Skynet Lintas Nusantara and the Trust Cost of Cheap Access in East Java

The second bill is the one that decides the margin

The Indonesian household that buys the cheapest access plan does not see the provider's economics on the first invoice. It sees a number: Rp110,000 a month for Skynet's listed 5 Mbps home package, Rp165,000 for 10 Mbps, or Rp195,000 for 15 Mbps, all described on the company's Sky Home page as unlimited, fiber-optic service with tax included (https://sky.net.id/sky-home/). The first bill is a promise that the internet can be made local, simple and affordable. The second bill arrives when that cheap line fails.

It is paid in mobile data bought in a hurry, a shop owner losing orders, a student missing a class, a small office moving files through a phone hotspot, or a reseller being blamed by neighbours even though the fault may sit somewhere upstream. That second bill is not printed by Skynet. It is imposed by downtime. In a low-price fixed-broadband market, that hidden bill is what turns a nominally cheap access service into either a trusted utility or a churn machine. The customer does not ask whether a BGP path is elegant. The customer asks whether someone answers the phone, whether a technician can reach the pole or rooftop, and whether the provider knows where the failure sits.

That is the right opening frame for PT Skynet Lintas Nusantara because its public materials do not describe a pure digital business. They describe a physical, repair-heavy access network based in Randuagung, Singosari, Malang, with home broadband, business bandwidth, cable pulling, reseller or subnet partnerships, and a route presence that reaches Jakarta exchange facilities. The constraint that decides margin is therefore not only backhaul price. It is the compound constraint of field repair, customer churn and Jakarta/upstream dependency. If the connection breaks often, cheap access destroys value. If local repair is credible and exchange reachability keeps common traffic cheap, the same low price can create a defensible neighbourhood franchise.

Skynet's public record is unusually direct about this bargain. Its homepage says it has operated in information technology since 2020, is centered in Malang, and offers home and business internet across a stated backbone footprint (https://sky.net.id/). It advertises legal ISP authorization from Kominfo under number 729/TEL.02.02/2020 and an operational fitness certificate under number 441/TEL.04.02/2021. Its company profile adds a separate telecommunications-network licence number, 346/TEL.01.02.2021, and an OSS commercial number, 0218000911659 (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). Its APJII member page lists PT Skynet Lintas Nusantara as an ISP member with registration 0810, brand SKYNET, domain SKY.NET.ID and the same Randuagung address in Kabupaten Malang (https://www.apjii.or.id/anggota/penyelengara?legality=&name=&page=28).

Those details matter because cheap access fails differently in Indonesia than it does in a dense fiber market with abundant duct, low retail churn and mature wholesale competition. Java has the population density to make broadband attractive, but it also has too many networks chasing similar customers. APJII commentary reported by AEI said Indonesia had 1,270 ISPs in February 2025, up from 600 in 2021 and around 300 before the pandemic; more than 800 were on Java, while fixed broadband penetration remained below 15% and industry revenue was being squeezed by price competition (https://aei.or.id/en/press-release/challenges-and-opportunities-for-internet-infrastructure-in-indonesia). Skynet's customer bill therefore sits inside a paradox: demand for home and small-business connectivity is real, but the supplier field is crowded enough that the monthly price alone cannot support a lazy network.

The company is interesting for that reason. It is not a national incumbent, a mobile operator, or a large data-center carrier. It is a small regional ISP trying to convert local presence, licensing, reseller support, exchange ports and a modest address block into trust. The question is not whether Skynet can beat the largest Indonesian operators on brand or national coverage. It almost certainly cannot. The question is whether it can own the local moment when the cheap connection fails and the customer needs a person, a ladder, a spare router, a route fix, or a clear answer.

Identity is clearer than the usual micro-ISP, but still needs boundaries

Skynet's legal and market identity can be triangulated from several public records. The company presents itself as PT. Skynet Lintas Nusantara, an internet service provider based at Gg VIII No.01, RT.03/RW.07, Gondang, Randuagung, Malang 65153 (https://sky.net.id/). The APJII member directory records PT SKYNET LINTAS NUSANTARA, brand SKYNET, as a "Keanggotaan Penyelenggara" with ISP licensing and domain SKY.NET.ID at Randuagung GG VIII, RT 03, RW 07, No. 01, Randuagung, Singosari, Kabupaten Malang, Jawa Timur 65153 (https://www.apjii.or.id/anggota/penyelengara?legality=&name=&page=28). PeeringDB lists the organization as PT Skynet Lintas Nusantara, also known as SKYNET, with the same Malang address and website override http://sky.net.id (https://www.peeringdb.com/org/29667).

The same identity appears at the network layer. PeeringDB's network record for AS141091 identifies the network as "Skynet Lintas Nusantara," organization PT Skynet Lintas Nusantara, also known as SkyNet, with network type Cable/DSL/ISP, two IPv4 prefixes, two IPv6 prefixes in the self-reported field, 5-10 Gbps traffic, and an open peering policy (https://www.peeringdb.com/net/27096). IDNIC/APNIC RDAP identifies AS141091 as SKYNET-AS-ID, registered on 2020-11-17 and last changed on 2024-03-25, with administrative and technical contact "Hostmaster SKYNET" at the Randuagung/Singosari address. RDAP for 103.156.128.0 - 103.156.129.255 names the block SKYNET-ID, type ALLOCATED PORTABLE, also registered on 2020-11-17, with a last-changed date of 2022-04-21 (https://idnic.rdap.apnic.net/autnum/141091 and https://idnic.rdap.apnic.net/ip/103.156.128.0).

That is a better public identity trail than many small ISPs expose. It links a legal company name, a market brand, a physical office, an APJII membership, a website, an autonomous-system number, and a portable IPv4 allocation. It does not, however, prove every commercial claim on the website. The Kominfo licence numbers are stated by Skynet and repeated in its company profile, while the publicly accessible APJII listing supports the fact that the company is treated as an ISP member. A reader should still distinguish a current regulator-standing check from a company-stated licence number. The strongest public conclusion is that Skynet is a real Indonesian ISP brand with APJII membership and directly registered network resources. The weaker conclusion, which would need government portal confirmation or documents from procurement, is that each licence and operational certificate remains in full current standing with no conditions.

The company's own language is also revealing. The homepage says Skynet has operated in IT since 2020 and is "salah satu penyedia layanan internet terkemuka di jawa timur," one of East Java's leading internet service providers (https://sky.net.id/). Glints lists PT. Skynet Lintas Nusantara in Kabupaten Malang, East Java, industry Internet, company size 1-10 employees, and verification status January 2024 (https://glints.com/id/companies/pt-skynet-lintas-nusantara/b72084cc-ba17-4c40-a335-2ae2f828f539). That personnel range should be treated as a platform listing rather than a payroll audit, but it matches the operating feel of the public record: a small company with a narrow management surface, not a large telco.

The company profile published on FlipHTML5 in July 2022 says Skynet's subnetworks were spread across Tasikmalaya, Sidoarjo, Madura, Bengkulu and Batam, and shows a "ready" and "plan" infrastructure map that includes Malang, Bali, Ponorogo, Mataram and Gorontalo as either current or future points (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). That document is useful because it shows management ambition and the business model Skynet wanted buyers and partners to see. It should not be read as proof that each market remains live in 2026. The live website is more cautious: it says Sky Biz using Metro bandwidth covers several cities in Java, Sumatra and Kalimantan, while Sky Home remains around branch-office areas (https://sky.net.id/).

This boundary is important. Skynet is not a mere web listing, but neither is it publicly proven as a nationwide retail provider. It is best understood as a Malang-centered licensed ISP and reseller-support platform with visible Jakarta interconnection and a small routed address footprint. That definition is narrower than the company's aspirational marketing. It is also more valuable for analysis, because the economics of a local ISP are decided by the gap between the claimed map and the routes, technicians and support desks that actually hold customers.

The business model is a legal umbrella plus local access

Skynet sells two related businesses. The first is direct access: Sky Home for households and Sky Biz for business users. The second is enablement: partnerships for RT/RW NET operators, BUMDes and subnet resellers who want to sell local connectivity under a clearer legal and technical umbrella. The official site places these next to each other. Sky Home is marketed as fast home WiFi over fiber, with packages from 5 Mbps to 15 Mbps and no quota limit (https://sky.net.id/sky-home/). Sky Biz is framed as a way for RT/RW networks to become a Skynet branch or service office, with a reseller or subnet arrangement, 1:1 bandwidth, legal distribution, resale rights, setup and maintenance support, and technical help for MikroTik or network issues (https://sky.net.id/subnet/).

That second line is the more distinctive one. Indonesia's neighbourhood broadband ecosystem includes many community-level access sellers, apartment networks, village operators, WiFi resellers and informal or semi-formal RT/RW networks. Some have strong field relationships but weak licensing, weak upstream purchase terms, or limited route engineering. A licensed ISP can monetize that gap by selling not only bandwidth but legitimacy, backbone pickup points, billing discipline, technical escalation and a recognizable brand. Skynet's 2022 company profile is explicit: product service includes SKYBIZ, SKYHOME and SUBNET/KEMITRAAN; the partnership product says Skynet "menaungi legalitas" for RT/RW NET actors, provides operational/commercial permission context, operational-fitness documentation, redistribution permission, dedicated 1:1 bandwidth, Metro Ethernet, radio and fiber last mile, and managed service for hotels, cafes, offices and housing (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/).

This model changes how one should read the small address block. Skynet's routed IPv4 space is only 103.156.128.0/23, or 512 IPv4 addresses, split into two visible /24 announcements in RIPEstat and BGP.tools (https://stat.ripe.net/data/announced-prefixes/data.json?resource=AS141091 and https://bgp.tools/as/141091). A company with only 512 IPv4 addresses is not built to give every household a unique public address. It must use NAT, private addressing, reseller aggregation, or customer-specific public-IP allocation. That is not a criticism; it is normal for small access ISPs in a scarce IPv4 environment. But it means the asset is not the address block alone. The asset is the operating system around that block: who gets public IPs, how abuse is handled, how customers are segmented, how resellers are supported, and how upstream traffic is bought.

The direct home offer looks like an entry product, not the only profit pool. At Rp110,000 for 5 Mbps, a provider has little room for repeated truck rolls, bad debts, pole damage, router replacement and high transit dependency. Skynet's terms therefore matter. The homepage says service is prepaid, payment is due on the 20th of each month, the minimum subscription is three months, and if a customer terminates early the penalty is calculated from monthly fee times remaining contract months plus arrears if any (https://sky.net.id/). It also says customers may not connect Skynet service outside the registered home address, with sanctions possible. Those terms are not decorative. They are the commercial defenses of a low-price ISP: reduce receivable risk, reduce early churn, and prevent a household line from becoming an uncontrolled resale line.

The reseller product has similar discipline. The Sky Biz page lists an initial registration fee of Rp5,000,000, a purchase order for the bandwidth needed, pickup from the nearest Skynet backbone point using cable or wireless, and monthly payment plus tax by the 10th of each month (https://sky.net.id/subnet/). It is a small but revealing set of terms. Skynet is asking the reseller to share setup risk, commit to ordered capacity, and bring the last-mile problem to a defined pickup point. That is the economic line between a local access wholesaler and a national operator. Skynet cannot make every village cheap by itself. It can make selected local operators easier to legalize, supply and troubleshoot.

A Scribd-posted invoice, not an official publication and therefore only a weak market signal, shows a 2024 bill labelled 1G Metro E service from PT. Skynet Lintas Nusantara at Rp11,100,000 for one monthly period (https://www.scribd.com/document/815196379/740-Invoice-Bandwidth-Oktober-2024-Agus-Sukoanyar). If genuine, it suggests a wholesale or high-capacity customer paid roughly two orders of magnitude more than a home customer. That is consistent with the business model: a few reseller or business circuits can carry more margin than many very cheap household lines, provided support costs and upstream costs are controlled. Because the document is user-posted, not company-hosted, it should not be used as audited revenue evidence. It is useful as a price-level clue.

The business model is therefore not "small ISP sells internet." It is more specific: a Malang ISP sells cheap home access, structured business capacity, reseller legitimacy and field support. Its best market is not the anonymous national broadband buyer. It is the household, small office, cafe, local institution, village body or RT/RW operator that wants someone nearby to carry the legal, routing and repair burden.

Route visibility gives Skynet credibility, but not scale

Skynet's network-resource evidence is concrete. PeeringDB records AS141091, network type Cable/DSL/ISP, two IPv4 prefixes, two IPv6 prefixes in the PeeringDB field, 5-10 Gbps traffic, a mostly inbound ratio and an open peering policy (https://www.peeringdb.com/net/27096). The same record shows 10G operational peering at IIX-Jakarta on 123.108.9.10 with IPv6 2001:7fa:2:5::10a, and 10G operational peering at JKT-IX on 119.11.184.194. It also lists Datacenter APJII-Cyber in Jakarta Selatan as the interconnection facility. IIX-Jakarta's own PeeringDB page describes the exchange as the Indonesia Internet Exchange under APJII, with 783 peers, 804 connections and 16.0T total capacity, and says the program is meant to establish a national interconnection network for licensed Indonesian ISPs (https://www.peeringdb.com/ix/210). JKT-IX's PeeringDB page separately places Skynet at 10G on 119.11.184.194 (https://www.peeringdb.com/ix/2476).

That is meaningful for a small regional ISP. A 10G exchange port does not prove the company has 10G of paid transit, thousands of customers, or continuous use at port capacity. It does show that Skynet is visible at the core Jakarta interconnection layer rather than operating only as a hidden reseller behind another ISP. For a customer in Malang or a reseller in another city, that visibility matters because domestic content, CDN reachability and peer selection can change perceived quality. Many customer complaints about "slow internet" are really about congestion, poor local caching, bad upstream choice, overloaded WiFi, or a route that takes a strange path for popular content. IX presence gives Skynet at least a route-level tool to reduce those frictions.

BGP.tools gives a narrower live view. It shows AS141091 as active, allocated under APNIC, network type Eyeball, originating two IPv4 prefixes and zero IPv6 prefixes in its current view, with upstreams AS45296 PT Rabik Bangun Nusantara and AS24534 PT Trans Hybrid Communication (https://bgp.tools/as/141091). RIPEstat announced-prefixes also shows exactly 103.156.128.0/24 and 103.156.129.0/24 in the observed window ending 2026-07-03 (https://stat.ripe.net/data/announced-prefixes/data.json?resource=AS141091). Hurricane Electric's BGP page similarly reports two IPv4 prefixes, zero IPv6, 512 originated IPv4 addresses, 22 observed IPv4 peers and RPKI-originated valid status for the two v4 prefixes (https://bgp.he.net/AS141091).

The IPv6 discrepancy is worth naming. PeeringDB says IPv6 is supported and shows an IPv6 address at IIX-Jakarta; BGP.tools and Hurricane Electric show no live originated IPv6 prefixes. That could mean IPv6 peering is configured at the exchange while customer or originated IPv6 is not visible globally, or that PeeringDB data is ahead of live route practice. Either way, IPv6 should not be treated as a proven customer feature until Skynet publishes product terms or route data shows stable v6 origination. The small-ISP economics are still mostly IPv4 and NAT economics.

The upstream dependency is equally important. BGP.tools sees two upstreams: PT Rabik Bangun Nusantara and PT Trans Hybrid Communication. The public record does not disclose contract terms, committed rates, pricing or redundancy design. It does suggest that Skynet's reachability is not only an APJII exchange story. When content is not local to IIX/JKT-IX, or when a peer route is not available, Skynet depends on upstream suppliers. That is normal. It is also the source of margin risk. A small provider selling low monthly prices can look excellent when domestic traffic stays local and content caches are nearby; it can look weak when international transit, upstream congestion or a supplier outage determines the customer experience.

The route evidence therefore supports a balanced conclusion. Skynet is not a ghost network. It has its own ASN, portable IPv4 space, RPKI-valid route visibility, Jakarta exchange ports and named upstreams. It is also not a large backbone. Its public route footprint is small enough that customer trust depends on operational discipline more than raw network scale. The route layer gives Skynet a seat at the table. It does not by itself pay the field technician, control churn, or guarantee that a cheap home line works at night.

Pricing turns support into the scarce commodity

Skynet's direct home prices are low in absolute terms. The official Sky Home page lists 5 Mbps at Rp110,000 per month, 10 Mbps at Rp165,000 and 15 Mbps at Rp195,000, all described as unlimited, fiber-optic service with tax included (https://sky.net.id/sky-home/). Its older company profile showed 5 Mbps at Rp200,000, 5 Mbps plus TV at Rp280,000, 10 Mbps at Rp250,000 and 10 Mbps plus TV at Rp330,000 (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). The current site appears cheaper than the 2022 profile, which is consistent with the national story of falling rates and price pressure.

The economics of such plans are unforgiving. A 5 Mbps line at Rp110,000 leaves limited monthly gross revenue after tax, upstream cost, local infrastructure, customer equipment, pole or rooftop work, payment collection, customer support, and regulatory levies. If one technician visit consumes several hours, fuel and replacement parts, the month's margin can disappear. If the same household churns after the minimum term, acquisition cost is not recovered. If a reseller overuses a pickup link or allows poor WiFi practice to damage the brand, Skynet may absorb reputation costs without controlling the last few meters.

This is why the service terms are as important as the advertised speeds. Prepayment reduces collections risk. The 20th-of-month due date creates a billing rhythm. The three-month minimum term protects installation economics. The ban on reconnecting service outside the registered address protects the provider from uncontrolled resale that would turn a household subscription into a small unlicensed network (https://sky.net.id/). Low-price broadband is not only a technical product; it is a contract discipline product.

The broader market increases the pressure. Opensignal's November 2025 Indonesia fixed-broadband report compares eight major providers: Biznet Home, CBN, Icon Plus, IndiHome, Indosat HiFi, MyRepublic, Oxygen.id and XL Home. It says Telkomsel's fixed-line IndiHome and FWA Orbit services together give the incumbent a 67% fixed-broadband market share, while ICON+ controls nearly 9%, MyRepublic around 6.5%, and XLSMART more than 6% after the XL-Smart merger (https://insights.opensignal.com/reports/2025/11/indonesia/fixed-broadband-experience). In East Java, Opensignal reports Biznet leading download speed at 35.1 Mbps, Indosat HiFi leading consistent quality at 66.2% and reliability at 429 points, and Biznet close behind on reliability at 421 points. Skynet is too small to appear in that national measurement set.

That absence is not a failure; it defines the competitive lane. Skynet cannot sell itself as the measured national champion. It must sell itself as a local provider whose support can be more reachable than a large brand's call center, or as a legal and technical partner for neighbourhood resellers whose end customers may never buy directly from a national ISP. The price ladder also has to be understood locally. A 10 Mbps line at Rp165,000 can be attractive where larger brands do not serve a street, where installation fees are lower, or where the customer's alternative is mobile data. It is less attractive where Biznet, MyRepublic, IndiHome, Icon Plus or Indosat offer faster packages with reliable installation and service.

Skynet's business product changes the revenue curve. The Sky Biz page says subnet partners pay Rp5,000,000 at registration, order bandwidth, pick up from the nearest Skynet backbone point and pay monthly after activation (https://sky.net.id/subnet/). The 2022 profile advertises business service with dedicated ratio 1:1, 99.8% SLA, and broadband business ratio 1:4 (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). These products may carry higher revenue per relationship, but they also carry higher support expectations. A reseller with dozens or hundreds of households can churn from Skynet in one decision if upstream performance or field response disappoints. A business customer will not tolerate consumer-grade ambiguity around outages.

The most plausible revenue logic is therefore mixed. Skynet likely needs direct home customers to build local brand density, reseller/subnet partners to raise bandwidth volume, and small-business/corporate users to improve average revenue per support case. Too much home-only growth raises repair burden. Too much reseller growth increases brand risk through third-party last mile. Too much business ambition requires stronger service assurance and redundancy. The attractive middle is where Skynet sells enough capacity to resellers and business users to fund a serious NOC and field team, while its home product keeps the local brand visible.

The cost base is hidden in poles, routers, power and waiting time

For a local ISP, the cost base is more physical than the retail website suggests. Skynet's homepage says it prioritizes fiber optic for all available services, while wireless via antenna may be used where installation conditions require it (https://sky.net.id/). The Sky Biz page says pickup from the nearest backbone point can use cable or wireless, and it specifically mentions help with MikroTik configuration and network troubleshooting (https://sky.net.id/subnet/). The Glints job posting for a fiber-optic network technician in Malang describes installation, maintenance and repair of fiber optic cables, problem analysis, routine maintenance, reporting and compliance with safety and technical rules; the listed salary band was Rp1,000,000 to Rp3,000,000 per month for a closed on-site position requiring one to three years of experience (https://glints.com/id/opportunities/jobs/teknisi-jaringan-fiber-optik/bfcc5f0d-5108-49c3-9a06-40edc8f2195c).

That job posting is not an audited staffing plan, but it points to the real operating surface: skilled technicians are the scarce asset. If a home customer pays Rp110,000 a month, even a low-wage technician visit can wipe out several months of margin after equipment, transport and overhead. If the company underpays or understaffs field work, customer trust deteriorates. If it staffs heavily, the low-price product cannot carry the cost. The local-ISP margin is therefore a scheduling problem as much as a bandwidth problem.

The route layer adds another cost. Skynet's presence at IIX-Jakarta and JKT-IX suggests it values domestic peering, but peering is not free after one counts port charges, transport to Jakarta, equipment, colocation, cross-connects, technical staff and monitoring. The public record places Skynet at Datacenter APJII-Cyber in Jakarta Selatan for interconnection (https://www.peeringdb.com/net/27096). A Malang customer, however, is not physically in that facility. Skynet must carry traffic from the customer access area to regional aggregation and then to Jakarta exchange or upstream points. The farther the last mile and aggregation path from controlled fiber, the more the provider depends on leased capacity, wireless backhaul, partner networks or reseller discipline.

Power and site access are also costs. A small access network needs active equipment in offices, rooftops, cabinets, towers or customer sites; it needs UPS batteries where outages matter; it needs replacement routers and optical parts; it needs vehicles or motorcycles; and it needs people who can reach a fault during rain, holidays or local events. Skynet's own customer support hours on the homepage list Monday to Saturday 08:00-17:00 and Sunday 09:00-15:00, while the 2022 profile says customer service standby 24/7 with outage notifications and payments integrated in one application (https://sky.net.id/ and https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). The gap between public-hours contact and claimed service excellence is exactly where customer trust is won or lost.

Regulatory costs are another layer. Komdigi's Ministerial Regulation No. 5 of 2021 defines BHP Telekomunikasi as an obligation of telecom operators and KPU/USO contribution as a contribution based on a percentage of gross telecom revenue; it states KPU/USO is 1.25% of gross telecom revenue and sets reporting, payment and verification duties (https://jdih.komdigi.go.id/produk_hukum/view/id/768/t/peraturan%2Bmenteri%2Bkomunikasi%2Bdan%2Binformatika%2Bnomor%2B5%2Btahun%2B2021). For a small ISP, these levies are not the whole cost structure, but they matter because they turn revenue into compliance exposure. Skynet's reseller pitch partly monetizes this: local RT/RW operators get a legal umbrella and distribution permission context, while Skynet carries some of the formal burden.

The highest cost, however, may be waiting time. When a cheap connection fails, customers do not immediately know whether the fault is their router, the drop cable, a local wireless link, a reseller aggregation point, a city fiber cut, a Jakarta exchange issue, or upstream congestion. The provider must diagnose quickly enough to keep trust. The 2022 company profile lists an escalation procedure of 0-10 minutes through Skynet's application or email, 10-30 minutes to an operational manager, and 30-60 minutes to a director (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). That is a bold promise for a small company. If it works, it is a competitive advantage. If it is merely brochure text, the second bill of downtime returns.

Upstream dependency is the hinge between local trust and national reach

Skynet's local promise depends on infrastructure it does not fully own. BGP.tools reports two upstreams for AS141091: PT Rabik Bangun Nusantara and PT Trans Hybrid Communication (https://bgp.tools/as/141091). The PeeringDB record adds local exchange presence at IIX-Jakarta and JKT-IX, both at 10G, plus APJII-Cyber as facility (https://www.peeringdb.com/net/27096). That combination means Skynet can improve domestic traffic economics through peering, but it still depends on upstream suppliers for routes that are not directly peered or cached.

This dependency is not unusual. Every small ISP depends on upstreams, exchanges, tower landlords, pole access, fiber routes, router vendors and content networks. The question is whether those dependencies are diversified enough for the product being sold. If Skynet is selling 5-15 Mbps home service in a narrow local area, two upstreams plus exchange peering may be adequate if field support is strong. If it is selling 1 Gbps business or reseller capacity across multiple provinces, customers should ask harder questions about path diversity, backup transport, committed information rates, DDoS handling, route filtering, maintenance windows and service credits.

The company profile's "strong backbone network" page says Skynet had IIX, CDN, East Java, West Java and Central Java segments, data center JKT, latest technology, redundancy link, total potential 100 Gb and AS141091 as partner context (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). Because this is self-published and dated 2022, it should be treated as a statement of intended architecture rather than current proof. The route records provide the proof that matters most in 2026: the ASN is live, two IPv4 /24s are visible, and exchange memberships are listed. They do not confirm a 100 Gb potential, current customer counts, or current capacity on regional segments.

This distinction matters for customer dependency. A household buying a cheap plan may tolerate occasional congestion if a technician responds and the price is low. A reseller cannot tolerate chronic uncertainty because its own subscribers will blame the reseller. A small office may tolerate lower nominal speed if the connection is stable, but it will not accept vague support during working hours. Skynet's ability to turn route visibility into trust therefore depends on whether it translates upstream events into clear customer communication.

There is also a route-quality argument. PeeringDB reports Skynet's traffic ratio as mostly inbound (https://www.peeringdb.com/net/27096). That is typical for eyeball access networks: customers download more than they upload. Inbound-heavy traffic makes domestic caching and peering valuable. If common content can be reached through local peers or caches at IIX/JKT-IX, upstream transit costs and latency pressure decline. If traffic shifts to platforms, games or cloud services not well reached through those paths, upstream dependency rises. A small provider's cost base can change without any change in retail price simply because the customer mix of video, gaming, cloud backup or work applications changes.

Supplier dependency also creates negotiation risk. Large operators buy upstream and equipment on better terms. A small operator may pay more per Mbps, per router, per cross-connect or per spare. It can compensate through lower overhead, local relationships and reseller focus, but those advantages are not automatic. The market will punish a small ISP that tries to price like a national discount provider while buying inputs like a small customer. Skynet's most defensible route is not to be the cheapest in every market. It is to be cheap enough, close enough and responsive enough that the customer values the local relationship.

Customer dependency is the real moat if it exists

The strongest argument for Skynet is not that it has a rare route asset. It is that certain customers may depend on a provider close enough to solve mundane problems. The live website says customers can register through contact channels, contact customer service by email, and visit or contact the Randuagung office (https://sky.net.id/). The Sky Biz page promises technical support for reseller setup and troubleshooting, while the blog post on RT/RW NET support says Skynet offers high-speed connectivity, scalable packages and reliable technical support for RT/RW operators (https://sky.net.id/subnet/ and https://sky.net.id/skynet-rtrw/). The 2022 company profile speaks of one application for customer service, 24/7 support, outage notification, payment and partner revenue information (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/).

These claims should be tested, not simply accepted. But they reveal what Skynet thinks customers buy. It is not only Mbps. It is a support relationship. A household pays for internet that does not require the customer to understand upstream suppliers. A small office pays for someone who will diagnose whether the problem is the line, router, wireless, payment status or broader network. A reseller pays for a legal and technical sponsor that can help keep the reseller's own customers calm. A cafe or hotel pays for visitors not to complain about WiFi. In all cases, the real product is reduced anxiety.

That anxiety is more acute when the national market is crowded. AEI's APJII-sourced discussion says the number of ISPs rose rapidly while fixed broadband remained underpenetrated, causing price competition, infrastructure stacking and smaller margins on Java (https://aei.or.id/en/press-release/challenges-and-opportunities-for-internet-infrastructure-in-indonesia). In a market like that, customers can be promised many things by many sellers. Trust becomes local and behavioural. Did the installer arrive? Did the support number answer? Did the outage message come before customers called? Did the reseller get a clear explanation? Did the company admit when an upstream or exchange issue was involved?

Skynet's public materials include signs of this trust strategy. Its Facebook and Instagram presence under Skynet Lintas Nusantara and skynetlinus markets promotions, contact numbers and local access products. Its Instagram profile says "Get Connected, Stay Connected!" and references Metro E promotions for Java; search snippets for 2025 posts show payment-method education and installation promotions (https://www.instagram.com/skynetlinus/?hl=en). The social-media following is not large enough to prove a major customer base, but the content is consistent with a company selling to households and small partners through direct contact rather than through anonymous online checkout.

The Glints profile cuts both ways. A 1-10 employee listing suggests a lean team, which can be good for local accountability but risky for 24/7 operations (https://glints.com/id/companies/pt-skynet-lintas-nusantara/b72084cc-ba17-4c40-a335-2ae2f828f539). The fiber technician job posting shows the company or its recruiter sought field repair capacity, but the low salary band also implies cost discipline and labour-market constraints (https://glints.com/id/opportunities/jobs/teknisi-jaringan-fiber-optik/bfcc5f0d-5108-49c3-9a06-40edc8f2195c). A small ISP's moat disappears quickly if the experienced technician leaves, the owner handles too many escalations personally, or the support team cannot scale as reseller customers grow.

Customer dependency therefore remains plausible but unproven. The public record supports the existence of local access and reseller relationships. It does not disclose churn, ticket volumes, mean repair time, service credits, active household count, reseller count, or customer satisfaction. Skynet's competitive claim should be judged by those missing metrics. A national operator can lose a few angry customers without changing strategy. A local ISP cannot. Its brand lives in the memory of the last outage.

Competition is not only national brands; it is every pole and reseller

Skynet competes against large fixed-broadband brands, mobile substitution, power-utility fiber, other local ISPs, informal resellers and its own partners' ability to switch upstream. Opensignal's market overview shows the national fixed-broadband field is led by Telkomsel/IndiHome with 67% share, with ICON+, MyRepublic and XLSMART also material players (https://insights.opensignal.com/reports/2025/11/indonesia/fixed-broadband-experience). In East Java, Opensignal's measured providers include Biznet Home, MyRepublic, Indosat HiFi, IndiHome and Icon Plus, with Biznet leading download speed and Indosat HiFi leading several quality metrics. These are not hypothetical competitors. They are the brands a Malang household or business may see on a billboard, sales page, neighbour recommendation or building sign.

The competitive threat from large brands is obvious: scale lowers procurement cost, national marketing raises trust, and larger network teams can standardize support. But large brands also leave gaps. They may not cover a particular street, building, village, boarding house, cafe, or small industrial cluster. They may take too long to install. They may treat a small business as a consumer ticket. They may not want to support a neighbourhood reseller arrangement. Skynet's opportunity is in those gaps.

The harder competition is from operators that look like Skynet. APJII's membership pages show hundreds of Indonesian ISP members, including many small providers with regional addresses, ISP licensing and local brand names (https://www.apjii.or.id/anggota/penyelengara?legality=&name=&page=28). AEI's APJII-sourced article says more than 800 ISPs were operating on Java alone, and that infrastructure was piling up in some places rather than spreading evenly (https://aei.or.id/en/press-release/challenges-and-opportunities-for-internet-infrastructure-in-indonesia). In such a market, a local ISP may compete with another local ISP on the same road, and both may compete with an RT/RW operator who has a different upstream.

That can create destructive economics. If every provider cuts the monthly price but must still pay for technicians, poles, backhaul and support, service quality eventually falls. A customer then churns not because the price was too high, but because the provider could not afford to maintain what it sold. The irony is that the cheapest operator may make the market worse for everyone if it underprices repair. Skynet's public terms show some awareness of this: three-month minimums, penalties and limits on unauthorized redistribution protect the network from the race to the bottom.

New fixed-wireless and satellite options add another layer. Indonesia's government has been preparing 1.4 GHz fixed-wireless broadband policy aimed at lower-cost home internet, and media reports have framed the target as affordable 100 Mbps service around Rp100,000 to Rp150,000 in some discussions (https://intimedia.id/read/komdigi-prepares-affordable-fixed-broadband-internet-rp-100000-for-100mbps). Starlink's Indonesian licensing and activation history also matters for remote or underserved areas, even if its economics differ from cheap urban Java fiber. For Skynet, these are not immediate one-for-one substitutes in every Malang neighbourhood, but they change customer expectations. A household that hears "100 Mbps for near my current price" will ask why 10 Mbps costs anything close to that. A reseller will ask whether wireless broadband can bypass a local upstream.

Skynet's likely answer has to be service density. It cannot win a national marketing fight. It can win by knowing exactly which streets, rooftops, resellers, cafes and small offices it can serve better than anyone else. That requires discipline: avoid overclaiming coverage, price enough for repair, keep upstreams honest, and make reseller support reliable. The competition is not only the provider with a bigger speed test. It is the provider whose customer knows the repair person by name.

Regulation and operating risk are inseparable from the resale pitch

Skynet's public pitch depends heavily on legality. The homepage says its Kominfo ISP licence is number 729/TEL.02.02/2020 and its SKLO is number 441/TEL.04.02/2021 (https://sky.net.id/). The company profile adds a telecommunications-network licence number 346/TEL.01.02.2021 and OSS commercial permission number 0218000911659 (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). The APJII directory lists the company as an ISP member (https://www.apjii.or.id/anggota/penyelengara?legality=&name=&page=28). The resale product then offers to support RT/RW NET actors with legality and redistribution permission context.

This is not a side issue. It is the central reason a small reseller might choose Skynet instead of simply buying bandwidth from the cheapest upstream. Indonesian telecom regulation imposes formal licensing, reporting, tariff, contribution and operational duties. Komdigi Regulation No. 5 of 2021 sets tariff principles and says operators must avoid tariff practices that disrupt consumer protection, healthy competition or service continuity; it also sets BHP and KPU/USO duties tied to gross telecom revenue, including a 1.25% KPU/USO contribution (https://jdih.komdigi.go.id/produk_hukum/view/id/768/t/peraturan%2Bmenteri%2Bkomunikasi%2Bdan%2Binformatika%2Bnomor%2B5%2Btahun%2B2021). A reseller without a strong legal structure may be cheaper in the short run but more fragile if the regulator, a landlord, a competitor or a customer challenges it.

For Skynet, regulation is both moat and risk. The moat is that legal documentation and APJII membership are valuable to partners. The risk is that the same documentation must remain current and operationally meaningful. If a licence lapses, a report is missed, USO/BHP payment is mishandled, or a reseller operates beyond authorized terms, the legal umbrella becomes a liability. The public record does not show such a failure. It also does not show a current regulator status page proving perfect compliance. A prudent buyer or reseller should ask for current documents, payment standing and service terms before relying on legality as the main reason to buy.

Operational risk is just as material. Indonesia's broadband field is shaped by aerial cable congestion, excavation, power interruptions, tower and rooftop permissions, weather, vendor supply, customer equipment quality, payment behaviour and local politics around poles or right of way. A small ISP may know its local geography better than a large provider, but it may also lack redundancy. Skynet's company profile talks about redundancy links and a Jakarta data-center presence, but the live public route evidence is the firmer proof (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/ and https://bgp.tools/as/141091).

Geopolitical risk is indirect but real. Indonesia wants broader broadband access, domestic digital resilience and more affordable fixed connectivity. Government support for new fixed-wireless bands or satellite service can change competitive economics. Large national operators may consolidate, push into smaller towns, or use bundled mobile-fixed offers to compress local prices. International content and cloud platforms may shift caching and peering patterns, changing upstream cost. None of this is a Skynet-specific crisis. It is the environment in which a small ISP must keep margin without controlling policy.

The key regulatory question is whether Skynet can keep its reseller model clean. The public terms prohibit a household customer from extending service outside the registered address (https://sky.net.id/). The Sky Biz product, by contrast, explicitly supports resale and partnership under defined terms (https://sky.net.id/subnet/). That separation is essential. If the company can keep consumer lines from becoming uncontrolled resale while channeling real resellers into licensed subnet arrangements, it protects both revenue and compliance. If that separation breaks, the provider may face congestion, customer disputes and legal ambiguity.

Unofficial signals show activity, but also small-company fragility

The non-official market signals around Skynet are useful because small ISPs often leave more evidence in job ads, social posts and customer documents than in formal filings. Glints says PT. Skynet Lintas Nusantara is a verified January 2024 company profile in Kabupaten Malang, East Java, with 1-10 employees and Internet industry classification (https://glints.com/id/companies/pt-skynet-lintas-nusantara/b72084cc-ba17-4c40-a335-2ae2f828f539). A closed Glints listing sought a fiber-optic network technician in Malang for installation, maintenance, repair and routine work, with IP networking, wireless networking and networking skills (https://glints.com/id/opportunities/jobs/teknisi-jaringan-fiber-optik/bfcc5f0d-5108-49c3-9a06-40edc8f2195c). Those signals support the view that Skynet was hiring for the work its business requires: field repair and network operations.

Social media adds current market colour. The Instagram account skynetlinus describes itself as the official account of Skynet Lintas Nusantara business in Malang, with the tagline "Get Connected, Stay Connected!" and posts about Metro E promotions, payment methods and installation promotions (https://www.instagram.com/skynetlinus/?hl=en). Facebook search results show PT. SkyNet Lintas Nusantara promoting Metro E bandwidth for Java and local promotions in Malang areas. Social data can be noisy and incomplete, but it confirms a live sales posture rather than a dormant network registration.

The FlipHTML5 company profile is another mixed signal. It is visually promotional and was published by an individual account, not necessarily a formal investor-relations channel, so it should not be treated as audited disclosure. But its text is detailed enough to show how Skynet wanted to explain itself: founded in 2020, headquartered in Malang, subnetworks across multiple Indonesian regions, legal licences, network access, partnerships, digital business solutions, 2022 targets, 100 Gb potential, Metro Ethernet, reseller legality, managed service and escalation procedures (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). That is valuable because it reveals management's strategic vocabulary. Skynet saw itself as more than a neighbourhood WiFi seller. It wanted to be a legal, technical and commercial platform for smaller access sellers.

The Scribd invoice signal is more delicate. It appears to show a 2024 invoice for 1G Metro E service at Rp11,100,000 to a named customer, with PT. Skynet Lintas Nusantara bank details (https://www.scribd.com/document/815196379/740-Invoice-Bandwidth-Oktober-2024-Agus-Sukoanyar). Because it is uploaded by a third party and includes customer details, it should be handled cautiously. It should not be repeated as a confirmed customer relationship beyond the fact that such a document is publicly visible. Economically, though, it aligns with the reseller/high-capacity product and gives a plausible reference point for business-bandwidth pricing.

The website itself is also a signal. It contains typos, mixed Indonesian-English labels, WordPress design artifacts and repeated content. That does not mean the network is poor; many small infrastructure companies have weak websites and strong field teams. It does mean the brand presentation is less mature than larger competitors. In a market where trust is the product, web hygiene matters more than small operators sometimes admit. A broken or sloppy website can make a customer question whether the same carelessness applies to billing, tickets or security. Conversely, Skynet's live site does expose useful terms, prices, contacts and licence numbers, which is better than hiding everything behind WhatsApp.

The absence of broad public complaint chatter should not be overread. Small local ISPs often produce few searchable reviews because customers complain in WhatsApp groups, local Facebook groups, reseller channels or direct calls. Lack of complaints is not proof of satisfaction. It may simply mean the customer base is small or local. The best unofficial signal would be repeated independent customer comments about outage response, installation times, real speeds and reseller treatment. Public search did not surface enough of that to draw a strong conclusion.

Overall, the unofficial signals support a living, small, operationally active ISP with a reseller orientation. They also reveal fragility: small team, promotional documentation, uncertain current coverage breadth and a public brand surface that is less polished than national competitors. For a local ISP, that fragility is not fatal. It is exactly what management has to convert into trust through rapid repair and honest communication.

What would change the judgement

The current judgement is cautiously constructive. Skynet Lintas Nusantara appears to be a real Malang-based ISP with APJII membership, visible network resources, Jakarta exchange presence, cheap home access, business/subnet products and public evidence of field-support hiring. Its economics make sense if reseller and business bandwidth carry enough contribution margin to fund local repair, while domestic peering and upstream choices keep the cheap home product from becoming a loss leader. Its risk is that Indonesia's crowded Java broadband market compresses price faster than a small field team can maintain quality.

Several facts would materially change that view. The first is current licence status. Public APJII membership and company-stated Kominfo/SKLO numbers are useful, but a current regulator portal record, recent certificate copy or procurement-grade documentation would strengthen confidence. A finding that a licence had lapsed, a levy was unpaid, or a reseller permission was invalid would weaken the entire legal-umbrella thesis.

The second is customer count and mix. Skynet's 2022 profile listed targets and claims around home, corporate and partner counts, but public sources do not verify current active subscribers (https://fliphtml5.com/dewxr/krfw/Company_Profile_PT._Skynet_Lintas_Nusantara/). If most revenue comes from a few resellers, the company is exposed to partner churn. If most comes from very cheap home accounts, repair cost may overwhelm margin. If a balanced mix of home, reseller and business customers exists, the model is more durable.

The third is repair performance. Mean time to respond, mean time to repair, first-time fix rate, outage notification quality and repeat-ticket rates would tell more about Skynet's value than any speed headline. The company's own escalation promises create a standard. Evidence that customers receive clear messages within minutes and meaningful field action within hours would raise the rating. Evidence of chronic silence or repeated unresolved failures would lower it.

The fourth is upstream and backhaul contracting. BGP shows two upstreams and exchange presence, but not committed capacity, diversity or cost. If Skynet has physically diverse backhaul from Malang to Jakarta and resilient upstream contracts, its trust proposition is stronger. If the route path depends on fragile leased links or one practical supplier despite two visible upstreams, the cheap-access promise is more vulnerable.

The fifth is actual coverage. The live website says Sky Home remains around branch-office areas while Sky Biz covers several cities across Java, Sumatra and Kalimantan (https://sky.net.id/). The 2022 profile lists broader ready and planned regions. A current coverage map, active reseller list or installation data would settle whether Skynet is a Malang specialist, a multi-city wholesaler, or a brand whose earlier expansion plan did not fully materialize.

The sixth is price discipline. If Skynet keeps cutting entry prices without raising business/reseller margin or improving automation, the cost of field support will eventually catch it. If it can use local density to lower install and repair cost, the low prices may be sustainable. A public tariff table for Metro E, reseller tiers and service credits would clarify whether the business product subsidizes the home product or whether each line stands on its own.

The final fact is customer reputation. Searchable public chatter is too thin to settle trust. A pattern of local recommendations, verifiable business references, stable reseller testimonials and transparent outage communication would support the thesis that Skynet turns local support into a moat. A pattern of unresolved complaint threads, payment disputes, route instability or inactive contacts would invert the thesis.

Until those facts are known, Skynet should be read as a credible but exposed small ISP. It has enough public infrastructure evidence to be taken seriously and enough market pressure to make complacency dangerous. Its margin is not decided by a logo, an ASN, a licence number or a cheap plan by itself. It is decided on the day a customer's cheap connection fails and Skynet either absorbs the second bill through repair, communication and route control, or lets the customer pay it alone.