When the Internet Expires at Midnight: SIGLO REDES DE MEXICO and the Operating System of Prepaid Access

Summary

• SIGLO REDES DE MEXICO SAPI DE C.V turns a fixed rooftop connection into a prepaid service: installation establishes the physical path, while each recharge opens a seven-, fifteen- or thirty-day access window. Its product is therefore the coordination of coverage, labour, equipment, payment and support, not merely a radio link.
• Public records prove a real Mexican service provider, a 2021 telecommunications-commercialisation authorisation, a 2025 single commercial concession and an active autonomous system. They do not prove nationwide built coverage, exclusive spectrum, ownership of every network component, customer totals, capacity or redundant upstream connectivity.
• The low headline price is economically meaningful but incomplete. Short-duration packages cost more per day, cash channels can add fees, the household supplies roof access and electricity, and the operator carries site, field-service, radio, transit and support costs whether or not a customer renews on day thirty-one.
• A serious buying test must be performed at the actual address and during the busy evening period. It should verify line of sight, installed-equipment ownership, wired speed, latency, packet loss, recovery time, support escalation, privacy practice, upstream resilience and the cost and mechanics of leaving.

The moment a fixed connection becomes temporary

At first glance, prepaid home internet sounds like a contradiction. A mobile user expects to buy a finite allowance and walk away. A home connection normally implies a continuing bill attached to a cable, a modem and an address. SIGLO REDES DE MEXICO SAPI DE C.V combines the physical permanence of the second model with the financial clock of the first. A technician puts an antenna on or near the roof, runs a cable into the home and connects a Wi-Fi router. That apparatus can remain in place after the paid period ends. What disappears is permission to use the service.

The company's current public offer makes time the unit of sale. Its home page advertises unlimited data, no compulsory term and average upload and download speeds of up to 100 megabits per second. The displayed price is MXN339 for thirty days, with shorter options available. Its June 2025 price notice lists MXN110 for seven days, MXN210 for fifteen days and MXN339 for thirty days, alongside six-month and annual options. The registered consumer agreement is more precise about the trigger: in prepaid service, the recharge date begins the service period. That means the commercial boundary is not an invoice month or a technician visit. It is the moment value reaches the account.

This is the opening mechanism for understanding Siglo. The recharge is a small transaction that activates a much larger chain. The company must already have judged the address serviceable, secured an elevated path between the home and a radio site, installed and aligned customer equipment, allocated capacity on a sector, connected that sector to aggregation and upstream networks, associated the household with an account, and made support reachable. It must then recognise cash, card or bank payment, change the account state promptly and keep the physical link usable for the purchased period.

If any element fails, an apparently simple package of days becomes a sequence of hand-offs.

The model fits a place in which income and expenses do not always arrive on a monthly direct-debit rhythm. It also fits a dense urban form where extending fibre to every address can be slow or uneconomic. Nezahualcóyotl, east of Mexico City, is central to Siglo's early story. An independent Thomson Reuters Foundation account of Neza described a city built incrementally, where residents added floors and services over time. That physical history matters to fixed wireless. A roof can provide a path over a street that is hard to trench, but irregular heights can also block another household's path. Local density is both the opportunity and the interference problem.

Prepayment transfers one kind of risk away from the operator: it need not extend a month of credit and chase every unpaid bill. It does not remove the network's fixed costs. The radio site, transit circuit, support shift and field team still exist during a customer's unpaid interval. Nor does prepayment remove household dependence. A child may need connectivity for school or a shop may need it to issue a quotation even when the next recharge has been delayed. Siglo's design therefore rests on an exact balance: make stopping financially possible without making restarting operationally difficult.

That is why this is not a story about a cheap tariff alone. It is about whether an access operator can make every transition—interest to survey, survey to installation, cash to service, fault to repair, expiry to renewal and address change to realignment—behave like one coherent system.

The legal name on the roof

The assigned company is not an invented label derived from a network register. It is the legal provider named across customer and regulatory records. Siglo's site footer identifies Siglo Redes de México S.A.P.I. de C.V. and calls it a subsidiary of Siglo Network, Inc. The Y Combinator company page describes “Siglo” as an active Summer 2021 company founded in 2020 to provide wireless internet in urban Latin America. That establishes the startup and brand context. It does not replace the Mexican contracting party.

The stronger bridge is the PROFECO-registered adhesion agreement, registered in January 2021. It names Siglo Redes de México, S.A.P.I. de C.V., the Siglo brand and the same internet service, and supplies its Mexican tax identity. It also cites an IFT authorisation valid to November 2030 for a telecommunications services commercialiser. The agreement says the provider is responsible to the subscriber, that the service is fixed internet at the registered address, that a prepaid recharge starts the active period and that the contract has no compulsory minimum term.

The regulatory position then changed. In September 2024 the company applied to provide internet using licence-exempt bands in named municipalities in the State of Mexico and in Iztapalapa, Mexico City. The IFT's February 2025 resolution granted the exact legal entity a single commercial concession for thirty years, with national geographic scope. That is substantial evidence of authorisation. It also contains two essential limits on interpretation.

First, national scope is permission, not a map of constructed service. The application described La Paz, Nezahualcóyotl, Chimalhuacán, Chicoloapan, Texcoco and Iztapalapa, while Siglo's own ordering flow still asks a household to verify coverage. No public build schedule in the evidence reviewed here converts the concession's national reach into nationwide physical presence. Second, the IFT stated that use of licence-exempt spectrum cannot itself be treated as ownership of network infrastructure; eligible parties may use those bands subject to technical rules. A concession does not make a shared radio band exclusive.

The legal sequence is therefore coherent without being expansive: a Mexican provider operated under a commercialiser authorisation reflected in its 2021 registered agreement, then obtained a single commercial concession in 2025. Its public brand has a US parent context, but Mexican households contract with the named Mexican entity. Its authority extends nationally, but its delivered footprint must still be proved at each address. Keeping those distinctions intact prevents a common research error—turning permission, branding or an ASN name into an unsupported claim of asset ownership or scale.

There is also a disclosure-maintenance question. Siglo's website terms, dated March 2023, describe accounts, installation, payment through Openpay, OxxoPay and SigloPay, a seven-day trial and a promissory note connected to antenna equipment. Some language appears to carry an unrelated website name. The company's commercial-practices code is much richer on the actual fixed-wireless service, yet contains unfinished wording in a few operational clauses. These documents are evidence that procedures exist, not assurance that every public document is the latest operative version. A customer or procurement team should obtain the current registered contract and order sheet, then make those documents control over an older web PDF.

From postcode to line of sight

Siglo begins with an address, not a nationwide coverage claim. A prospective household provides location and contact information; the operator must decide whether service is technically feasible. The public FAQ says installation can occur within a maximum of forty-eight hours and lists three basic conditions: the home must be inside coverage, an adult must be present and the installer must have roof access for the antenna. Each condition reveals part of the operating architecture.

Coverage is not a simple coloured polygon. In a fixed-wireless access network, an address can sit within the nominal reach of a site and still fail because a building, tree, rooftop extension or terrain feature blocks the useful path. The commercial-practices code explicitly lists line-of-sight obstacles, terrain, weather and movement or manipulation of equipment among conditions that can impair service. A site survey therefore needs more than a postcode match. It should identify the serving sector, the mounting point, the cable run, the expected signal margin and any likely obstruction growth.

The answer may differ between adjacent houses and even between floors in the same building.

Once accepted, the household-side chain is straightforward to describe. The FAQ identifies a dual-band TP-Link router using 2.4 and 5 GHz Wi-Fi and says a typical arrangement can serve about five wireless and four wired devices. A cable links the indoor device to the rooftop antenna. The antenna and router require electricity, so a local power failure removes the connection even if the distant site remains online. Moving only the router is not necessarily harmless: the FAQ says support must validate that the replacement cable still reaches the antenna. Moving or misaligning the antenna requires technical intervention.

The indoor router is the customer's experience of the product, but it is not a reliable measure of the radio service by itself. A slow phone in a back room may indicate Wi-Fi interference or weak indoor coverage rather than congestion on Siglo's sector. Conversely, a strong Wi-Fi icon says nothing about the rooftop link or upstream path. Siglo advises testing with a wired connection. That is the correct starting point for measurement because it separates the home wireless environment from the access link, though it does not by itself isolate the serving sector from backhaul and transit.

Two supplier accounts illuminate parts of the network without establishing a universal design. An early Siklu success story says Siglo initially used short-range 5 GHz access links in Neza and Siklu millimetre-wave EtherHaul connections at 2 Gbps full duplex for backhaul. It says the company wanted eventually to use 60 GHz access technology. The document also reports more than 1,200 connected homes six months after a November 2020 launch and a 50/25 Mbps offer at MXN400. These are historical supplier and company statements, useful for architecture and chronology but not current fleet, customer or performance measurements.

A later Cyber Antennas case account hosted by Cambium Networks describes a December 2025 field comparison on a production Siglo sector using Cambium ePMP 3000 equipment at 5.3 GHz in a 20 MHz channel. The supplier reported forty-nine connected stations in one comparison and more than sixty-seven in another, with better signal and lower retransmission after changing horn antennas. Those measurements support a narrow conclusion: at least one observed Siglo deployment used point-to-multipoint 5 GHz access whose performance was sensitive to antenna pattern, signal-to-noise ratio and retransmission. They do not establish that every sector uses that equipment, that the reported gains persisted, or that supplier-selected measurements equal customer experience.

The 2025 concession application adds another clue by naming licence-exempt 5, 60, 70 and 80 GHz bands. It cannot prove that each is deployed. Taken together, the sources support a layered possibility rather than a single owned network: household Wi-Fi; a customer-premises radio; a short fixed-wireless access path; site-level aggregation; wireless or fibre backhaul depending on location; IP routing; and purchased or exchanged connectivity to the wider internet. Equipment, sites and circuits may be owned, leased or supplied under differing terms. Public evidence does not allocate title across every layer.

This layered view matters because each layer fails differently. Indoor Wi-Fi is affected by walls and neighbours. A rooftop link is affected by alignment and obstruction. A shared 5 GHz sector is affected by interference and load. Millimetre-wave backhaul needs careful path engineering and weather margin. A fibre hand-off can be cut. A route can be withdrawn. A payment can be late. “The internet is down” is the household's accurate description, but the operator's job is to locate the break quickly enough that the prepaid promise remains credible.

The installer is part of the product

Software cannot open a locked roof, choose a safe mount or explain to a family why the antenna must face a particular direction. Siglo's economics depend on local labour because installation converts theoretical radio reach into a paying connection. The field visit is also a moment of trust: a worker enters or climbs a customer's property, handles network equipment, observes the home environment and may help the customer activate or pay.

Mexico's NOM-184-SCFI-2018 sets consumer-facing requirements for telecommunications service, including registered contracts, fulfilment of advertising promises and identifiable installation personnel with a work order. Siglo's practices code describes sales, distribution, installation, repair and fault attention as connected functions. That is operationally sensible. If sales accepts an infeasible address, the installer absorbs the conflict. If installation records do not capture the serving site, signal, cable route and equipment identifiers, support begins every fault without context. If payment status is separate from the work order, the customer may have a completed physical installation but no usable service.

Siglo's public support window runs Monday through Saturday from 9 a.m. to 9 p.m. Central Mexico time, with WhatsApp as a prominent channel. That is long enough to cover much household use, but not a published round-the-clock restoration commitment. The FAQ quotes MXN149 for moving service to another address, subject to coverage. A move is not an account-field change alone. It may require a new survey, a different sector, another mounting decision and equipment realignment. The modest stated fee suggests either a deliberately subsidised move, a limited scope, or costs recovered elsewhere; public material does not show which.

Founder Isaac Phillips has publicly described an installer system that goes beyond ad hoc dispatch. In a LinkedIn account of Siglo's field programme, he referred to monthly breakfast and training sessions for independent installers, more than 1,000 household visits per month across seven municipalities, and performance levels tied to successful work, fewer tickets and priority zones. Those are founder claims rather than audited workforce statistics. They nevertheless expose the correct management variables: first-time completion, repeat visits, ticket creation, travel by zone, safety, workmanship and installer retention.

Independent labour can expand coverage faster than a fully employed field force, particularly across fragmented neighbourhoods. It can also widen variance. The household should not have to discover whether a mount, weather seal, cable bend or router setting depends on which installer arrived. The operator needs standard survey evidence, photographed work, equipment tracking, escalation and quality sampling. Incentives should reward durable connections, not just completed jobs. A low installation cost is illusory if a poor link generates three visits and a cancelled customer.

Support is similarly local. A WhatsApp conversation can distinguish expired service from a technical fault, produce a cash payment reference and arrange access to the roof. That combination can outperform a distant call centre for the target workflow. It can also become a bottleneck if one channel handles billing, faults and installation without clear queues. The relevant support measure is not whether WhatsApp exists; it is the distribution of time to first useful response, remote resolution, field dispatch and full restoration, split by cause and by paid-package length.

A tariff measured in days, a network financed in years

Siglo lowered its thirty-day price from MXN399 to MXN339 in June 2025. The same notice lists seven days at MXN110, fifteen at MXN210, six months at MXN1,999 and a year at MXN3,999. The headline is affordable flexibility. The arithmetic shows how the operator prices that flexibility.

Ignoring payment-channel fees, seven days cost about MXN15.71 per day, fifteen days MXN14, and thirty days MXN11.30. Six months, if treated as 180 days, cost roughly MXN11.11 a day; the annual package is about MXN10.96. A household that bought fifty-two seven-day periods would spend MXN5,720 over 364 days, around 43 per cent more than the annual package. Twelve thirty-day periods cost MXN4,068, only MXN69 more than the annual offer but cover 360 rather than 365 days. The sharpest discount is therefore between intermittent weekly access and a long commitment, not between monthly and annual payment.

Cash adds another layer. Siglo's notice says card payments, bank transfer and certain Walmart-group counters have no Siglo-listed fee; OXXO adds MXN15, while SigloPay collection through an installer or seller adds MXN20. A seven-day purchase at OXXO therefore costs MXN125, or about MXN17.86 per day. The collection fee is small in absolute terms but large relative to the shortest package. The household buying flexibility most often may pay the highest effective rate.

These differences are not necessarily unfair. Short packages create more payment transactions, more expiry and restart contacts, less predictable demand and greater churn risk. Cash collection has real distribution and reconciliation costs. The issue is whether the customer sees the all-in amount before choosing a channel and whether activation is immediate enough that paid days are not consumed by reconciliation delay. A useful receipt should show amount, fee, activation timestamp, expiry timestamp, account and a support route for a missing credit.

The economic bargain is more complex for Siglo. Radio equipment, site access, backhaul, addresses, upstream connectivity, software, customer-premises stock, installation and support require money before a household has renewed many times. Yet no compulsory term means revenue can stop at each expiry. Prepayment removes bad-debt exposure but increases the importance of renewal frequency and reactivation experience.

The operator needs to know the fully loaded acquisition cost, installation recovery period, contribution after transit and support, sector utilisation, failed-recharge rate, thirty-, ninety- and 365-day survival, and the share of dormant equipment that can be recovered or reactivated.

The loyalty design tries to influence that rhythm. The price notice awards points across packages, rising with duration, and the company's public materials promote referrals and local seller participation. Such mechanisms can reduce acquisition cost and encourage longer payment, but they are not free growth. Points create a future liability or discount; referrals can concentrate households on a sector; seller commissions add cost; and local collection expands the number of people handling customer information and money. The gain must be measured against churn, fraud, support and capacity, not just registrations.

There is also a useful distinction between affordability and liquidity. MXN3,999 a year is cheaper per day than weekly purchases, but a household may be unable or unwilling to advance that sum. The seven-day option is valuable because it lets connectivity follow available cash, even though its annualised cost is higher. Judging Siglo solely by the lowest equivalent monthly rate would miss the product. The service is selling control over timing.

Demand, however, is not perfectly elastic. When many households recharge for weekends, school periods or paydays, the operator cannot instantly add spectrum or site capacity. A sector that looks efficient during dormant days may congest when customers become active together. The prepaid ledger must therefore feed radio planning: active accounts, paid days and traffic concentration need to be forecast at sector level. Selling unlimited gigabytes removes an allowance boundary, but it does not remove shared capacity.

What AS273254 proves—and what it does not

An autonomous system number is stronger evidence than a marketing adjective and weaker evidence than an end-to-end network inventory. bgp.tools' live view of AS273254 names Siglo Redes de México S.A.P.I. de C.V., links the LACNIC registration and shows an active network originating three IPv4 /24 routes and two IPv6 routes. The observed IPv4 space totals 768 addresses. Route-origin authorisations are shown as valid, and one visible upstream is AS32098, Flō Networks, formerly Transtelco.

IPinfo's independent AS273254 view broadly corroborates the three IPv4 /24s, IPv6 announcements and valid route-origin state. It adds active-measurement and geolocation indications but does not turn addresses into households. An IP address may serve a router, shared address translation, management function or other purpose. “Hosted domains” is a poor proxy for a residential access network, and a scan result is not a subscriber census.

APNIC offers a different kind of estimate. Its Mexico network-population measurement for April 28, 2026 associated AS273254 with an estimated 17,945 users and roughly 0.02 per cent of the country's measured internet population. That is a model-derived eyeball estimate, not a disclosed customer count. One subscription can serve several people, sampling varies, shared addressing complicates inference, and an access provider can carry traffic through more than one arrangement. It is useful as a sign that the ASN is seen serving end users, not as a revenue or market-share figure.

The upstream observation is similarly bounded. Flō's own network page offers IP transit and connectivity across Mexico and other Latin American markets. PeeringDB's AS32098 record, maintained from operator-supplied information, shows a broad exchange presence and an open peering policy. This gives context to the route collector's visible path: Siglo can reach the wider internet through a substantial regional carrier. It does not disclose Siglo's contract, committed rate, hand-off location, price, protection path or restoration terms.

One visible upstream is not proof of single-homing. Route collectors do not see private links, paths used only after failure, or every commercial arrangement. The reverse is also true: a carrier with many exchange points does not automatically make a customer's particular circuit redundant. Procurement should ask Siglo for the logical and physical diversity that matters at the offered address—separate upstreams where claimed, separate entrances, power independence, automatic failover, tested capacity and the points at which two paths converge.

Valid route-origin authorisation is positive. It reduces one class of routing mistake by allowing networks to verify that AS273254 is authorised to originate the covered prefixes. It does not validate the entire path, stop every route leak, secure the rooftop radio, protect the subscriber's router, encrypt application traffic or prove fast recovery. Network-resource evidence should be read at its proper resolution.

The chronology is nevertheless informative. The ASN record dates from October 2024, during the period between Siglo's commercialiser status and the February 2025 concession. It shows that the company developed a visible routing identity as its regulatory footing evolved. Watchpoints now include changes in originated prefixes, invalid route-origin state, new or withdrawn upstream paths, IPv6 delivery to customers and reachability during an upstream failure. Those changes would say more about operational maturation than a generic claim of owning “the network.”

The contention hidden inside “up to 100”

Siglo's current home page combines the word “average” with “up to 100 Megas” for download and upload. Those terms answer different statistical questions. “Up to” is a ceiling or tier description; “average” requires a population, period and measurement method. A buyer needs the registered tariff and the applicable average busy-hour speeds for the exact package, not an attempt to merge both formulations.

The IFT's fixed-service quality framework is useful here. It requires fixed providers to report audited quality information and treats modem-based measurements of average download, upload and latency as material. It also says advertising should communicate average speed during the busiest period rather than only a maximum. That does not tell us Siglo's measured result, but it establishes the questions the operator should be able to answer: What is the package's nominal speed? What average is registered? Which hours define peak demand? Where is the measurement point? How are failed tests treated?

Siglo's practices code calls the service “best effort” and says performance can vary with network traffic. That is normal for consumer shared access, but it does not make every result acceptable. The Mexican minimum rights charter for telecommunications users describes rights to promised quality, information, equipment warranty and compensation or discount for attributable service failure. The registered contract and current tariff should translate those rights into a practical complaint and credit process.

The radio layer makes peak-hour testing especially important. The 2025 antenna supplier case reported signal and retransmission improvements after changing a sector antenna. This is a plausible mechanism: a better pattern can reduce unwanted interference and improve signal-to-noise ratio. Yet a stronger radio signal does not guarantee a faster evening application if sector airtime, backhaul or transit is saturated. Nor does an unloaded daytime speed test describe a household's experience when neighbouring users stream simultaneously.

Fixed wireless also needs an availability budget, not merely a speed number. The International Telecommunication Union's Recommendation P.530-19 addresses design and propagation for line-of-sight terrestrial links, including clear-air and rain effects. It does not assess Siglo. It explains why a competent radio design should document path clearance, frequency, distance, fade margin, expected availability and mitigation. The lower 5 GHz access layer and higher-frequency backhaul or access links face different propagation and interference conditions.

A household trial should therefore use a small measurement matrix. Connect a computer by Ethernet. Test at several times, especially 8–11 p.m., for at least seven consecutive days. Record download and upload throughput, latency to a nearby stable destination, jitter, packet loss, session interruptions and time to recover after power returns. Repeat during rain if weather is relevant. Keep application tests—video calls, school systems, cloud backups, streaming—separate from synthetic speed. Ask support to diagnose one controlled issue and record the escalation path.

The goal is not to demand dedicated-enterprise performance from a low-cost household service. It is to learn the shape of the variance. A stable 40 Mbps connection can be more useful than an occasional 100 Mbps peak that collapses at night. For a shop or home worker, upstream stability and latency may matter more than download maximum. For a family, fair sharing across devices and the router's indoor reach may dominate. “Unlimited” answers the data-volume question; it leaves capacity, quality and recovery open.

Wholesale dependence starts where the public map fades

Siglo controls important parts of the customer experience, but no regional internet provider is self-contained. The public record supports an operator with customer-premises installations, access radios, an ASN and visible IP routes. It does not reveal a complete schedule of tower ownership, rooftop leases, fibre purchases, radio licences, transit commitments, payment suppliers, cloud systems or outsourced support. Dependencies should be identified rather than assigned without evidence.

At the access edge, Siglo depends on permission to enter and use private property. It needs a suitable mounting point, safe access, an electrical supply and a clear radio path. It may depend on landlords or neighbours when the resident does not control the roof. The equipment depends on vendors for firmware, spares and support. The later Cambium-hosted case and earlier Siklu account show two supplier families at different layers and times, but public material does not disclose fleet concentration or replacement stock.

At the site, the operator needs power, security, mounting structures and a route onward. Backhaul can be wireless, fibre or a mix. The Siklu account says millimetre-wave backhaul was selected in early Neza deployments partly to avoid waiting more than a month for new fibre. That is a supplier's historical explanation, not proof that current sites lack fibre. It captures the strategic trade: wireless can compress deployment time, while fibre may offer different capacity and weather characteristics. A mature network may use each where its economics fit.

Beyond aggregation sits wholesale internet connectivity. The visible AS path points to Flō, but a route collector cannot show invoice terms or the last physical common point. Siglo's retail margin must absorb the committed wholesale rate, burst or overage structure, cross-connects, address resources, equipment and support. If transit costs fall with scale while installation remains local and labour-intensive, growth improves some unit costs and strains others. A new household on a lightly used sector may be attractive; the next household on a full sector can trigger a costly upgrade.

Payment is another wholesale layer. OXXO, Walmart-group counters, banks, card systems, Openpay and local SigloPay entities extend collection reach. They also introduce reconciliation, fraud, availability and privacy dependencies. If the network is working but the payment reference cannot be generated, the prepaid product is unavailable. If cash is accepted but the account is not credited promptly, the operator has collected money without delivering the purchased time. Payment uptime belongs in service operations.

The app and web account are control surfaces rather than the access network itself. They display expiry, produce payment references, record loyalty points, change a Wi-Fi password and support installation checks. A failure there may not stop an already active radio session, but it can prevent renewal or support. This is why a prepaid ISP should separate and monitor four states: physical link, internet session, account entitlement and payment confirmation. A customer-facing status should say which one is failing.

The most dangerous procurement shortcut is to ask whether Siglo “owns its network” and accept yes or no. Ownership is granular. The company may own some customer devices, radios or routers while leasing sites, buying backhaul and purchasing transit. The contract may make the customer owner of the indoor modem after payment while reserving antenna or network equipment for Siglo. Operational control, repair rights and replacement times matter more than a blanket label. The right schedule lists each component, owner, maintenance party, return condition and dependency.

Privacy begins on the installation form

The security boundary starts before the first packet crosses the roof. Siglo must collect enough information to identify the customer, qualify an address, enter the premises, activate an account, take payment and support the service. Its privacy notice names the Mexican legal entity and lists a broad set of data: name, email, telephone, date of birth, sex, marital status, device and IP information, image and social-network information. It says information may be gathered in person, by photograph, telephone, email, website or app, and used for service, diagnostics, complaints, demographics, research and marketing.

The notice also contemplates transfers in Mexico or abroad to a parent, affiliates and Siglo Ventures Inc., and allows images to be used in advertising. It explains access, rectification, cancellation and objection rights. Those disclosures create questions that a simple home-internet purchase should answer clearly. Which fields are necessary to deliver service? When is a photograph required? Does a roof or installation photo include household members or neighbouring property? Which corporate recipient receives which field? How long are failed-lead, former-customer, support-chat and payment-reference records retained?

The Android listing adds a notable inconsistency. The Google Play page for Siglo Internet says the app can show service expiry, generate payment references, manage SigloPay points, change a password and validate installations. The developer's data-safety declaration says no data is collected or shared, while also stating that data is not encrypted and cannot be deleted. The privacy notice describes extensive collection across the wider service. These statements could refer to different scopes or be out of date; they are not enough to diagnose a technical flaw. They do need reconciliation.

A buyer should ask for a current data map that separates the mobile app, website, customer account, payment processor, WhatsApp support, field-installation record and network logs. It should identify purposes, legal bases, recipients, retention, deletion processes and safeguards. A mobile-store declaration is self-reported, not an independent security assessment. A privacy notice is a legal disclosure, not evidence that every control operates as described.

Technical security also spans several owners. The household should receive a unique Wi-Fi password and instructions to change it. Installer access should be time-limited and attributable. Router firmware and remote-management responsibility should be stated. Account recovery should resist takeover of a telephone number or WhatsApp session. Payment references should not expose more identity data than necessary. Support workers should verify a customer without asking them to send sensitive material into an uncontrolled conversation.

At the radio and network layers, encryption, management-plane separation, device inventory, patching, configuration backup, logging and denial-of-service response matter. Public route-origin validity is one favourable control at the internet edge, but it does not answer any of those questions. Nor does the absence of a reported breach prove security. The evidence reviewed did not reveal a public independent security assessment, a subprocessor list, a technical security paper or a detailed retention schedule for Siglo.

The proportional standard is not that a regional ISP publish the same assurance library as a global cloud provider. It is that the company can explain what it holds, who can reach it, how access is removed, what happens when a device is replaced, how customers are notified and how records leave the company after cancellation. Because field labour, photographs, cash and messaging sit close to the customer, privacy practice must be designed into daily work, not confined to a footer.

Failure without a public status history

No credible public source in the frozen evidence set established a material Siglo outage, security breach or regulatory sanction. That sentence should not be converted into “Siglo has had no incidents.” The public site does not expose a long-term status archive, achieved-availability series, fault volumes, mean restoration time or incident reports from which such a conclusion could be drawn.

The practices code describes several failure categories. It treats the service as best effort and discusses interruptions caused by non-payment, government or court action, violence, cyberattack, infrastructure theft, weather, line-of-sight changes, customer handling and force majeure. It says customers should be notified through SMS or WhatsApp within six hours for force-majeure interruption and describes time compensation for extended interruption; attributable failures remain subject to legal remedies. These are contractual positions, not observed incident statistics.

The taxonomy is useful for operations. Expiry is not a network outage. Loss of household power is not a remote-site failure. A misaligned antenna differs from sector congestion. A backhaul cut differs from upstream route loss. A payment-credit delay is a service failure even when packets could flow. Support should code these causes consistently and publish enough aggregate information for customers to understand reliability.

A useful public history would include start and restoration times, affected zones, paid-time credits and plain-language causes without exposing security-sensitive detail. It would also disclose planned maintenance. In a prepaid model, the fairest remedy is often an automatic extension of the active period, because the customer bought days rather than a calendar subscription. The system must preserve that extension accurately across the next recharge.

Procurement should request twelve months of address-relevant or sector-level availability, ticket and restoration data. If Siglo cannot share it, the trial period carries more weight. The absence of public reports increases uncertainty; it is not itself evidence of poor service. The correct response is a smaller initial commitment, better measurement and explicit remedies.

Three clocks of competition

Siglo competes in Mexico at an awkward moment. The national fixed market is moving quickly toward fibre even as many individual addresses remain difficult to serve. An IFT release based on OECD data reported that fibre's share of Mexican fixed-broadband access rose from 41.1 per cent at the end of 2022 to 64.5 per cent at the end of 2023. This means fixed wireless should not be evaluated only as a remedy for total absence. In many zones it competes against a rapidly improving wired alternative.

Current household offers make that pressure visible. Telmex's home page listed a 120 Mbps internet-only package at MXN349 per month when accessed for this article, subject to address technology and current promotional terms. That is only MXN10 above Siglo's thirty-day base price. The comparison is not one-for-one: Telmex is a recurring wired service with its own installation, contract, promotion and coverage conditions; Siglo is prepaid and fixed wireless. But where fibre is actually installed, Siglo cannot rely on price alone. It must win on availability, speed of connection, payment flexibility, local support or the customer's dislike of a continuing bill.

A second clock is cellular-style fixed access. Bait's official Internet en Casa page describes a fixed-address home service that uses a customer kit and online recharges and says it cannot simply be moved away from the registered address. Such offers substitute wide-area mobile infrastructure and retail distribution for a roof-aligned local WISP link. Their trade-offs can include device cost, radio coverage, speed and data-policy limits. For a household, the comparison should be based on the actual allowance, post-allowance behaviour, indoor signal and peak-hour performance—not the shared word “prepaid.”

The third clock is satellite. Starlink's Mexican residential plans listed Residential Lite at MXN1,050 per month and Residential at MXN1,350, with unlimited data and a thirty-day trial at the observed date. Satellite reaches places that a local terrestrial sector cannot, but equipment, sky view and a monthly price roughly three to four times Siglo's thirty-day package change the buying case. In dense Neza, a clear path to a nearby local site and a clear view of the sky are distinct installation tests.

The deeper competitive issue is time. A fibre operator amortises a cable and seeks a continuing monthly account. A mobile-derived home offer sells a device, coverage and a data policy. A satellite operator sells a terminal and access to a global constellation. Siglo sells short windows over a locally installed fixed link. Its advantage is strongest when a household values rapid local installation, unlimited volume, cash recharge and the right to let service lapse.

Its weakness is strongest when fibre is already at the door at nearly the same price, when roof access is unavailable, or when a customer needs documented business-grade restoration.

Mexico's demand base is large but uneven. The INEGI ENDUTIH 2024 results estimated that 73.6 per cent of households had internet through fixed or mobile means and that the urban individual-use rate, 86.9 per cent, remained well above the rural rate of 68.5 per cent. It also found substantial urban-rural gaps in online banking and payments. Siglo's store and cash channels are therefore not incidental distribution; they are part of reaching customers who use the internet but may not organise household bills around online financial tools.

No national statistic resolves the household choice. Availability must be checked at the address; price must include installation, equipment and channel fees; performance must be measured at busy times; support must be tested; and switching must be understood. Competition in access is a set of local counterfactuals, not a national speed table.

No compulsory term does not mean no switching cost

The registered Siglo agreement says the service is indefinite and can be cancelled without an early-termination penalty. The public offer emphasises no compulsory term. Those are real forms of flexibility. They do not make the switch to or from the service costless.

The first cost is physical. A household must grant roof access, accept a mount and cable route, wait for an installation and keep equipment powered. If it moves to another provider, it may need another survey and another installation. The second is equipment. Siglo's FAQ says the modem becomes the customer's property after the full package amount is paid. Other public terms refer to a promissory note for a Siglo-owned antenna or network equipment and to recovery on cancellation.

The exact order should resolve which device is sold, lent or secured; its replacement price; normal-wear rules; removal work; and whether an unreturned item creates a debt.

The third cost is service learning. The household knows how to obtain a Siglo reference, where to pay and which support conversation can reach the local team. A substitute may require a bank card, a new app, a continuing debit or a different shop. The fourth is technical. A home worker may have configured cameras, printers or remote access around the existing router and address translation. A new provider may change indoor coverage, IPv6, public-address availability, port behaviour and latency.

The fifth cost is uncertainty. Cancelling a working prepaid link before proving the alternative can turn a nominally easy switch into downtime. The prudent method is overlap: install and test the replacement, move critical devices, verify billing and support, then close the old service and document equipment return. Because Siglo can be allowed to expire, a household may keep the physical installation as an occasional fallback if the contract permits and equipment obligations are clear. That option has value, but it may leave dormant hardware and personal data with the provider.

Siglo's own switching risk runs in the opposite direction. A customer can cease recharging without a cancellation conversation. The operator may see silence rather than a reason: lost income, fibre arrival, move, poor radio performance, support frustration or seasonal pause. Retention analytics must distinguish temporary dormancy from permanent loss without pressuring customers into an unwanted term. A respectful exit survey, equipment process and data-deletion route can produce better evidence than repeated sales messages.

The address-level procurement test

A household buying seven days can learn by use. A school, clinic, shop, community space or home-based business needs a more explicit test because connectivity failure has downstream cost. The procurement unit should still be the actual address, not the company's national concession or a generic coverage map.

1. Bind the seller. The quotation, registered contract, privacy notice and receipt should name SIGLO REDES DE MEXICO SAPI DE C.V as provider. The order should cite the current tariff registration and distinguish the Mexican provider from the US parent, payment firms, installer and equipment suppliers. Any promise made by a salesperson should appear in the order.

2. Freeze the all-in price. Record installation, router, antenna, mounting, cable, tax, deposit or promissory-note amount, payment-channel fee, package price, renewal price and move or removal charge. Confirm whether MXN339 buys exactly thirty consecutive days from payment and how overlapping recharges are treated. Do not compare that number with a monthly fibre price until all one-time costs and promotion expiry are included.

3. Survey the path. Require the serving zone or site, frequency family, mounting point, line-of-sight result, expected signal, cable length, grounding and weather sealing to be recorded. Ask what happens if later construction blocks the path. Photograph the completed work with consent and avoid capturing unnecessary household or neighbouring detail.

4. List the assets. Identify the indoor router, power supply, rooftop unit, mount and cable by make, model and serial number. For each, state owner, warranty, maintenance party, replacement time and return rule. Reconcile the FAQ's modem-ownership statement with the contract's treatment of antenna or network equipment before signing.

5. Define the service tier. Obtain nominal and registered average peak-hour download and upload speed, latency measure, any traffic-management rule and the meaning of unlimited data. Ask whether address translation is shared, whether a public IPv4 address is available, whether IPv6 is delivered and whether inbound connections are supported. These details affect cameras, remote work and business systems.

6. Run a wired acceptance test. For at least a week, measure throughput, latency, jitter, packet loss and interruptions over Ethernet at morning, afternoon and evening periods. Note weather and household load. Use several destinations and retain results. A single installer speed test is evidence of activation, not of sustained busy-hour quality.

7. Test power and recovery. Confirm whether the rooftop unit is powered from the home and how long service takes to return after a brief outage. Ask about site backup power. If connectivity is important, price a compatible household uninterruptible power supply and verify that it supports both router and rooftop device without unsafe modification.

8. Exercise support. Open a low-risk question through WhatsApp and any telephone channel. Record response time, identity verification, diagnostic quality and escalation. Ask for priority and restoration treatment outside the Monday-to-Saturday published window. Obtain the credit rule for attributable failure and the process for extending prepaid days.

9. Map dependencies. Ask which backhaul and upstream diversity serves the zone, where claimed paths converge and when failover was last tested. The public view of Flō as an upstream is a useful starting point, not the answer. Ask also about critical equipment lead times and whether a failed sector radio has a local spare.

10. Reconcile privacy and app statements. Obtain the current data map, retention periods, corporate and payment recipients, deletion route, app transport safeguards, field-photo practice and installer access controls. Ask why the Android declaration and general privacy notice appear to describe different collection and protection states. Require a direct answer rather than assuming either document is comprehensive.

11. Review incident evidence. Request twelve months of zone-relevant availability, maintenance, ticket and restoration measures plus any material security notifications. Ask how the operator distinguishes expired entitlement, payment failure, home power, indoor Wi-Fi, antenna alignment, sector congestion, backhaul loss and upstream failure. Require customer communication and credit rules for each applicable class.

12. Rehearse exit. Before the initial trial ends, document cancellation, data deletion, equipment recovery and final account closure. If an antenna remains, establish who may remove it and who repairs the mounting surface. Test an alternative connection before a critical use depends on Siglo alone.

These tests are intentionally practical. They do not require the provider to reveal sensitive network design or promise enterprise-grade service at a household price. They require evidence for the claims that determine whether the product works: the right seller, a feasible path, an honest price, measurable performance, reachable support, controlled data and a clean exit.

Watchpoints after the first recharge

Siglo's public record is unusually rich for a young regional access provider in some areas and thin in others. The legal entity, registered consumer agreement, concession and autonomous system are visible. Prices, cash fees, installation requirements and support hours are public. Supplier accounts give glimpses of radio and backhaul engineering. Missing are consistent operating series that would let an outsider test how the model performs as it scales.

The first watchpoint is coverage evidence. The concession permits national service, while the application and historical accounts point to a specific eastern Mexico City cluster. Siglo should be judged by verified serviceable addresses, connected sectors and capacity, not by the concession boundary. Expansion into a new municipality should be accompanied by local support and restoration readiness, not only sales availability.

The second is peak-hour capacity. A price cut and an unlimited-volume offer can accelerate activation. Watch registered average speeds, sector load, retransmission, backhaul upgrades and complaints during evening use. Supplier antenna gains are promising engineering evidence only when they persist across seasons and translate into household quality.

The third is renewal economics. Monitor the mix of seven-, fifteen-, thirty-day and long packages; payment-channel fees; failed or delayed credits; dormancy; reactivation; and equipment recovery. Growth in registered users can coexist with weak paid-day utilisation. The durable measure is contribution after installation, field service, site, backhaul, transit, payment and support.

The fourth is network independence. Changes in AS273254's routes, valid route-origin coverage, IPv6 delivery and observed upstream paths will indicate technical development. They must still be read cautiously. More prefixes are not automatically more customers, and more visible paths are not automatically physically diverse. Published failover and restoration evidence would be more valuable.

The fifth is document consistency. Current tariffs, the registered contract, website terms, commercial-practices code, privacy notice and app declarations should tell one compatible story. The old unrelated site reference, unfinished clauses and conflicting data-safety language are not proof of bad conduct; they are avoidable uncertainty. Version dates and a clearly controlling document set would improve trust.

The sixth is incident transparency. A public maintenance and outage history, even a modest one, would let customers distinguish silence from reliability. Automatic paid-time extensions should be visible on the account. Security disclosures should explain scope, customer action and lasting repair when an incident warrants notice.

The seventh is competitive response. Fibre prices are now close enough to Siglo's thirty-day price in served areas that flexibility and execution must carry the proposition. If Siglo raises speed, cuts price or expands, the question is whether local radio capacity and field quality move with demand. If it retreats to zones fibre does not reach, the question is whether those zones can support the site and backhaul economics.

The final watchpoint is the household's own counterfactual. Prepaid access is valuable because it preserves choice. The service succeeds when a resident can connect quickly, pay through a suitable channel, use the promised days with tolerable variance, receive help locally and stop cleanly. It fails when flexibility at the billing edge hides rigidity elsewhere—an obstructed roof, ambiguous equipment debt, congested sector, opaque outage or payment that does not activate.

SIGLO REDES DE MEXICO SAPI DE C.V has built enough of the chain to leave a verifiable public footprint. Its exact legal identity is connected to a real consumer service, a regulatory path, active internet routes and field deployments. The harder claim is not whether it is a real operator. It is whether the company can keep the recharge, radio and repair clocks aligned as households and traffic grow.

That is the operating system of prepaid home internet. The interface is a payment for seven, fifteen or thirty days. The product underneath is a continuing promise that the rooftop path, local worker, account ledger, upstream connection and consumer right will all be ready when the customer chooses to switch the internet back on.