SATELIT SERVIS and the price of owning network reliability

Summary

• SATELIT SERVIS has a real local operating footprint: an address-level Kramatorsk access network, public retail tariffs, 52 reported employees, AS9205, 9,216 uniquely originated IPv4 addresses, an IPv6 route and three currently observed routing neighbours. Those records establish an operating internet provider, but not physical route diversity, delivered uptime or profitable contracts.
• Filed figures reproduced by a Ukrainian company-data service show 2025 revenue rising 18.5% to UAH26.12 million and the net loss narrowing to UAH612,400. The improvement matters, but it was still a fifth consecutive annual loss; liabilities rose 37.6%, faster than assets, while the company raised its main household tariffs again in May 2026.
• The economic opportunity is to charge for reachable local support, resilient PON access and competent route control in a frontline city. The downside is unusually concentrated: customers can evacuate, cables and equipment can be damaged, national operators and another local fibre provider offer close substitutes, and a small set of upstream paths may have to carry more reliability than public records can verify.

The UAH225 reliability question

The first price in SATELIT SERVIS's current household offer is UAH225 a month. That buys up to 40Mbps over twisted pair or 60Mbps over PON, depending on the connection. UAH255 buys up to 100Mbps on copper or 120Mbps on PON. The May 2026 notice adds 300Mbps for UAH265 on either access type and 1Gbps PON for UAH325. A public IPv4 address costs another UAH30 a month. These are not luxury prices. They are the narrow monthly contributions from which a local provider has to fund everything between an apartment and the wider internet.

The incentive to own network capability rather than act as a pure reseller is straightforward. A provider with its own local plant, address resources, routing policy and technical staff can choose more than one external route, move traffic when a supplier fails, issue addresses directly, repair the access line and answer to the customer under one service obligation. It can also retain more of the retail payment than an intermediary that merely rebills somebody else's access. In a place where power loss, physical damage and curfew restrictions can delay repairs, that control can be valuable.

Control is not free. The May 2026 price notice says the company's equipment is regularly damaged by shelling and presents the increase as necessary for stable operation and rapid restoration. That is the company's account, not an independently costed damage schedule. It nevertheless identifies the bill that management is trying to recover: not just traffic, but repeated repair of a physical network in Kramatorsk.

The financial record shows why the price question is urgent. Opendatabot's company page, which reproduces filed annual figures, reports UAH26.12 million of 2025 revenue and a UAH612,400 net loss. Revenue improved by 18.5% from 2024 and the loss narrowed by 73.4%. Yet the company did not cross into profit. A provider can grow sales, migrate customers to fibre and announce more address space while still destroying value if the incremental payment is absorbed by suppliers, labour, repairs, taxes and depreciation.

SATELIT SERVIS's test is therefore stricter than survival. It has already survived years of war and remains visibly active. The economic test is whether local control produces a premium or a cost saving large enough to cover the capital and operating burden. Customers have to pay enough for independence, and enough of them have to remain in the service area, for reliability to become a return rather than a public duty financed by the balance sheet.

A Kramatorsk operator, despite the name

The word "Satelit" creates an obvious identity risk. Public evidence does not support treating this company as a satellite-communications operator. Its own company page identifies legal code 33270712, a Kramatorsk address and the SATELIT SERVIS name. Opendatabot gives the same code and address, a 17 November 2004 founding date and wired telecommunications as the principal activity. The current retail offer is Ethernet and PON home internet. Nothing in the reviewed service catalogue establishes a satellite fleet, satellite capacity resale or a satellite broadband product.

The operating boundary is local and terrestrial. The company's coverage page lists apartment buildings street by street across Kramatorsk and separately identifies 20 business-centre or commercial points of presence. That is stronger evidence than a generic claim to serve Ukraine. It shows where customer acquisition, cable access and field support are likely to be concentrated. It does not show how many addresses are active, how many have PON, what proportion are businesses, or how many subscribers have evacuated.

The company is also more than an access reseller. The RIPE Database connects legal code 33270712 to an active local internet registry organisation and AS9205. Current route collectors see prefixes originated by that autonomous system. The company's public terms describe an operator that takes applications, checks technical feasibility, installs a connection, maintains a personal account and provides continuing internet access. A 2025 decision by Ukraine's competition authority separately identifies SATELIT SERVIS as the provider of specified static and shared NAT addresses to subscribers in an unrelated procurement investigation. The company was an information provider in that case, not an accused party; the record is useful because it independently confirms real subscriber address assignment.

The legal company and the routing identity do not share the same birthday. Opendatabot reports that the company was formed in 2004, whereas RIPE records show AS9205 created in December 2001 and RIPEstat has an even earlier first-seen route observation. The safe conclusion is that the network identity has a long history. It would be unsafe to claim that the current limited company operated the same commercial network before it legally existed. Assets, maintainers or predecessor arrangements can change while an ASN remains.

That distinction matters because network records are evidence, not the company itself. An ASN cannot hire repair crews or earn a margin. A prefix does not reveal whether the customer pays on time. SATELIT SERVIS is the legal and operating company; AS9205 and its address space are tools used in delivering connectivity.

What the customer is actually buying

The basic household product combines access, installation, continuing route availability and local fault handling. The public service page charges UAH150 for a standard connection, including initial configuration of one computer and up to ten metres of cable into the premises. Extra cable is charged per metre. A technician visit inside the premises costs UAH100, with specified extras for cable and router work. The company says connection normally follows full payment and can take up to five working days. These details reveal a labour-bearing service, not a frictionless bandwidth subscription.

The payment model shifts some working-capital risk to the customer. The published service procedure defines a personal account funded by advance payments. The 2023 customer agreement says acceptance occurs through payment, after which an installation date is agreed. Prepayment reduces receivables and disconnection risk for the operator. It does not remove the need to buy equipment, hold spares, pay staff or secure upstream capacity before the monthly payment arrives.

For a household, the visible benefit is a working connection and somebody local to call. For a business, the value can be larger: a public address, a known local technician, connection at a listed commercial building and continuity for payments, administration, security or remote work. Public procurement records show small contracts for internet access to local public bodies, including a UAH7,200 2026 police-security contract and a UAH6,900 cultural-administration contract. These contracts establish a business-to-government service channel. Their modest size does not establish that public customers dominate revenue.

The paid unit is therefore best understood as an address or site under local responsibility. The buyer is not paying only for a maximum speed. It is paying SATELIT SERVIS to connect the site, authenticate the account, carry traffic to external networks, respond to faults and restore service. The economic premium should come from fewer coordination failures: the same provider owns the local customer relationship and controls at least part of the routing and access plant.

That premium remains unmeasured. No public source reviewed for this article gives active subscriber count, average revenue per user, churn, installation backlog, gross margin by access technology, mean time to repair, service credits or customer acquisition cost. Business tariffs are unavailable in a comparable schedule. The company also states that household service is unavailable to non-residential premises, so the retail table cannot be applied to business contracts.

Without those metrics, speed is a weak proxy for value. A 1Gbps line that fails when local power disappears may be less useful than a slower path with backed-up equipment. A public address may be worth UAH30 to one household and essential to a business. A technician who can reach an address quickly can justify a local premium, but only if repair outcomes are better than those of the national and local alternatives. Reliability has to be experienced and renewed; it cannot be inferred from the tariff name.

Growth has not yet become value creation

The annual figures describe a company that has recovered revenue without recovering profitability. Opendatabot reports revenue of UAH27.54 million in 2020, UAH30.39 million in 2021, UAH19.22 million in 2022, UAH17.81 million in 2023, UAH22.05 million in 2024 and UAH26.12 million in 2025. By 2025 sales had risen 46.7% from the 2023 low, but they were still 5.2% below 2020 in nominal hryvnia.

Profit moved in the wrong direction for most of that period. The same source reports UAH1.54 million of net profit in 2020, followed by losses of UAH625,200 in 2021, UAH1.07 million in 2022, UAH1.02 million in 2023, UAH2.30 million in 2024 and UAH612,400 in 2025. The 2025 net margin was negative 2.34%, a considerable improvement from negative 10.45% in 2024 but still a loss. Five consecutive loss-making years are not explained by one isolated repair.

The 2025 recovery is meaningful. Revenue added UAH4.07 million while the loss improved by UAH1.69 million. This could reflect tariff increases, more customers, higher business sales, restoration of service, a richer product mix or lower costs. Public filings reproduced on the page do not split the cause. It would be equally wrong to dismiss the improvement and to declare a turnaround complete.

The balance sheet adds caution. Assets rose 14.6% in 2025 to UAH12.26 million, while liabilities rose 37.6% to UAH7.80 million. The difference between reported assets and liabilities fell from about UAH5.02 million to UAH4.46 million, consistent with another annual loss. Liabilities were 63.6% of assets at year end, up from 53.0% a year earlier. These are small-company accounts, and the public summary does not show debt maturity, cash, trade creditors or lease obligations. It does show that creditors and other obligations funded more of the asset base as the company grew.

Reported headcount was 52 in both 2024 and 2025, up from 48 in 2022. Revenue per employee rose from about UAH424,000 in 2024 to UAH502,000 in 2025. That is a useful productivity signal, but not a labour-margin calculation. It excludes contractor labour, says nothing about wage inflation and cannot distinguish network engineers from sales, administration and field crews. The provider still lost about UAH11,800 per reported employee in 2025.

Public tender volume does not explain the recovery. Opendatabot lists UAH167,187 of 2025 tender sales, less than 1% of reported company revenue, after UAH557,149 in 2024. Tender databases can omit private contracts and do not capture all public purchasing arrangements, so the ratio is not a complete customer map. It does suggest that the visible procurement channel is too small to be treated as the main business.

The most plausible revenue base is household and private commercial connectivity in the local footprint, supplemented by modest public contracts. That remains an inference from the service, coverage and procurement evidence. The exact residential-business split is not disclosed. This uncertainty is important because each segment carries different bargaining power. Households are numerous but price-sensitive. Businesses can pay more for public addresses, support and continuity but may be concentrated. Public bodies can offer recurring demand but impose procurement and payment procedures.

Revenue growth becomes value creation only when the incremental gross profit exceeds the extra cost of routes, equipment, power, labour, tax and repair. SATELIT SERVIS's 2025 accounts got much closer. They did not yet cross that line.

A compressed price ladder

SATELIT SERVIS has raised the same headline household tiers in stages. Its August 2023 notice priced 40Mbps at UAH170 and 100Mbps at UAH200. The March 2025 notice moved them to UAH195 and UAH225. In May 2026 they became UAH225 and UAH255. Across the full interval, the entry tier rose 32.4% and the 100Mbps tier 27.5%.

The increases do not, by themselves, prove pricing power. Ukraine's annual consumer inflation reached 12.0% in 2024 and 8.0% in 2025, according to the National Bank of Ukraine. Equipment and software are exposed to foreign-currency and logistics costs, while frontline field work carries risks that national consumer inflation cannot capture. A nominal tariff increase can leave the real contribution unchanged.

The shape of the new ladder is more revealing than the percentage increase. Moving from 60Mbps PON at UAH225 to 300Mbps costs only UAH40 more. Moving from 300Mbps to 1Gbps costs another UAH60. The top tier is only UAH100, or 44.4%, more expensive than the slowest PON line. This is a classic fixed-network strategy: once fibre and equipment are installed and upstream capacity is available, the provider uses speed to move customers up the ladder because the marginal cost of an extra permitted megabit can be much lower than the installation cost.

The risk is that the ladder gives away too much throughput before capital is recovered. A gigabit port may require newer optical equipment, customer terminals, aggregation capacity and power backup. Peak demand, not advertised speed, determines upstream and switch costs, but SATELIT SERVIS publishes no peak traffic or oversubscription policy. If customers upgrade without increasing usage, the higher tier can be attractive. If a small number of heavy users consume the new ceiling and force capacity purchases, UAH325 may be a thin return.

PON can improve the equation. Passive splitters in the distribution network need no local power, reducing the number of powered field devices compared with an all-active Ethernet design. The optical line and customer terminal still need power, and the operator's central equipment, aggregation and upstream routes still need backup. The company does not publish the share of customers on PON, the cost per conversion or battery duration. The tariff notice proves that PON products exist, not that the entire footprint is power-resilient.

Pricing also reveals the limit of differentiation. A direct local competitor, Elite Line, lists apartment-building PON at UAH230 for 110Mbps, UAH260 for 300Mbps and UAH320 for 1Gbps. SATELIT SERVIS is UAH5 cheaper at the entry PON tier, UAH5 more expensive at 300Mbps and UAH5 more expensive at 1Gbps. That is near price parity, not a visible reliability premium. SATELIT SERVIS must therefore win on address coverage, restoration, route quality or trust, because headline price and speed alone offer little shelter.

The network is real, but redundancy is not yet proven

Current routing evidence is substantial for a provider of this financial size. RIPEstat's routing-status observation on 10 July 2026 showed AS9205 visible to all 327 IPv4 and all 321 IPv6 full-feed peers in the returned sample. It reported four IPv4 prefixes covering 9,216 unique addresses, one IPv6 /48 and three observed neighbours. The announced-prefix list identifies 87.76.224.0/19, 87.76.254.0/23, 193.110.112.0/22, 193.110.115.0/24 and 2a04:5e40:4::/48.

Two IPv4 announcements are more-specific routes inside larger announcements, so they should not be added twice when measuring unique space. The /19 contributes 8,192 addresses and the separate /22 contributes 1,024, for 9,216 unique IPv4 addresses. More-specific announcements can support traffic engineering or route control; they can also add routing-table entries. Public observations do not reveal the company's purpose.

Route-origin security is a positive. RIPEstat returned a valid RPKI state for each of the five observed prefixes when checked against origin AS9205. RPKI helps external networks reject an unauthorised origin announcement. It does not stop fibre cuts, power loss, router failure, congestion, distributed denial-of-service attacks or an authorised operator's configuration error. It secures one layer of route authorisation, not the whole service.

The resource footprint changed recently. The RIPE organisation object connecting SATELIT SERVIS and code 33270712 was created in December 2025. The record for the 87.76.224.0/19 block was created in February 2026 and names the company organisation as holder. The IPv6 /48 record dates from 2023. These dates show active resource administration and a much larger currently originated IPv4 pool. They do not reveal whether the /19 was bought, transferred, sponsored or obtained under another arrangement, nor what it cost.

Economically, 9,216 addresses can support customer assignment, carrier-grade NAT pools, business products or future growth. The company charges UAH30 per month for a public address, but it would be wrong to multiply every address by that price. Many addresses are required for infrastructure, routing, network services, reserves or blocks assigned under different products. Some may not be assigned to paying customers. Address count is capacity evidence, not recurring revenue.

The company's RIPE-linked geofeed asks geolocation users to map all listed IPv4 blocks to Kyiv. That does not mean every router or subscriber is physically in Kyiv. A geofeed is a statement about intended IP geolocation. The distinction matters in Donetsk because foreign platforms can apply sanctions or risk controls based on location databases. A 2024 local interview with the company's then director described Kramatorsk users losing access to foreign services when platforms associated their IP addresses with sanctioned territory. The Kyiv geofeed may reduce that problem, but the public evidence does not establish its motive or effectiveness.

One absence is equally important: PeeringDB's public API returns no network record for AS9205. That does not prove there is no peering, no exchange connection and no private interconnection. It means there is no self-reported PeeringDB entry from which to verify facilities, ports, traffic scale or policy. Public route collectors show reachability; they do not show where physical paths meet.

Independence is still bought from suppliers

An autonomous system gives SATELIT SERVIS policy control, not self-sufficiency. RIPEstat's neighbour observation currently sees AS3261, AS35297 and AS6768 adjacent to AS9205. bgp.tools labels them LLC FTICOM, Dataline LLC and EUROTELE-PLUS and classifies all three as upstreams. That classification is useful but derived; only contracts can establish who pays whom, how much capacity is committed and what service guarantees apply.

The registered AS9205 policy names AS29559, AS3261, AS6768 and AS44863 in import or export statements. The difference between registered policy and currently observed neighbours is not necessarily a contradiction. Registry policy can be historical or prospective, and collectors may not see every path. It is a reminder that neither source alone is a supplier invoice.

Three observed neighbours are better than one, but they do not prove three independent failure domains. Two logical routes can share the same duct, power source, city hand-off, long-haul fibre or upstream-of-upstream. A provider could have excellent BGP diversity and poor physical diversity, or modest visible adjacency and strong backup through routes not selected in the global table. SATELIT SERVIS publishes no fibre map, committed bandwidth, path-disjointness test, failover time or supplier concentration.

The company's own notices show that external routing changes are operationally material. One technical-work notice warned of possible interruptions during changes to external routing. This is normal network maintenance, not evidence of poor engineering. It does demonstrate that route independence has a labour and change-risk cost.

Supplier bargaining power is likely to be important because SATELIT SERVIS's total revenue is only UAH26.1 million. A national backbone or large upstream can spread routers, security systems, engineers and cross-border capacity across a much larger base. The local provider can offset that scale disadvantage by buying from several suppliers and aggregating local demand. It loses if supplier minimums, imported equipment or emergency capacity consume the retail margin.

The facts needed to judge independence are therefore commercial and physical: transit spend as a share of revenue, committed and peak capacity, route-level latency, packet loss, path overlap, failover tests and supplier terms. Public resource control is a necessary capability. It is not the same as low-cost independence.

A concentrated market that can physically leave

SATELIT SERVIS's most defensible advantage is also its greatest concentration risk. The coverage page shows a dense Kramatorsk footprint. Local technicians can know the buildings, ducts, distribution points and recurring faults better than a distant call centre. The provider can focus investment where it already has access and customer relationships. That is how a small local operator can compete with a national brand.

The address base is exposed to one city and one conflict. The Donetsk regional administration said 1,800 people left the Kramatorsk community in the first part of June 2026, the largest number among communities in that update. Official orders have expanded mandatory evacuation from specified Kramatorsk areas. A departed household no longer pays for a fixed line, even if the cable and port remain in place. An evacuated business can take its demand and staff with it. Fixed costs do not leave as quickly.

Attacks also remove demand and damage the plant simultaneously. The regional administration reported repeated strikes on Kramatorsk through 2025 and 2026, including residential and infrastructure damage. SATELIT SERVIS says shelling regularly damages its own equipment. These are separate evidence streams: official sources establish the city's continuing exposure, while the company attributes its own repair burden to attacks. Neither source quantifies the provider's cumulative loss.

The global reconstruction assessment shows the wider scale. The fifth Rapid Damage and Needs Assessment estimated total direct damage to Ukraine at $195.1 billion through December 2025; the detailed report put telecommunications, digital and media damage at about $2.5 billion. That national figure cannot be allocated to SATELIT SERVIS. It explains why replacement equipment, contractors, power and finance are contested resources across the sector.

Customer concentration is therefore geographic even if no single customer dominates revenue. The company has a local network in a place where population, buildings, power and purchasing capacity can change abruptly. Expanding to another city might diversify demand but would require new access rights, crews, aggregation and marketing. Selling business transport beyond Kramatorsk could diversify revenue but public materials do not establish that product at material scale.

The 2026 IPv4 expansion and Kyiv geofeed raise a strategic question. They may be part of an effort to operate a broader or more geographically flexible resource plane. They do not prove that local access revenue has diversified. The company needs paying lines in more than a registry database; it needs customers and infrastructure that are not all exposed to the same event.

Competition sets the ceiling on resilience pricing

The local customer has realistic substitutes. Elite Line sells PON in the same city at almost identical household prices and advertises service during power cuts. Kyivstar's Kramatorsk home-internet page invites address checks, offers fibre, and says 99% of its fixed network has backup power for up to 12 hours, while noting that performance still depends on equipment at the particular building. Mobile service can provide an immediate backup for lighter use. Satellite terminals can provide a different physical path where power and a clear view are available.

These alternatives are not interchangeable. A national operator may have deeper procurement, backbone and financing scale but slower local repair or less address-specific flexibility. A local fibre rival may match field knowledge and offer a passive optical path. Mobile capacity can congest during fixed-network outages and may be unsuitable for heavy or low-latency work. Satellite service carries equipment, subscription, power and sky-view constraints and is generally a more expensive household substitute.

SATELIT SERVIS can defend itself in four ways. First, it can cover buildings where rivals do not. Second, it can provide a better restoration record. Third, it can use multiple upstreams and good route policy to deliver lower loss and latency. Fourth, it can bundle local public addresses, business connectivity and support in a way that reduces customer coordination cost.

Only the first is clearly evidenced at address level. The route records support the capability behind the third but not measured quality. The tariff and contract documents support the fourth in form but not business revenue or margin. No public uptime series proves the second.

Price matching is a rational response when residents are under financial pressure and can switch. It is also dangerous for a company with a history of losses. A national operator can subsidise a local access promotion from a much wider product base. A local provider cannot indefinitely sell below full cost merely because a competitor does. SATELIT SERVIS needs either lower cost per active line or a service difference customers will renew.

The PON transition can deliver the lower-cost route. Passive distribution can reduce field power points, and higher speed tiers can raise revenue over the same fibre. Yet every conversion consumes equipment and labour before the monthly uplift arrives. The difference between UAH225 and UAH325 is only UAH1,200 a year. If a gigabit upgrade requires a new optical terminal, router work, aggregation and a service visit, payback can be long unless the equipment is inexpensive, customer-funded or shared across many lines.

Reliability in a frontline city is an operating process

The company's August 2024 repair notice is unusually informative. It named the affected streets and said work had to begin at 6am because the curfew prevented overnight repair. That turns an abstract outage into an operating constraint: even when staff and replacement parts are available, security rules can delay the lowest-impact maintenance window. Customers bear daytime disconnection; the provider bears labour, repair and reputational cost.

The current home page says support operates from 6:30am to 10pm. The 2023 agreement describes a round-the-clock telephone help service, and archived paid local-channel advertisements also claim 24-hour support. These statements do not align. The company may have changed hours, may distinguish consultation channels or may have outdated documents. Until it publishes a clear current service commitment, readers should not treat "24/7" as a verified response guarantee.

Ukraine's communications authorities describe a much broader resilience burden. The State Service of Special Communications says physical and cyber attacks have made stable operation and rapid restoration central missions, alongside coordination during emergency power cuts. The NCEC 2025 report says fixed-internet revenue grew 8.2% to UAH24.4 billion nationally and sector capital investment rose 35% to UAH33.9 billion. The market is growing, but growth is being earned in part through heavy reconstruction and resilience spending.

For SATELIT SERVIS, the resilience stack likely includes upstream contracts, optical and copper plant, switches and routers, customer terminals, power backup, monitoring, spares, transport and skilled people. "Likely" matters because the company does not disclose cost lines. Its 52 employees and service documents establish a labour component; routing and access records establish network components; tariff notices establish repair pressure. The mix and cost are unknown.

Reliability also needs a definition. It can mean access equipment stays powered, an upstream route remains available, DNS works, foreign platforms accept the IP location, a technician answers, or damaged cable is restored. A provider can perform well on one layer and fail on another. SATELIT SERVIS publishes no multi-year measure that joins them into an end-to-end outcome.

The buyer should therefore ask for observable terms: expected availability, backup duration, restoration target, escalation route, compensation, and whether a second path is physically separate. The provider should price those commitments. Giving every customer an undefined promise of reliability creates an uncapped obligation. Selling a measurable service creates a product.

Regulation and cross-border access add another cost layer

SATELIT SERVIS appears in the Ukrainian regulator's supplier register under code 33270712. Its contract says it was included in the updated register on 31 March 2023; a June 2024 NCEC appendix lists it at entry 2844. Registration supports legal operating status. It does not certify network quality or financial strength.

Compliance extends beyond registration. Providers submit regulatory reports, respond to lawful information requests and implement access restrictions required under Ukrainian sanctions and security measures. NCEC has reminded providers of monthly reporting concerning blocked sanctioned resources. SATELIT SERVIS's website says it pays profit tax under the general regime, while the State Tax Service has said suppliers of electronic communications services cannot use the simplified tax system under the post-2022 legal definition. These obligations increase the minimum administrative capability required of even a local provider.

Blocking carries technical and reputational costs. A 2020 Digital Security Laboratory study found that Kramatorsk's SATELIT SERVIS blocked more than 270 sites from the then sanctions list, placing it among the more complete implementers in the sample. That is historical evidence of compliance, not a current count and not a judgment on the merits of any restriction.

The cross-border problem can run in the opposite direction. Foreign services may block a Ukrainian user because commercial geolocation places the address in a sanctioned zone. The local interview about TikTok, Zoom and other services shows that internet availability is not only a matter of packet delivery. A route can work technically while an application refuses the customer. Correct registration and geofeeds therefore become part of the product.

This creates work that a pure reseller can pass upward but an autonomous operator must own: registry updates, geolocation corrections, abuse handling, RPKI, route objects, sanctions implementation and customer explanation. It can deepen SATELIT SERVIS's local value because the provider controls its address space. It also consumes skilled time that does not appear as a separate monthly line item.

Regulatory credibility can help win business and public contracts. It can also raise fixed costs per subscriber as the local base contracts. The same compliance team is more affordable across 10,000 lines than across 5,000. Active-line count is therefore one of the most important missing facts.

Who pays, who benefits and who carries the downside

The household or business pays the monthly tariff, usually in advance. The immediate beneficiary is the person or organisation using the connection. Wider beneficiaries include employers, schools, local administration, emergency services and relatives who depend on communication. SATELIT SERVIS benefits if it retains enough of the payment after direct network costs to fund replacement and profit.

Upstream carriers, equipment suppliers, landlords, power providers and employees are paid before the residual return is known. Their claims can remain even when customers evacuate or a damaged address stops billing. Creditors carry some risk, but the increase in liabilities means they also have a larger claim on the asset base. The owners carry the final loss if revenue and asset value do not cover those obligations.

Customers carry outage risk through lost work, interrupted services and the cost of backup. SATELIT SERVIS carries repair and churn risk. Public bodies carry continuity risk where local services depend on its connections. No source shows that the state guarantees the company's economics or that customers compensate it fully for war damage.

The distribution of benefits explains why reliability can command a premium. A business may lose much more from an hour offline than the difference between UAH255 and UAH325 a month. But that buyer should pay the premium only when the provider can specify the extra protection. The consumer tariff does not promise a second physical route, a restoration time or a service credit. The market cannot reliably price what it cannot observe.

The strongest strategy is not to describe independence. It is to allocate resources to it: PON conversion where power resilience and density justify the cost, diverse transit where paths are genuinely separate, spares where repair delay is expensive, and measurable business service where willingness to pay is highest. Anything else is marketing funded by a thin balance sheet.

Unofficial signals point to demand, not proven quality

The company is actively seeking customers. A public archive of paid posts in a Kramatorsk Telegram channel repeats claims of stable high-speed service, uninterrupted connection and 24-hour support. Individual posts attracted roughly 1,500 to 2,000 views in the visible archive. This is evidence of local marketing activity, not subscriber conversion, uptime or customer satisfaction.

The wording also lags the official offer. The archived advertisements mention speed up to 100Mbps, while the May 2026 tariff notice lists 300Mbps and 1Gbps products. The ads claim 24-hour support, while the current homepage gives 6:30am to 10pm. Repetition can build awareness; inconsistency can weaken trust.

The Work.ua company page describes SATELIT SERVIS as a regional internet provider. That supports market positioning but adds little operating evidence. Review sites and address aggregators contain too little verified, current data to support a service-quality conclusion. Sparse complaints are not proof of excellent reliability, especially when customers may resolve faults by phone, leave the city or switch providers without a public review.

The responsible conclusion from these signals is limited: the brand is active, still markets locally and is trying to sell reliability. Whether buyers believe and renew that promise must be tested against churn, uptime and cash, none of which is public.

Evidence register

The most important records support different parts of the judgment and have different limits.

Evidence	What it supports	What it does not prove
Company identity page	Legal name, code, Kramatorsk address, contact and general tax status	Ownership economics or service quality
2026 tariffs	Copper and PON prices, public-address price and company-reported damage pressure	Subscriber count, margin, uptime or PON coverage
Coverage page	Address-level Kramatorsk footprint and listed business locations	Active lines or current occupancy
Opendatabot	Filed revenue, profit, assets, liabilities, headcount, tenders and registry identity	Product economics, cash flow or audited causes of changes
RIPE member record	RIPE NCC membership and official service-region context	Sale of transit, cloud or managed services
RIPE Database AS9205	ASN registration, organisation, registered policy and maintenance history	Current contracts, path capacity or physical diversity
RIPEstat routing status	Current route visibility, unique announced space, IPv6 and observed-neighbour count	Traffic, latency, uptime or profitability
RIPEstat neighbours	Three currently observed adjacent networks	Paid transit terms or separate fibre paths
Company geofeed	Intended Kyiv geolocation for listed IPv4 space	Physical location of customers or routers
NCEC 2025 report	National fixed-internet growth and sector investment context	SATELIT SERVIS market share
RDNA5	National and sector damage context through 2025	Company-specific damage or compensation
Elite Line tariffs	A close local PON price and product substitute	Address-by-address availability or comparative quality
Kyivstar Kramatorsk offer	National-operator availability, fibre and public power-backup claim	Performance at a particular building
Competition-authority decision	Independent confirmation of subscriber static and shared address assignment	Wrongdoing by SATELIT SERVIS or current customer identity
August 2024 repair notice	Named outage area, restoration work and curfew constraint	Long-run outage rate or restoration average

What would change the judgment

The current evidence supports a real regional operator that improved sharply in 2025 but has not yet proved that local network ownership earns an adequate return. Several disclosures would turn that conditional judgment into a firmer one.

First, SATELIT SERVIS should show active lines by access type. Household Ethernet, household PON, business internet, public addresses and other services should be separated. Additions, disconnections and evacuation-related suspensions would show whether revenue growth came from price, volume or mix. The most important denominator is not address space; it is paying connections.

Second, it should disclose contribution economics without revealing customer names. Average monthly revenue, direct upstream cost, support cost, installation capital and gross margin per active line would show whether UAH225 can recover the service obligation. PON conversion cost and the share funded by customers would reveal the payback period. Business-service margin would show whether higher-value contracts subsidise household access.

Third, cash conversion needs to be visible. The public summary shows losses and rising liabilities but not operating cash flow, debt maturity, payables, cash or capital expenditure. A move to positive operating cash after maintenance investment would be stronger than revenue growth alone. Falling liabilities relative to assets would show that growth is funding itself.

Fourth, reliability must be measured. Monthly availability, material incidents, restoration time, support answer time, repeat faults and service credits should be published by access type. Backup duration at the central site and field nodes should be tested under realistic outages. A map of PON-enabled and powered-Ethernet buildings would let customers buy the right product rather than infer resilience from a technology label.

Fifth, route diversity should be demonstrated without exposing sensitive detail. The company can publish the number of contracted upstreams, aggregate committed capacity, failover test results and whether principal paths are physically disjoint. It can explain why registered policy and observed neighbours differ. A PeeringDB record is not essential, but transparent interconnection information would make the independence claim easier to assess.

Sixth, the recent resource changes need an economic explanation. Management could state whether the /19 association represents an acquisition, transfer, sponsorship or administrative consolidation; how much address space is in customer use; and why all listed IPv4 space is geofed to Kyiv. None of this requires revealing individual assignments. It would connect resource administration to customer value and cost.

Seventh, customer concentration needs boundaries. The share of revenue from the ten largest customers, households, business and government would show bargaining and collection risk. Geographic revenue outside Kramatorsk would indicate whether the company is diversifying. The tender record alone cannot answer these questions.

Finally, the support promise needs one current version. If telephone consultation is available around the clock but field or account support runs from 6:30am to 10pm, say so. If restoration targets differ between household and business products, price them. Reliable local accountability is the strongest potential reason to choose SATELIT SERVIS. Contradictory hours weaken the product at almost no saving.

A positive judgment would follow if 2026 brings positive operating cash, another improvement in net result, stable active-line count despite evacuation, increasing PON share, measured restoration performance and demonstrated path diversity. A negative judgment would follow if tariff increases coincide with continued losses, shrinking equity, rising obligations, subscriber contraction or repeated outages without compensation.

Owning the problem is valuable only when the price covers it

SATELIT SERVIS has already crossed the threshold that separates a nominal reseller from a real operator. It maintains a local access footprint, installs customers, assigns addresses, originates routes, runs IPv4 and IPv6 space and manages more than one observed external adjacency. Its 2025 revenue recovery and smaller loss show economic movement in the right direction.

The company has not yet crossed the more important threshold into demonstrated value creation. Five consecutive losses, rising balance-sheet obligations, a concentrated frontline service area and close price competition leave little room for error. The May 2026 increase may fund repair and PON migration. It may also be only the next attempt to catch costs that keep moving.

Customers who remain in Kramatorsk have a reason to value a provider that knows their buildings and owns the fault from the apartment to the upstream hand-off. Businesses have an even stronger reason to pay for reachable staff, public addressing and a tested backup path. The willingness to pay exists where an outage is costly. SATELIT SERVIS must turn that willingness into a specified service rather than an undefined promise.

The fairest conclusion is conditional. Network ownership gives SATELIT SERVIS the capability to create value through local accountability, PON resilience and route control. It also leaves the company carrying field repairs, supplier concentration, compliance, power and customer-flight risk. The UAH225 entry tariff is attractive to the buyer. Whether it is attractive to the owner depends on the missing facts: active lines, gross margin, cash, restoration time and physical diversity. Until those appear, reliability is a credible product ambition and a visible operating burden, but not yet a proven return.