be local fiber, a route filter, a congested upstream, a remote SaaS issue, a data-center path, a DNS path, a peering policy or a return-route asymmetry. A business account buys MegaLink's ability to sort those layers quickly enough that the customer does not have to become a network engineer.

The public downstream and neighbour record hints at the kinds of stakes involved, though it should not be overstated. BGP.Tools and RIPEstat show right-side or downstream ASNs with Bolivian institutional names as well as MegaLink and WebLink-linked ASNs. That does not prove current customer contracts or exclusive dependence. It does show that AS22541 is visible in a routing position near organization-scale networks rather than only in an anonymous residential pool. Sources: https://bgp.tools/as/22541 and https://stat.ripe.net/data/asn-neighbours/data.json?resource=AS22541

What would change the judgement

Several facts would materially improve the confidence level. The first is current upstream contract evidence or operator-confirmed route maps. Public IRR text and live route collectors are useful, but they do not tell us which providers are active commercial suppliers, which are backup-only, which are stale policy entries and which carry the most traffic. A current network map with upstream capacity and failover rules would make the Miami dependency thesis stronger or narrower.

The second is measured performance. Public BGP tables show reachability, not customer experience. Latency from a La Paz customer to Bolivian banks, tax services, cloud regions, Lima, Santiago, Miami and U.S. East Coast destinations would show where MegaLink's route choices matter most. Outage logs, mean time to repair, packet-loss measurements and customer trouble-ticket resolution data would be even stronger. None of those are public in the sources reviewed.

The third is financial and customer-segment data. MegaLink's tariff table lets us infer pricing logic, but it does not show revenue mix. We do not know how many customers are on Internet Online, how many buy private transport, how many rely on VPN, how many are business accounts, how many are residential or small office users, or how much wholesale or downstream traffic contributes to revenue. We also do not know capex, debt, upstream cost, dollar exposure, labour cost or margin.

The fourth is IPv6 and security posture. Public data shows IPv4 strength and RPKI-valid IPv4 routes in HE's view, while IPv6 global announcements were not visible in the snapshots reviewed. If MegaLink has a concrete IPv6 deployment plan, customer IPv6 service or route-security documentation beyond public RPKI observations, that would change the assessment of future readiness. If not, IPv6 lag becomes a watchpoint for a provider selling business reliability.

The fifth is local content and cache evidence. PIT Bolivia's growth matters because local exchange can reduce dependency on long-haul transit. But the actual customer experience depends on whether the services customers use are local, cached or still outside Bolivia. Evidence of local caches, CDN participation, direct government and banking exchange, or improved routes to common cloud platforms would sharpen the judgement.

The conclusion: small operator, large trust surface

MegaLink is not a national telecom incumbent and should not be valued as one. It is also not a thin directory listing. The public evidence shows a real Bolivian regional ISP and business-connectivity operator with a current service catalogue, regulated tariff disclosure, La Paz identity, AS22541, visible IPv4 routing, exchange participation, open peering posture and a route-policy history that reaches into Miami and other international transit surfaces.

That combination is the point. The company matters because the Bolivian customer is buying a service whose economic centre is local and whose trust surface is international. A dedicated 50 Mbps or 100 Mbps symmetric account is not only a speed tier. It is a claim that MegaLink can connect the office, keep upload and download useful, manage BGP, use local exchanges where possible, carry traffic out of Bolivia where necessary, and answer when the chain breaks.

The risk is that the same chain contains many actors MegaLink does not fully control: foreign carriers, local exchange participation, imported hardware markets, foreign-exchange liquidity, national competitors, customer-premises equipment, cloud platforms and regulatory pressure. The opportunity is that customers who understand those risks may prefer a smaller operator that can speak the language of routes, local fiber and support over a cheaper plan that treats all megabits as equal.

That is why "MegaLink makes Bolivian internet trust travel through Miami routes" is not a claim about one immutable path. It is a way to read the company. MegaLink sells local business connectivity in a country where the internet is increasingly essential, fixed fiber is widespread but not universal, mobile is the mass-market default, and cross-border transit remains part of the customer experience. The operator's value is the ability to make that distant route feel accountable to a Bolivian buyer.