Summary

  • The economic incentive in Daginfonet is not to become a national carrier. It is to keep enough dense Derbent access lines on-net that a 1,000 to 1,800 ruble residential bill, a 1,500 to 3,000 ruble commercial or legal-entity bill, and a small paid public-IP add-on can cover wholesale connectivity, field visits, power, compliance storage and periodic renewal capital.
  • Elias Ward's judgment is cautious but not dismissive: Daginfonet looks like a viable local access cash-flow business when it protects local density and support response, but it is not a scale story. The disclosed revenue decline into 2025, IPv4-only public routing footprint, thin public customer disclosure and visible substitute tariffs mean the margin is vulnerable if churn rises faster than the company can reprice.
  • The best evidence for resilience is not marketing language. It is the company's own willingness to raise prices in 2025 and again in 2026, the continued public visibility of its autonomous system and IPv4 aggregate, a wide cash-payment footprint around Derbent, business tariffs materially above household tariffs, and an open installer vacancy that shows continuing field demand.
  • The best evidence against resilience is that Daginfonet's public accounts show 2025 revenue of 43.933 million rubles after a 57.129 million ruble opening figure, while cost of sales was 35.019 million rubles. That leaves an implied gross spread of 8.914 million rubles, or about 20 percent of revenue, before treating every public number as complete enough to carry the whole company story.
  • The facts that would change the judgment are concrete: a current subscriber count, churn by tariff, take-up by street or building, upstream contract pricing, access-network age, capex backlog, bad-debt rate, cash collection leakage, any large enterprise account, and whether the observed 2025 revenue decline was volume loss, accounting timing, tariff migration, or a one-off reporting effect.

One Line, One Spread

Daginfonet's economics start with one active access line, not with the autonomous system. A household taking the 100 Mbps plan at 1,000 rubles a month is buying a local promise: the line works, the router is configured, the account can be paid without friction, and a technician can appear when the last drop cable, apartment entry or home router fails. At that price, the company is not selling raw internet capacity by the megabit. It is selling the last-mile bundle around it.

The price has to absorb the traffic mix of ordinary household video, messaging, gaming, software updates and shared apartment use, while the costs arrive through upstream connectivity, building access, support wages, power and equipment replacement.

That is why the first serious question is not whether Daginfonet's 100 Mbps headline speed is cheaper than every other advertised offer in Derbent. It is whether the retained monthly contribution per line stays positive after local density is considered. A regional ISP with dense clusters can do well at prices that look modest because the same field team, billing footprint and uplink can support many accounts within a small physical radius. A regional ISP with scattered customers has the opposite problem: every fault visit, customer call and node upgrade consumes more of the monthly fee.

Daginfonet's own public materials point to the first model. The company publishes a long list of payment terminals in Derbent, a central office on Pushkina Street, and residential, commercial, cyberclub and legal-entity plans. Those are signs of a city-centred access franchise rather than a diffuse backbone business.

The tariff table sets the basic revenue ladder. Residential unlimited plans are 100 Mbps for 1,000 rubles, 150 Mbps for 1,200 rubles and 200 Mbps for 1,800 rubles. The step from 100 to 150 Mbps lowers the price per headline megabit, while the step to 200 Mbps raises the bill enough to protect heavier usage. Commercial points pay 1,500 rubles for 50 Mbps and 2,000 rubles for 100 Mbps. Legal entities pay 2,000 rubles for 20 Mbps, 2,500 rubles for 50 Mbps and 3,000 rubles for 100 Mbps. Cyberclub plans reach 6,000 to 10,000 rubles for 100 to 300 Mbps.

Public IPv4 addresses are priced separately at 300 rubles a month for individuals and 500 rubles for legal entities.

The shape of that table matters. Daginfonet is using the same access plant to segment willingness to pay. A household line at 1,000 rubles can be profitable if support and churn stay contained. A legal-entity 100 Mbps line at 3,000 rubles can subsidize heavier service obligations and lower contention. A cyberclub at 6,000 to 10,000 rubles is a different bet: it can contribute strongly if usage is predictable and low-latency support is valued, but it can also expose the network to concentrated evening demand. The white-IP add-on is a small but telling yield tool because Daginfonet's visible public IPv4 block is only 1,024 addresses.

Selling public addresses separately pushes scarce address space toward customers that value it.

The company has already shown it will move price when costs move. In April 2025 it told subscribers that rising telecom service costs, equipment, electricity and related legal expenses forced a tariff revision. In February 2026 it repeated the explanation, adding tax burden and announcing that from April 2026 it would raise access speeds with a modest price increase and move subscribers automatically to new plans. That is not a cosmetic notice. It is the centre of the investment case. If Daginfonet cannot pass a meaningful share of inflation and compliance cost into the bill, its small scale will be punished.

If it can reprice without losing too many accounts, the local access franchise remains economically alive.

Identity, Boundary And Scale

The public identity boundary is reasonably clear. The operating company is LLC Daginfonet, a Russian limited liability company registered in July 2006, with INN 0542029112 and OGRN 1060542004731. The company site lists an operating address in Derbent and a legal address on Heydar Aliyev Street, gives the director as Ramazan Radzhabovich Asadullayev, and publishes contact numbers and an email address. Public business registries are broadly consistent on the registration identifiers, main activity and private-ownership status, although they are not identical on every executive-history detail.

That identity boundary is important because the visible network assets map back to the same name. AS209219 is shown in RIPEstat as "asdin LLC Daginfonet." The RIPE Whois record lists the autonomous system as assigned, with the as-name "asdin" and organisation ORG-LD111-RIPE. Multiple public BGP views associate 185.12.224.0/22 with LLC Daginfonet. The company is therefore not merely a reseller with a tariff page and no network-resource footprint. It has a registered AS, a visible IPv4 aggregate and a regulatory-facing communications contract form.

The scale boundary is just as important. Public routing records show one visible IPv4 prefix, 185.12.224.0/22, equal to 1,024 IPv4 addresses, and no visible IPv6 announced by AS209219 in the checked RIPE routing-status view. The routing-status API reported the IPv4 route visible to all queried RIS full-feed peers in its window and no IPv6 visibility. That is a small footprint, consistent with a local access operator. It does not prove the number of subscribers, homes passed, buildings connected, fibres owned, switches deployed or customers served.

A 1,024-address allocation can sit behind carrier-grade NAT, dynamic assignment, business public-IP sales, internal management use or a mixture of all those. The address count is evidence of operating scale but not a subscriber count.

The financial boundary is also thin but useful. RBC Companies reports 2025 revenue of 43.933 million rubles, cost of sales of 35.019 million rubles and profit or gross profit of 8.914 million rubles, with 12 average employees. It also shows the 2025 opening revenue figure as 57.129 million rubles. Tochka's counterparty page, using earlier public data, reports 2024 revenue of 57.129 million rubles, expenses of 45.368 million rubles and net profit of 11.761 million rubles. OpenWeb reports 2021 income of 37.663 million rubles, expenses of 31.204 million rubles, profit of 6.459 million rubles and 10 average employees.

These figures should not be over-read, but they establish the order of magnitude: Daginfonet is a tens-of-millions-of-rubles local operator, not a hundreds-of-millions regional consolidator.

That scale can be attractive or dangerous depending on density. The attractive version is a compact Derbent ISP with several thousand recurring lines, local brand memory, cash collection points, technicians who know the buildings, and enough business customers to raise average revenue. The dangerous version is a flat or shrinking customer base facing national operator bundles, while every compliance, electricity and equipment cost arrives in hard cash. The public record cannot choose between those versions with precision. It does show that the question is real.

Unit Economics From Public Numbers

Daginfonet does not publish subscriber counts, churn, gross adds, capex, traffic volume, upstream prices or bad debt. That means any unit-economics work has to be framed as a sensitivity, not a hidden estimate. The useful starting point is the 2025 revenue figure. At 43.933 million rubles a year, Daginfonet averaged about 3.661 million rubles of revenue a month. If all of that revenue were household access at 1,000 rubles per month, the revenue base would mathematically equal about 3,661 billed lines. If all of it were 1,800 ruble household access, it would equal about 2,034 lines.

If the blended bill were 1,200 rubles, it would equal about 3,051 lines. None of those figures is the subscriber count, because the company also has business plans, public-IP add-ons, cyberclub plans and IT outsourcing services. But the calculation tells the reader the size of the economic object.

The same sensitivity shows why Daginfonet's business mix matters. A single 3,000 ruble legal-entity line can carry the same monthly revenue as three 1,000 ruble household lines. A 6,000 ruble cyberclub line can carry six. If those customers are support-heavy, the gross revenue advantage can disappear. If they are stable and value a local technician, they can materially improve contribution. The published tariff menu suggests Daginfonet is trying to collect different prices from different use cases rather than treating internet access as one flat commodity.

The public accounting spread is not large enough to tolerate complacency. RBC's 2025 revenue of 43.933 million rubles less cost of sales of 35.019 million rubles gives an 8.914 million ruble gross spread, or roughly 20.3 percent of revenue. That spread is before the reader tries to allocate overhead, working capital, renewal capital and any debt service not visible in the page summary. If one treats the 12-employee figure as current, revenue per employee is roughly 3.66 million rubles per year. If one uses the 11-employee figure seen in another registry snapshot, revenue per employee is roughly 3.99 million rubles.

Those are not weak numbers for a local operator, but they are not fat enough to absorb uncontrolled field labour, frequent truck rolls or unpaid accounts.

The 2025 decline is the caution sign. A drop from 57.129 million rubles to 43.933 million rubles is a decline of about 13.196 million rubles, or 23.1 percent. The public pages do not explain whether that reflects fewer customers, lower non-access revenue, a reporting period effect, a classification change, delayed receipts or another cause. It is still material. A local ISP can survive lower revenue if costs fall faster or if the customer base becomes higher quality. It cannot treat a revenue step-down as noise while simultaneously warning customers about rising telecom, equipment and electricity costs.

Tariff geometry gives the company some defence. The 150 Mbps plan at 1,200 rubles looks designed to lift the bill without making the customer feel stuck on the old 100 Mbps product. The 200 Mbps plan at 1,800 rubles captures heavier willingness to pay. Business and legal-entity tariffs price a given megabit well above residential service, which is rational if these customers value uptime, routing stability, public addressing and support. The separate public-IP charge protects scarce address capacity.

Price segmentation is a serious tool for a small network, because it lets the operator defend margin without applying one blunt increase to all users.

The weakness is that tariff tables do not reveal retained contribution. A 1,000 ruble household can be excellent if it sits in a building with many other Daginfonet lines, pays on time and rarely calls. It can be unattractive if it sits at the edge of the footprint, needs repeated router visits, or churns after a promotional offer from a larger operator. A 3,000 ruble legal-entity customer can be excellent if it signs a stable contract and needs little hand-holding. It can be poor if support expectations are closer to managed service than access service. That is why the judgment remains conditional on density, support discipline and churn.

The practical test is contribution after the second year, not the first bill. A new line can look profitable at connection if the customer pays promptly and the installation uses existing building access. The same line becomes weaker if the router must be replaced, a damaged drop cable requires a second visit, payment discipline slips, or a national bundle forces a retention discount. Daginfonet's price increases therefore have to be read alongside service quality. Higher tariffs help only if the extra rubles are not spent rescuing unhappy subscribers.

In a dense city footprint, every saved truck roll and every reused access path turns into margin. In a thin footprint, the same tariff table can hide a field-support deficit.

Cost, Capital And Risk Transfer

Daginfonet's own notices name the cost stack. The company points to telecom services, equipment, electricity, taxes, related legal expenses and statutory storage obligations. Those categories map directly to a regional ISP's cash burden. Wholesale capacity and upstream connectivity are recurring external inputs. Access switches, optical gear, routers, power systems and customer equipment need replacement. Field staff must connect subscribers, diagnose faults, configure routers and support IPTV. Electricity keeps access nodes, office systems and any storage or network equipment alive.

Legal compliance adds documentation, data-retention and reporting burden.

The 2022 tariff notice is especially useful because it links a price increase to Russian laws requiring operators to create infrastructure for information storage. The exact capex burden is not public, and Daginfonet does not disclose storage vendor, capacity or timing. Still, the notice proves the company viewed compliance storage as a serious cost, not a remote legal abstraction. For a small operator, mandated retention and technical compliance can be more painful than for a national carrier because fixed compliance costs are spread across fewer accounts.

The 2025 and 2026 notices show price-pass-through in action. In 2025 Daginfonet said it had to increase service prices from May. In 2026 it said it would increase speeds and slightly raise prices from April, with automatic subscriber migration. That is risk transfer: cost pressure is pushed into the customer bill, partly softened by higher headline speeds. The customer receives a faster tier; the operator receives a higher monthly fee. The economic question is whether customers view the trade as fair enough to stay.

The contract form adds another risk-transfer detail. The household communications agreement says subscribers acknowledge the service terms and tariff rules, use a login for the personal account and payments, and accept that the operator may unilaterally change the service procedure with advance notice. It also records subscriber consent for operator equipment on building structures and common property. That does not eliminate operational risk, but it gives Daginfonet a contractual frame for building access and tariff administration.

Power is not a theoretical cost in Dagestan. The regional energy ministry announced that from July 2026 Dagestan's guaranteeing electricity supplier functions were moving from Rosseti North Caucasus to Stavropolenergosbyt through the Dagenersgosbyt branch, with legal entities and entrepreneurs required to contract with the new supplier by September. Separate tariff documents cover 2026 electricity prices and transmission tariffs.

For Daginfonet, the issue is less the precise household tariff band and more the operational exposure: network nodes consume electricity continuously, billing has to follow the regional supplier change, and outages or contract friction can show up directly in service quality.

Capital renewal is the hardest public blind spot. Daginfonet's site does not publish the age of its last-mile plant, the mix of Ethernet, fibre and wireless, node counts, spare inventory, capex budget or replacement cycle. Yet the tariff notices name equipment inflation, and the vacancy page asks for a specialist-installer able to connect subscribers, service customers, diagnose faults, configure routers and set up IPTV. That vacancy is a market signal. A local ISP earns revenue through field execution, and field execution consumes labour. If technicians are scarce or wage expectations rise, the monthly connection spread narrows quickly.

Network Evidence And Supplier Dependence

AS209219 gives Daginfonet a visible routing identity, but the identity is small and supplier-dependent. RIPEstat's AS overview says the AS is announced and held by "asdin LLC Daginfonet." The Whois record shows import and export policy with AS12389 and AS47626. RIPEstat's neighbour data, however, observed AS20485 and AS47626. BGP.tools also listed TransTeleCom and Timer as upstreams, while another BGP view showed Timer and MegaFon. IPIP and BigDataCloud preserve the Rostelecom and Timer policy view. CIDR Report saw TransTeleCom as an upstream in its collector context.

The right conclusion is not to choose one page as perfect. The right conclusion is that Daginfonet depends on external Russian wholesale and transit networks, and the public record does not fully settle current commercial supplier terms.

That ambiguity is common in small AS analysis. Whois policy can lag observed routing. Collector vantage points can differ. A relationship can be backup, transit, peering, legacy policy or an observed path through a provider. What matters economically is that Daginfonet is not a self-sufficient backbone. It must buy or otherwise obtain upstream reachability and preserve enough route diversity to keep customer experience acceptable. Two observed neighbours is better than one, but it is not broad interconnection.

The PeeringDB public API returned no network entity for AS209219, which means there was no visible public PeeringDB profile in that query. That does not prove no peering. It does mean Daginfonet is not advertising a rich public peering posture there.

The public IPv4 footprint is coherent. RIPE routing-status showed one IPv4 prefix and 1,024 IPv4 addresses, with first seen information dating back to 2019 and last seen on the checked 2026 window. Prefix-overview associated 185.12.224.0/22 with AS209219 and LLC Daginfonet. IP address databases locate the block around Derbent in Dagestan and show no IPv6 address count for the AS. IPinfo also describes activity with a pronounced day-night rhythm, a signal consistent with an eyeball or consumer-heavy network. That matters because a consumer-heavy network has peak-time contention and support patterns very different from a pure hosting network.

RPKI is a caution item. RIPEstat's RPKI validation endpoint returned unknown status for the AS209219 and 185.12.224.0/22 pair because it found no validating ROAs. That does not mean the route is invalid. It means the public validation check did not find a route origin authorization covering the prefix and origin in that query. In a market where route leaks and filtering discipline matter, lack of a visible ROA is an avoidable operational weakness. It is not a financial thesis by itself, but it is a governance and engineering flag.

The no-IPv6 public footprint is another flag. A small regional ISP can continue serving many Russian households over IPv4 plus NAT, and the separate public-IP tariff suggests IPv4 scarcity is being monetized. But no visible IPv6 means Daginfonet has less room to relieve address pressure through modern addressing, and business customers with IPv6 expectations may view national or more advanced operators as safer. In the short run, IPv6 absence may not drive churn in Derbent households. Over a renewal cycle, it is a sign of a conservative network posture.

The network evidence therefore supports a measured view. Daginfonet has real routing assets, not just a storefront. It has enough public visibility to be reachable and observable. It does not have the route diversity, IPv6 posture or disclosed peering depth of a carrier-scale network. Its economics will depend on using that limited network footprint very efficiently within its local access territory.

Customers, Concentration And Local Density

No public source names Daginfonet's major customers. That absence is itself material. If Daginfonet has one or two large anchor accounts, the public record does not disclose them. Spark's summary says the company did not participate in tenders; another registry snapshot says no government contracts. Public pages show no large enterprise logos, public-sector awards or wholesale customer base. The visible business is therefore best treated as a local retail and small-business access base unless better evidence appears.

The customer concentration risk is geographic rather than named-customer concentration. Daginfonet is tied to Derbent. The company site, payment terminals, office address and IP geolocation all point there. Derbent's estimated population is around 129,000 in 2026, according to a public population aggregator that cites Rosstat municipal files and its own projections. A compact city can be enough for a profitable ISP if the operator is dense in the right apartment blocks, commercial streets and small enterprise clusters. But the addressable base is finite.

Once the best buildings are wired, growth depends on upgrades, churn saves, local business share and price.

The payment footprint says Daginfonet understands that local collection is part of the product. The payment page lists the office, Meradom terminals and Sberbank cards or terminals, then enumerates many collection points around Derbent. The 2017 notice about Sberbank payment resumption also tells subscribers to use the correct login. This is not glamorous infrastructure, but it is economically important. Frictionless local payment reduces involuntary churn, preserves cash collection and makes the service feel embedded in ordinary city routines.

The company's support model is local as well. The vacancy page seeks an installer-specialist whose duties include connecting subscribers, servicing customers, diagnosing and fixing faults, setting routers and computers, and setting up IPTV. That tells us where the operating edge sits: a technician's time. A small ISP wins by solving the local fault faster than a national call centre can. It loses if technician load rises, wages rise, or response time deteriorates. In local access, the support unit is both cost and moat.

The service page broadens the offer into IT outsourcing: local network setup, office automation, system administration-style support and cost optimization for businesses without internal specialists. That is useful if real, because it can deepen the business relationship beyond a commodity access line. It is also dangerous if underpriced, because managed IT support can consume open-ended labour. The public page does not show prices or customer examples, so the service is best treated as optional upside rather than proven recurring revenue.

Customer concentration therefore remains unresolved. The most likely base is a spread of households, small businesses, commercial points and possibly cyberclubs. The published tariffs are designed for that mix. The public record does not prove a single large dependency, but it also does not prove enough diversification across buildings and segments. The facts that would matter most are active lines by tariff, top ten customers by revenue, bad debt by channel, churn after each price change, and the share of revenue from business tariffs.

Competition And Substitutes

Daginfonet does not compete in a vacuum. Public tariff aggregators for Derbent show national and alternative offers from Rostelecom, MegaFon and other brands, with advertised plans starting below or around Daginfonet's household prices in some snapshots. Nets showed Derbent offers as low as 375 rubles per month and a fastest plan up to 700 Mbps from Fregat-telecom in its captured page. Tarifnik's 2026 page showed Rostelecom and MegaFon offers beginning at 750 rubles per month and up to 100 Mbps.

Provayder.net lists multiple Derbent providers and tariffs, while local InterCom advertises residential 50 Mbps at 750 rubles and 100 Mbps at 1,000 rubles, with business and budget tariffs resembling the local ISP price ladder.

Aggregator pages are imperfect. They can be promotional, stale, partner-led, address-dependent or incomplete. They are still useful as market signals. They tell us that a Derbent household can see alternatives with lower introductory pricing, bundled TV, mobile minutes or a national brand. If Daginfonet's value is only a headline 100 Mbps price, it is exposed. If its value is local install speed, known support, existing building wiring, cash-payment convenience and business familiarity, it has a defensible niche.

Rostelecom is the most important fixed-line benchmark because it is both a national operator and a public financial comparator. Its 2025 results showed 12.8 million household optical broadband subscribers, 0.9 million B2B/G optical broadband subscribers, and B2C optical broadband ARPU of 426 rubles in the fourth quarter. That ARPU is far below Daginfonet's posted 1,000 ruble entry residential tariff, but the comparison is not simple. Rostelecom reports a national blended ARPU across a huge base, bundles, regions and discount structures. Daginfonet's tariff page is a posted local price, not realized ARPU.

Still, the contrast matters: national carriers can sustain lower blended broadband ARPU because of scale, bundling, cross-sell and capital access. A local operator has to win by local density, not by balance-sheet scale.

MegaFon and mobile substitutes add a different pressure. MegaFon's mobile home internet tariff shows a 350 ruble 30-day fee tied to LTE home internet, with device and tariff conditions. Mobile broadband is not the same product as fixed access for a heavy household, especially when congestion, indoor signal, data policy or router requirements appear. But it is a churn weapon for price-sensitive users or temporary households. Mobile bundles that include home internet, TV and handset allowances can make a standalone local ISP feel expensive even when the fixed line performs better.

The wider Russian market gives mixed signals. A government summary of 2025 telecom research said fixed internet had 38 million subscribers, with private users representing 94.3 percent of subscribers and 72.3 percent of traffic, while corporate subscribers used far more data per account. BusinesStat's summary reported that Russia's fixed wired broadband market grew to 310 billion rubles in 2025, helped by mobile-internet restrictions pushing users toward fixed connections. Those trends can help Daginfonet: if households value fixed lines more because mobile access is unreliable or restricted, local fixed operators gain pricing power.

The same trends also encourage national carriers to push harder into fixed access.

Technology substitution is another time horizon. ComNews reported Rostelecom's plan to move high-load network areas toward up to 10 Gbit/s speeds. Daginfonet's posted residential top tier is 200 Mbps, and cyberclub top tier is 300 Mbps. That does not mean Daginfonet must immediately match 10 Gbit/s marketing. Many households in Derbent will value reliability and price more than extreme speed. But if national fibre upgrades raise customer expectations, Daginfonet's renewal capital requirement rises.

Regulatory And Geopolitical Risk

Telecom regulation is built into the operating margin. Daginfonet's household contract form lists a Roskomnadzor communications licence number dated February 2017. Its site has a licence page, and Russian communications law requires public access to licence registry information. The company's own 2022 notice linked tariff increases to federal storage obligations. A local ISP therefore carries a compliance burden that is both legal and physical: data retention, customer identification, contract administration, licence continuity, personal-data handling and possible enforcement interaction.

Geopolitical risk appears through equipment and routing, not through a single named sanction in the public record. Daginfonet did not publish vendor exposure. But the 2025 and 2026 tariff notices explicitly mention equipment cost. Since 2022, Russian telecom operators have had to manage a harder equipment environment, longer supply chains, substitution risk and currency-sensitive replacement. For a small operator, the challenge is not simply the purchase price of a switch or optical terminal.

It is spare availability, firmware support, technician familiarity and the ability to standardize the access plant enough to keep fault resolution cheap.

Power and regional administration are also part of risk. The 2026 change in Dagestan's guaranteeing electricity supplier required legal entities and entrepreneurs to contract with the new supplier by a deadline, with a warning about off-contract consumption and possible restrictions. An ISP has less tolerance for administrative power disruption than many small businesses. Access nodes, office billing, support phones and customer equipment all depend on stable electricity arrangements. Daginfonet's 2026 tariff notice named electricity as a rising cost, so the power line item is not speculative.

Routing risk is modest but visible. Two observed upstream neighbours reduce single-provider risk compared with one, but the public record shows limited diversity and no PeeringDB profile. If one upstream degrades, pricing changes, filters routes, or suffers a regional issue, Daginfonet has fewer public options than a large carrier. The route has been visible for years, which is positive. Lack of a visible RPKI ROA and lack of IPv6 visibility are not fatal, but they reduce the engineering comfort margin.

Regulatory and geopolitical risk therefore reinforce the pricing thesis. Daginfonet cannot absorb every external cost quietly. It needs the commercial right and customer trust to raise prices, migrate tariffs and charge separately for scarce resources. The company's notices show it has used that right. The financial question is whether customers accepted the increases without a volume loss large enough to explain the 2025 revenue decline.

Unofficial Signals And Contradictions

The most useful unofficial signal is the payment map. A long list of local terminals is a sign of physical embeddedness and cash-collection pragmatism. It does not prove subscriber count. It does show that Daginfonet has organized the business around everyday Derbent payment habits rather than relying only on remote digital billing. For a local ISP, that can reduce lost accounts and late payments.

The second signal is the installer vacancy. A company does not publicly recruit field capacity for a purely dormant network. The role described is exactly the role that determines local ISP quality: connect, diagnose, repair, configure and support IPTV. But a vacancy can mean growth, turnover, labour scarcity or replacement. It is a signal of operating activity, not proof of expansion.

The third signal is consumer rhythm in IPinfo activity. A pronounced day-night pattern is consistent with an eyeball access network. That fits the tariff table and local payment footprint. It also means evening peaks and household support load matter. A hosting-heavy network would have different economics.

The contradictions are worth keeping. Public registry sources differ on details such as employee count, leader-history presentation, and whether a given financial item should be read as profit, gross profit or net profit. Routing sources differ on which upstreams are current or observed. Local aggregator pages differ on provider counts, starting prices, dates and maximum speeds. Those contradictions do not destroy the analysis. They define its uncertainty. The only responsible way to use them is to state the ranges and avoid invented precision.

The largest contradiction is price versus revenue. If Daginfonet raised tariffs in 2025 and again in 2026, why did the 2025 revenue figure fall from the reported opening level? It may not be a true operating decline. It could reflect accounting presentation, non-recurring revenue in the prior year, timing, classification or source limitations. But if it is subscriber loss or reduced activity, the unit economics are under pressure. The company can still be profitable, but it would need to explain how a smaller revenue base will fund equipment renewal and compliance.

Judgment

Elias Ward's judgment is that Daginfonet is a local access-density story with a narrow but real moat. The moat is not technology leadership, public peering depth or national scale. It is local billing convenience, installed access, field familiarity, segment pricing and a visible network identity. That can be enough in Derbent if the company owns the right buildings, keeps support fast, and uses business tariffs and add-ons to lift blended revenue.

The risk is that the moat can decay quietly. National carriers can bundle. Mobile operators can undercut temporary or price-sensitive users. Equipment and electricity costs can rise. Compliance storage can consume capital. A small AS can remain technically visible while the customer base erodes. A 23 percent apparent revenue decline is too large to ignore, even if the cause is unresolved. The company's own price notices prove management knows costs are rising; they do not prove customers will keep absorbing those costs.

The positive case would become stronger with five facts. First, current active lines by tariff showing stable or rising household density after the 2025 and 2026 increases. Second, churn data showing that the automatic migration to higher-speed plans did not trigger meaningful losses. Third, evidence that business and cyberclub tariffs contribute a durable share of revenue without disproportionate support cost. Fourth, a capex schedule showing the access network can be refreshed within retained cash flow. Fifth, clearer routing and RPKI posture, including current upstream contracts, redundancy and IPv6 plan.

The negative case would become stronger with five different facts. First, revenue decline tied to subscriber churn rather than one-off accounting. Second, high bad debt or payment leakage despite the local terminal footprint. Third, heavy reliance on one enterprise or building owner. Fourth, rising field labour costs without improved average revenue per line. Fifth, customer migration to national fibre or mobile bundles at lower all-in prices.

Until those facts appear, the most defensible conclusion is conditional. Daginfonet can keep working as a profitable local ISP if it protects density and passes through cost while customers still value local support. It is fragile if treated as a growth company. The economic incentive is to make every retained connection pay for the next renewal cycle before the renewal bill arrives.

Sources