Summary

  • KOYCEGIZ BILISIM HAB.TEK.SAN.TIC.LTD.STI. is not merely a brochure operator. The RIPE record shows AS212549, the company organization entity, a registered Turkish address in Mugla, a single routed IPv4 /24, a route object and RPKI validation for the visible prefix. RIPEstat shows the origin was first seen in October 2020 and still visible on 2026-07-28.
  • The operating problem is scale and repricing, not existence. The public network footprint is narrow, the address space is marked for dynamic CGNAT customers, no IPv6 visibility appears in the sampled RIPEstat routing view, and the retail offer must compete with national fixed broadband and mobile fixed-wireless substitutes while paying upstream, support, site and imported hardware costs.
  • The strongest economic signal is the company’s own tariff ladder: local fibre offers at 750, 850 and 950 Turkish lira per month, wireless offers at 950, 1,250 and 1,500 lira per month, and installation charges that rise sharply for wireless. That schedule suggests the company knows it must recover more from customers who need local access work rather than simple resale.
  • The judgment would change if the company published subscriber counts by access type, churn, utilization, wholesale cost pass-through, backhaul terms, equipment replacement cycles, IPv6 deployment, support staffing, outage records and the share of revenue from business or seasonal accounts. Without those numbers, the defensible conclusion is that Koycegiz Bilisim is an operating local access provider whose resilience depends on disciplined contribution per subscriber.

Start with one bill in lira

The useful way to read Koycegiz Bilisim is to begin with a single local household bill, not with the legal name, the logo or the autonomous-system number. A 750 lira monthly fibre customer, an 850 lira fibre customer and a 950 lira fibre customer do not all create the same margin simply because they sit on the same tariff page. The first may require less active capacity if usage is light, the second may push more traffic into upstream and access electronics, and the third may be bought by a household that expects the service to behave like a national operator. A wireless customer is more demanding still.

The advertised wireless prices begin at 950 lira per month and rise to 1,500 lira, with a 10,000 lira installation fee. That fee is not just a retail flourish. It is a sign that the customer path may involve equipment, labour, site reach, alignment and support that cannot be hidden inside a low monthly charge.

The company’s public pages make the tension visible. It markets high-speed broadband, no quota, no commitment, wide coverage, specialist support, e-Devlet onboarding, infrastructure queries and service in fibre, VDSL, metro and Airfiber-style access. The same pages also show a regional identity. The contact address is in Gulpinar, Fevzi Pasa Caddesi, Koycegiz, Mugla. The FAQ says a customer in Mugla’s Koycegiz district does not necessarily need existing infrastructure because the company can deliver internet wirelessly. The coverage page is not a national fibre-construction disclosure. It is a local-access promise.

That distinction matters because a local-access provider earns its relevance where national plant, apartment cabling, customer urgency or geography leaves space for a smaller operator to do practical work.

The public record supports that the company operates a real network. RIPE identifies the organization as KOYCEGIZ BILISIM HAB.TEK.SAN.TIC.LTD.STI., country Turkey, with a RIPE local internet registry type and registry number KÖYCEGIZ/1277. RIPEstat’s AS overview names the holder as the same company. The AS routing record shows AS212549 announced, with the visible prefix 185.203.89.0/24. RIPEstat’s routing status says the first seen origin was 2020-10-13 and the last seen record in the collected view was 2026-07-28. That is an operating history, not a parked shell.

But a real routed network does not prove a durable retail margin. RIPEstat shows one IPv4 prefix and 256 announced IPv4 addresses. The RIPE inetnum record says dynamic addresses for CGNAT customers are assigned from this range. That is a crucial economic clue. Carrier-grade NAT lets a provider serve more customers than its public IPv4 count would otherwise permit, but it also creates operational burdens: logging, troubleshooting, abuse handling, port-sensitive applications, support calls and eventual IPv6 pressure. A small provider can run CGNAT well, but it must price and staff for the complexity.

The published tariff table does not reveal whether the monthly bill carries that work.

The first economic question is therefore not whether a customer can see an attractive price. It is whether each customer adds enough contribution after value-added tax, wholesale inputs, upstream transit, access electronics, customer equipment, labour, payment friction, field visits, support calls, power, site rent, replacement parts and churn. A 750 lira monthly retail price can look strong beside a household income constraint and weak beside a replacement-radio invoice. A 10,000 lira installation charge can look high to a rural customer and prudent to an operator that must avoid burying capital in low-density links.

The analysis turns on utilization, not marketing.

What the public record can and cannot prove

Koycegiz Bilisim’s first-party pages give the commercial surface. The home page describes high-speed broadband under the KoycegizNet brand, unlimited internet and economic tariffs. The about page says the company serves individual and business customers under that brand using fibre, VDSL2, metro and Airfiber technologies, and says it has broadband agreements with Türk Telekom’s wholesale function and Turkcell Superonline. The tariff page shows the live retail ladder. The contact page gives a local office, email, call-centre phone number and working hours.

The e-Devlet post describes digital subscription approval, suggesting an onboarding process designed to reduce manual friction. These are all useful facts, but they are not audited financial statements.

The routing record gives a different layer. The aut-num entity for AS212549 lists imports and exports with AS12735 and AS204457. RIPEstat’s observed neighbour data on 2026-07-28 shows one observed neighbour, AS204457, with no IPv6 peers in that view. RIPEstat’s routing consistency page says AS204457 is both in BGP and in the RIPE whois import/export data, while AS12735 remains in whois but is not observed in BGP in that snapshot. This does not prove a broken contract or a lost supplier; it proves the public routing table and the registry entity do not show the same active picture at that moment.

For a small provider, that gap is worth watching because upstream diversity and active failover can be more important than the number of names written into an entity.

The immediate observed upstream, AS204457, is Atlantis Telekomunikasyon Bilisim Hizmetleri Sanayi ve Ticaret Ltd. Sti. according to RIPEstat. PeeringDB lists AtlantisNet as a regional Cable/DSL/ISP network with several exchange and facility presences and a reported traffic band far larger than Koycegiz Bilisim’s visible footprint. That makes Atlantis a plausible upstream or regional transport partner. The dormant registry relationship with AS12735 points toward TurkNet, a national network with a large PeeringDB footprint.

The public evidence therefore suggests Koycegiz Bilisim sits behind larger Turkish networks rather than operating an extensively meshed independent backbone.

That posture can be rational. A local access provider does not need to become a national backbone to create value in Koycegiz. It needs reliable local installation, support, suitable backhaul and enough upstream resilience to meet the expectations it sells. The risk is concentration. If one observed upstream carries the visible route and the company has one visible IPv4 /24, service economics become sensitive to the price and reliability of a small number of inputs.

A local outage, supplier dispute, route-policy error or equipment failure can compress the margin quickly because the operator cannot average the event across a national subscriber base.

The public pages also create a geography question. The about page speaks in wider terms, including service through Türk Telekom infrastructure in many provinces and districts or by wireless without telecommunications infrastructure. The FAQ, contact details, domain branding and tariff naming remain strongly tied to Koycegiz. Both can be true: a provider can resell or arrange wholesale-based access in more than one area while its own operating edge is concentrated locally.

The public evidence does not establish a national customer base, and it would be unsafe to treat the company as a national retail ISP simply because the about page uses broad language.

The strongest conservative reading is that Koycegiz Bilisim has three economic surfaces. The first is local fibre or local last-mile access under its own brand, where installation, support and field work matter. The second is wireless access, where the higher installation charge signals customer-premises equipment and alignment costs. The third is wholesale-dependent VAE or VDSL/fibre service, where Türk Telekom infrastructure and regulated wholesale arrangements shape the input cost. These surfaces have different risk profiles. A single blended subscriber number would not be enough; the mix matters.

The regulatory context also matters. BTK’s page on internet end-user tariffs says retail internet access is provided by ISPs authorized by the regulator and that, within the legal framework described there, Türk Telekom does not provide end users with direct internet access in the same way. BTK also explains that DSL services use the local access part of Türk Telekom’s fixed network and that ISPs buy wholesale services from Türk Telekom, while Türk Telekom’s wholesale internet access tariffs for ISPs are subject to BTK approval.

BTK adds that retail fees charged by ISPs to end users for internet access are not subject to BTK approval under the cited framework. In other words, Koycegiz Bilisim faces a regulated wholesale world on one side and a retail repricing problem on the other.

That is the core control boundary. Koycegiz Bilisim can choose retail positioning, support intensity, installation discipline, customer screening, upstream mix, replacement timing and the amount of margin it tries to preserve. It cannot fully control inflation, exchange rates, national wholesale reference offers, the availability of Türk Telekom local access at a given address, mobile operator substitute pricing, or the dollar or euro cost of imported access hardware. The company’s margin survives if the controllable decisions offset the uncontrollable inputs.

The tariff ladder is the economic map

The tariff ladder is more informative than the slogans. The local fibre pages list 50 Mbps at 750 lira per month, 100 Mbps at 850 lira and 200 Mbps at 950 lira, each with no commitment, no quota and a 2,000 lira installation fee. The step from 50 Mbps to 100 Mbps adds 100 lira. The step from 100 Mbps to 200 Mbps adds another 100 lira. That small gap can be a deliberate competitive move, but it leaves little visible room for usage-sensitive costs if customers on higher tiers use the service heavily.

In fixed networks, the expensive question is often not the nominal line speed but busy-hour contention, backhaul headroom, customer complaints and the capital cycle triggered when real usage grows.

The wireless ladder is steeper. The 16 Mbps wireless package is 950 lira per month, the 24 Mbps package is 1,250 lira and the 35 Mbps package is 1,500 lira. Each advertises no commitment and no quota, with a 10,000 lira installation fee. The price progression makes sense if the operator is rationing scarce wireless capacity, recovering higher field costs or serving customers who lack attractive wired alternatives.

A 35 Mbps wireless user may not consume more total data than a 200 Mbps fibre user, but the operator must protect the radio sector, maintain line of sight, handle weather or interference complaints, support customer equipment and possibly revisit sites. The public prices imply that wireless is not being sold as a cheap substitute for fibre; it is being sold as a more labour- and equipment-heavy product.

The old or wholesale-facing pages complicate the picture. VAE VDSL and VAE fibre pages show lower monthly numbers, including 300 lira for 50 Mbps and 400 lira for 100 Mbps on pages modified in 2024, while the live tariff page and 2026-modified fibre pages show higher current local fibre prices. The page text also contains inconsistencies, such as a title for a 50 Mbps VAE package while body language refers to 100 Mbps in places. That is not a basis to accuse the company of mispricing; it is a signal that the tariff archive needs careful reading.

For economic analysis, the recent tariff page and the pages modified in June 2026 carry more weight than older VAE pages whose text appears mechanically duplicated or stale.

Annualizing the visible current local prices clarifies the stakes. A 750 lira customer generates 9,000 lira of gross retail billings over twelve months before tax, processing, bad debt and service costs. An 850 lira customer generates 10,200 lira. A 950 lira fibre or 16 Mbps wireless customer generates 11,400 lira. A 1,500 lira wireless customer generates 18,000 lira. Those are not margins. They are top-line envelopes. A single replacement of customer equipment, one truck roll, a difficult installation or a heavy support case can consume a material part of the first year’s contribution if the installation charge does not cover the field cost.

The installation charges reveal how the company tries to protect that envelope. The 2,000 lira local fibre installation charge is meaningful but not prohibitive relative to the monthly price. The 10,000 lira wireless installation charge is a much stronger filter. It pushes more cost recovery to the moment of connection and discourages churn by making the initial commitment economic even when the product is advertised as no-commitment.

That structure is more defensible than hiding all installation cost in monthly service, especially if the company must buy imported radios, routers, mounts or optical equipment before it has earned any monthly revenue from the customer.

The danger is price anchoring by national competitors. A household does not compare only the technical burden on the provider. It compares a visible monthly number, advertised speed, installation charge and reputation. Türk Telekom’s customer acquisition pages, Turkcell’s Superbox pages and Vodafone’s home-internet pages show that national fixed and mobile-based substitutes are always part of the mental price set, even when availability differs by address. TurkNet’s network and public profile add another type of substitute where wholesale or own-network access exists.

Koycegiz Bilisim’s local advantage must therefore be something other than simply being cheaper. It must be faster installation, better support, availability where others are weak, or more accountable local service.

That makes support labour a revenue problem, not a customer-service footnote. The company advertises expert call-centre support and office contact. Local support can be an advantage because the operator understands addresses, terrain, building quirks and subscriber expectations. It is also a cost centre. If the company wins customers because it answers the phone and sends technicians, the monthly bill must finance that responsiveness. A national competitor can amortize call-centre systems and field processes across millions of lines. A local provider must make each service promise pay locally.

The no-commitment message has two sides. It lowers buyer anxiety and can help a local ISP win customers frustrated by long contracts. It also reduces guaranteed revenue duration. If customers can leave quickly after an expensive installation, the operator must either charge enough upfront, keep churn low through service quality or avoid installations whose payback depends on a long tenure. The wireless installation fee points to that logic. The fibre installation fee is less protective, so the operator must rely more on low churn, efficient installation and the ability to upsell or retain.

The tariff ladder therefore maps the business model. Low-to-mid local fibre prices seek relevance against national alternatives. Higher wireless prices protect scarce access where infrastructure is missing or difficult. The e-Devlet onboarding page reduces friction and may lower administrative cost. The infrastructure-query page channels demand before field work. The contract and privacy pages support formal consumer processes. None of this proves profitability, but it shows that the company’s public surface is built around the practical economics of adding and retaining local subscribers rather than a pure hosting or transit business.

Imported equipment meets lira revenue

The hardest cost mismatch is that local broadband revenue is collected in lira while many network inputs are priced directly or indirectly in foreign currency. Wireless radios, routers, switches, optical modules, power equipment and spare parts often follow dollar or euro supply chains even when purchased through Turkish distributors. MikroTik’s wireless product catalogue is an example of the type of equipment ecosystem a local WISP or access operator may depend on; the exact vendor mix for Koycegiz Bilisim is not public. The point is not to assign a specific bill of materials.

It is that access hardware is not insulated from exchange-rate movement merely because customers pay in lira.

The Turkish central bank’s exchange-rate pages and TurkStat’s inflation portal are relevant for that reason. A provider with imported equipment exposure faces two different clocks. Subscriber prices are changed deliberately and publicly. Hardware replacement costs, licence fees, tools, fuel, rent and wages move whenever suppliers reprice or local inflation feeds through. If the company waits too long to reprice, margins compress. If it reprices too aggressively, customers compare national offers, mobile substitutes or delayed upgrades.

Local broadband is therefore a repeated timing problem: raise prices late and absorb cost; raise prices early and test loyalty.

Wireless access magnifies the timing problem because the customer connection may require more bespoke work. A fibre resale or fibre-last-mile customer may need a router, optical equipment and installation labour. A wireless customer may need outdoor equipment, mounting, aiming, power, cable, weatherproofing and future troubleshooting. The tariff schedule’s 10,000 lira installation fee is a public clue that management recognizes this difference. A lower upfront price could have grown gross subscriber count while leaving the company exposed to unrecovered equipment capital.

The higher upfront charge protects the balance sheet but narrows the addressable market to customers who value the service enough to pay.

That creates an economic sorting mechanism. Customers with good fixed alternatives may reject the wireless installation charge. Customers with weak wired infrastructure, urgent business needs, rural locations, seasonal properties or dissatisfaction with national operators may accept it. Those customers can be valuable, but they often require individualized support. A rural or edge-of-coverage account may generate high loyalty and high operating burden at the same time.

The profitable version of that business is selective: connect customers where radio quality, installation feasibility and support cost are understood before the operator commits.

The company’s public pages do not disclose its site count, radio sectors, fibre route lengths, backhaul capacity, customer-premises equipment inventory or replacement cycle. That absence is normal for a small private ISP, but it limits the economic conclusion. If most customers ride wholesale fibre or VDSL with low field cost, the current tariff ladder may generate enough contribution even at modest scale. If a large share requires wireless field work and repeated truck rolls, the company needs higher average revenue per user, lower churn and disciplined installation screening.

The same advertised price can be healthy or fragile depending on the access mix.

Imported equipment also affects renewal, not only growth. A local ISP can look profitable while it delays replacement of radios, switches, batteries or customer routers. That creates hidden technical debt. The public routing record will not show whether access equipment is aging. Customers may not notice until weather, congestion or failures accumulate. A conservative economic view therefore asks whether the monthly tariff covers depreciation and renewal, not just today’s bandwidth and labour. A 950 lira monthly fibre bill that covers current operations but not future replacement is not durable margin.

The visible IPv4 position adds another form of scarcity. A single /24 can support far more than 256 subscribers only through CGNAT or private addressing behind the public pool. RIPE’s remark that dynamic addresses for CGNAT customers are assigned from the range makes this explicit. CGNAT avoids buying or leasing scarce public IPv4 at scale, but it can increase troubleshooting complexity. Some applications, games, remote access systems, cameras and business workflows behave poorly behind shared addresses. If customers demand public addresses or better inbound reachability, the operator may face additional cost or support work.

IPv6 could reduce some pressure, yet RIPEstat saw no IPv6 announced space for AS212549 in the sampled view.

The absence of visible IPv6 is not a moral failure; many small ISPs delay IPv6 because immediate retail demand is weak. Economically, though, it is a deferred decision. If IPv4 costs rise, CGNAT complaints grow or upstream partners push modern routing expectations, the company may need to invest in IPv6 planning, customer-router support, help-desk scripts and monitoring. Those costs arrive before most subscribers agree to pay extra for IPv6. A local operator must decide whether to invest early for resilience or late under pressure.

Wholesale access and upstream dependence

The about page’s reference to broadband agreements with Türk Telekom’s wholesale function and Turkcell Superonline places Koycegiz Bilisim in a familiar Turkish access model. BTK explains that DSL local access uses Türk Telekom’s fixed network and that ISPs buy wholesale services to reach their users. BTK also says Türk Telekom’s wholesale internet access tariffs for ISPs are subject to regulatory approval, while the retail prices charged by ISPs are not individually approved in the same way. The operator therefore lives between regulated inputs and competitive retail outputs.

That structure can be attractive to a small ISP because it avoids building every meter of access network. It can also compress differentiation. If the underlying line, wholesale fault process or address availability is controlled by a larger infrastructure owner, the smaller ISP wins only through retail packaging, support, local accountability and perhaps speed of customer handling. If wholesale prices or rules change, Koycegiz Bilisim may have limited ability to absorb the difference. If the retail market resists repricing, margin shrinks.

The network-layer evidence adds another dependency. RIPEstat’s AS neighbour view saw only AS204457 adjacent to AS212549 on 2026-07-28. The aut-num entity also lists AS12735, but RIPEstat’s consistency view did not observe it in BGP. The company may still have commercial arrangements, backup paths or dormant configurations not visible in that snapshot. The public conclusion must be narrower: the visible routing footprint at that time looked concentrated. Concentration matters because a small local provider has less bargaining power with upstreams than a large national network and fewer redundant paths to absorb a provider-side disruption.

PeeringDB’s AtlantisNet profile helps explain why a local ISP might use that upstream. Atlantis reports regional scope, multiple exchange and facility presences and a traffic band much larger than AS212549’s visible route. A regional upstream can aggregate demand, handle interconnection and provide transit economics that would be unrealistic for a small operator to build alone. The trade-off is that the local ISP’s retail promise depends on upstream quality, support and price. If upstream costs rise in lira terms or if the operator needs a second active path, the subscriber bill must pay for it.

The public record does not disclose backhaul pricing, commit levels, 95th percentile usage, burst policy, installation cross-connect cost, supplier contract duration or service-level credits. Those are exactly the numbers that decide whether a 100 Mbps retail user is profitable. A network can announce only one /24 and still serve a profitable niche if contention is managed and upstream costs are predictable. It can also lose money on attractive tariffs if busy-hour demand forces capacity upgrades faster than revenue grows. Without utilization data, the safest conclusion is conditional.

The conditional test is straightforward. First, does average revenue per active subscriber rise when equipment, wholesale and upstream costs rise? Second, do higher-speed customers generate enough extra contribution to cover the capacity they actually consume? Third, does the installation fee recover customer-specific capital within a period shorter than expected churn? Fourth, is upstream diversity real enough to prevent a single supplier event from causing customer loss? Fifth, are support contacts falling per subscriber as the company learns its local footprint, or rising as the network ages?

The answers are not public, but they are the correct management dashboard.

Competition and substitutes

Koycegiz Bilisim competes in more than one market at once. For a customer with good fixed infrastructure, the company competes against national fixed broadband offers and against the customer’s default expectation that a large brand should be cheaper or safer. For a customer outside strong fixed coverage, it competes against mobile fixed-wireless products such as Turkcell Superbox or Vodafone home-internet alternatives, against waiting for infrastructure, against tethering and against doing without a high-capacity fixed line. For small businesses, it competes against the cost of downtime and the value of local support.

That mixed competitive set is why the local subscriber bill cannot be read as expensive or cheap in isolation. A 950 lira 200 Mbps fibre offer may look different from a 950 lira 16 Mbps wireless offer, even though the monthly number is the same. The fibre customer is buying speed where plant exists. The wireless customer is buying availability where fixed options may be weak or absent. The installation charges reinforce the distinction. In a dense apartment with national fibre, Koycegiz Bilisim needs price, service or trust.

In a harder-to-serve location, it needs technical competence and enough upfront recovery to avoid subsidizing every difficult customer.

Mobile substitutes are both threat and price support. A mobile fixed-wireless product can be easy to buy and backed by a national operator, but it may carry quota, radio-capacity, indoor-reception or address-eligibility constraints. A local wireless ISP can sometimes provide better installed performance because it controls the outdoor link and knows the area. It can also be less convenient, more installation-heavy and more dependent on local field work. The rational customer compares total cost, reliability and support, not just nominal speed.

The company’s own content leans into that opening. The wireless blog post presents wireless internet as easier to install, wider in coverage and lower in cost than traditional connections, and emphasizes that ports may not be necessary. The FAQ says infrastructure is not required for customers in Mugla’s Koycegiz district because service can be delivered wirelessly. The infrastructure-query page still asks the customer to provide accurate address information so suitable options can be presented. That is a sensible funnel: sell the idea that access is possible, then check the address before promising the method.

Customer concentration is the missing economic risk. A local provider can depend heavily on a small town, a few neighbourhoods, a few business accounts, seasonal demand or a few radio sectors. The public record does not disclose the number of subscribers, the share of revenue from business customers, the share from seasonal properties, or the concentration of customers behind specific access nodes. That means the company’s resilience cannot be assessed from tariff data alone. A small number of high-paying wireless or business customers could make the company stronger than the visible /24 suggests.

A broad base of low-margin residential accounts with high support demand could make it weaker.

Seasonality is also plausible but unproven. Koycegiz and Mugla sit in a region where tourism, seasonal residence and local business cycles may affect demand for connectivity. The evidence here does not prove a seasonal subscriber base, so it should not be overstated. The economic implication is only that management should distinguish permanent residential lines, small business lines, seasonal lines and temporary high-support installations. Each has a different churn and support profile.

Local trust can offset brand scale. A customer who can visit an office, call a local number or get a technician who knows the district may prefer a regional operator even if a national brand has a larger marketing budget. But trust must be converted into renewal economics. If local responsiveness produces many unpaid visits, informal favours or custom fixes, the operator may build goodwill while eroding margin. The healthier version is standardized local service: clear installation rules, clear support boundaries, fast triage and pricing that reflects the real cost of difficult access.

Regulation, consumer process and public risk

Koycegiz Bilisim’s pages show attention to formal consumer process: privacy and security text, distance-sales terms, return and exchange language, subscriber contracts and forms, e-Devlet digital approval and published contact routes. These documents do not prove service quality, but they reduce one kind of operating risk. A local ISP that sells no-commitment service still needs clean consent, data handling, cancellation, installation, consumer communication and dispute processes. Weak process can turn a technical service issue into regulatory or reputational cost.

BTK’s authorization regime page matters because it frames the sector as one where companies gain operator status and become subject to rights and obligations under communications law. KoycegizNet’s FAQ says its wireless service is licensed by BTK. The public article should treat that as a company claim unless the exact operator row is independently retrieved from BTK’s current authorization search. The more important economic point is that regulatory compliance is not free. Notifications, subscriber records, lawful process, abuse handling, data protection and consumer complaints create overhead that must be funded by the subscriber base.

The CGNAT note in RIPE’s IP record intersects with that compliance burden. Dynamic CGNAT addressing can require more careful logs and customer mapping when abuse complaints or lawful requests arise. It can also create user complaints when services expect inbound reachability. These are operational realities for many access providers, not unique defects. They are still costs. A national ISP can spread CGNAT platforms, logging systems and abuse teams across a large base. A small provider must make sure the same obligations do not consume too much of each subscriber’s margin.

The RPKI validation result is a positive technical-control signal. RIPEstat reports a valid route origin authorization for AS212549 and 185.203.89.0/24 with max length 24. That means the visible route is not merely present; it has modern routing-authentication support. For a local ISP, this is valuable because it reduces the risk that its only visible IPv4 route is treated as suspicious by networks that validate RPKI. It does not solve upstream concentration, but it shows the operator or its maintainers have handled a nontrivial routing hygiene task.

The lack of a PeeringDB record for AS212549 is not a failure by itself. Many local access networks do not maintain a public interconnection profile. It does limit outside visibility. Without a PeeringDB profile, a reader must rely more on RIPE, RIPEstat and indirect upstream records to understand interconnection posture. If Koycegiz Bilisim wants to signal maturity to partners, a maintained public profile, accurate AS-set information and transparent NOC/abuse contacts could help. If it remains a purely local retail operator, the absence may be commercially irrelevant.

The regulatory and operational risk therefore sits in the middle ground. The company is not invisible: RIPE, RPKI, public tariffs, public contracts and official Turkish regulatory context all exist. But it is also not transparent in the way a listed telecom company or national operator is. No public financials, subscriber counts, outage reports, churn data or wholesale pass-through disclosures appear in the reviewed source set. That limits the strength of any investment-style conclusion.

What would change the judgment

The judgment would become more positive if the company showed that higher-speed and wireless customers carry materially higher contribution after support and equipment cost. The current tariff ladder suggests that wireless is priced more defensibly than a simple flat offer, but the gap between fibre speeds is narrow. If usage data showed that 200 Mbps customers are not meaningfully more expensive than 50 Mbps customers because the network is lightly utilized or because upgrades are mostly marketing, the small price gap would be rational. If 200 Mbps customers drive busy-hour upgrades, the gap may be too thin.

The judgment would also improve if upstream diversity were active rather than merely listed. The aut-num entity lists two external networks, while RIPEstat observed one. A demonstrated active backup path, tested failover, or a second visible neighbour would reduce operating fragility. So would IPv6 deployment. IPv6 is not a revenue product for most households, but it can reduce future dependence on scarce IPv4, improve technical maturity and lower long-run CGNAT pressure.

Subscriber mix is the largest unknown. A few hundred loyal local accounts, a stable business base and disciplined wireless installation could support a small ISP with one visible /24. A larger but low-margin base with heavy support calls could struggle. The public routing table does not tell the difference. Nor does the number of tariffs. Management’s internal truth is probably found in churn by product, gross margin by access method, support minutes per subscriber, truck rolls per installation, busy-hour utilization and cash payback period.

A stronger public record would separate three claims that are currently blended by marketing language: where the company can sell wholesale-based service, where it controls or materially manages local fibre or wireless access, and where it has actual active coverage. A clear map of service types, not just coverage aspiration, would help customers and analysts. It would also reduce the risk that broad claims about countrywide service are read as proof of active national scale.

The most important evidence would be renewal economics. If customers stay for multiple years, installation cost can be amortized and local support can become efficient. If customers churn quickly, no-commitment positioning becomes expensive. A local ISP can survive currency pressure if it reprices, selects installations carefully and retains customers long enough for the capital cycle to work. It cannot survive by adding unprofitable lines just to look larger.

The defensible conclusion

Koycegiz Bilisim’s public record is stronger than a generic small-company profile but weaker than a full economic proof. The company has a real AS, a visible prefix, a valid RPKI state, RIPE organization data, CGNAT-marked dynamic customer address space, public tariffs, local contact details, subscriber process pages and a service model that visibly combines local access, wireless, wholesale and support. Those facts support treating it as an operating regional ISP.

The economic risk is not that the company lacks a website or a route. The risk is that every part of its model pulls in a different direction. Local customers pay in lira and compare against national offers. Imported equipment and replacement parts follow harder-currency supply chains. Wireless access solves infrastructure gaps but requires field capital. Wholesale access lowers build cost but creates dependence on regulated and supplier-controlled inputs. CGNAT stretches IPv4 but increases support and compliance complexity. Local support differentiates the brand but consumes labour.

That combination can work if management is disciplined. The company needs to know which customers are profitable, not merely which customers can be connected. It needs to keep installation charges high enough to avoid subsidizing hard access, while keeping monthly prices acceptable against national substitutes. It needs to maintain upstream resilience without buying more backbone complexity than its base can pay for. It needs to decide when IPv6 and renewal investment are cheaper than continued CGNAT and aging equipment workarounds.

Above all, it needs to raise contribution per subscriber faster than the cost of equipment, backhaul, support and wholesale inputs.

For customers, the practical question is whether KoycegizNet solves an address-specific access problem better than national alternatives. For suppliers, the question is whether the company can pay reliably through lira volatility. For regulators, the question is whether small ISPs preserve competition and reach without accumulating hidden service-quality or compliance risks. For the company, the question is narrower and harder: can each new subscriber pay back the installation, consume capacity within plan, require support within budget and stay long enough for renewal economics to work?

The answer is not visible from public sources. What is visible is the shape of the test. Koycegiz Bilisim has the ingredients of a defensible local access business, but its margin is likely determined at the edge: one installation, one support call, one upstream invoice, one imported device and one monthly lira bill at a time.

Sources