Summary

  • ActiveNet's most visible access product is a Bratislava House of Technology business internet account priced at EUR 21.12, EUR 38.24 or EUR 55.36 per month for 10/10, 20/20 or 30/30 Mbps service, each with a EUR 20 installation fee and 12-month commitment on the current public tariff.
  • The economic question is not whether those speeds beat national consumer fiber bundles on headline megabits. It is whether a tenant values symmetric upload, local installation, 24/7 support contact, hosted-server cross-sell and a known building infrastructure enough to pay a higher price per advertised megabit.
  • Network evidence identifies ActiveNet as the operator of AS209998 with the RIPE allocation 152.89.16.0/22, but public BGP views show a small, single-upstream footprint. That supports the operating identity, not a claim of high resilience or low latency for every customer.
  • The strongest commercial reading is that ActiveNet is a micro-operator combining access, hosting and local IT labour. Its risk is the same combination: a small team, narrow visible routing base and heavy reliance on the economics of a specific building and nearby customers.

A tenant's choice is not only speed

Imagine a small Bratislava tenant choosing connectivity for an office in the House of Technology. The firm may have five to twenty people, a VoIP phone setup, a cloud accounting system, a shared drive, remote access for an accountant, a reservation or billing application, and perhaps a small on-premise server that is gradually being replaced by hosted services. The procurement question looks simple at first: is the 10/10 Mbps line enough, should the firm buy 20/20 Mbps, or should it pay for 30/30 Mbps?

That is not the same decision a household makes when it compares a 500 Mbps or 1 Gbps residential fiber bundle. The small office is not only buying download capacity for streaming or web browsing. It is buying upload capacity for sending invoices, uploading design files, keeping video calls stable, backing up databases, connecting remote staff, synchronising point-of-sale records and supporting a hosted phone system. In that setting, the "10/10" label matters because the second number is not an afterthought. Symmetry is the paid unit.

ActiveNet's public House of Technology page gives the cleanest view of that unit. The current visible tariff lists Internet 10/10 at EUR 21.12 per month, Internet 20/20 at EUR 38.24 per month and Internet 30/30 at EUR 55.36 per month, with each package carrying a EUR 20 setup fee and 12-month commitment. All prices are listed as including VAT. On that schedule the monthly charge per symmetric megabit declines as the buyer steps up: EUR 2.112 per Mbps at 10/10, EUR 1.912 per Mbps at 20/20 and about EUR 1.845 per Mbps at 30/30. The saving is not dramatic, but it is enough to tell the buyer that the first circuit is carrying a lot of fixed account and support cost, while incremental bandwidth is somewhat cheaper.

That shape is typical of a small local business access product. The operator must answer the phone, maintain a port, install or configure customer equipment, issue invoices, monitor its building infrastructure and keep enough upstream capacity available. The cost of creating the account does not halve just because a customer buys 10 Mbps instead of 20 Mbps. The additional capacity may be cheaper than the relationship around it.

The page also says ActiveNet offers internet in the House of Technology through its own data infrastructure, describes the building as connected by two independent optical routes, and says the optics are terminated in an air-conditioned and monitored server room using Cisco and MikroTik equipment. It further says customers in the House of Technology and nearby buildings are connected from that room, and that optical lines, unlicensed-band radio links and digital circuits can be used for data connections. Those statements place the product at building and near-building scale. They do not prove any individual customer's measured latency, packet loss or repair history, but they explain why a tenant's alternative is not simply "the national internet market." The relevant market begins with what is physically available in that building, on that floor, and at that demarcation point.

That is the narrow but interesting economic unit: a symmetric business internet circuit in one Bratislava building, sold by a very small operator that also runs hosting and IT services.

The company is a small IT operator before it is a carrier

ActiveNet, s.r.o. is not presented in public records as a national telecommunications incumbent. The Slovak accounting register lists ActiveNet, s.r.o., ICO 36773760, at Skultetyho 1, Bratislava-Nove Mesto, with SK NACE 62010, computer programming, a creation date of 5 May 2007 and a size category of two employees. FinStat's company page likewise describes the company as based at Skultetyho 1 and reports 2025 revenue growth to EUR 406,520, profit of EUR 11,103 and a 2026 employee category of two.

Those numbers are important because they keep the article grounded. A company with a few hundred thousand euros of annual revenue and a two-employee category cannot be analysed like Orange Slovensko, Slovak Telekom, SWAN or Slovanet. It can still provide useful connectivity in a building. It may even be more flexible for a specific tenant than a large carrier. But its economics are micro-operator economics: owner knowledge, local support, software work, hosting revenue and the reuse of a small network platform across several services.

The official financial report gives more texture. In the 2025 profit and loss statement, ActiveNet reported EUR 406,210 of sales from its own products and services, EUR 409,075 of operating revenue, EUR 260,182 of services costs, EUR 28,846 of material, energy and other non-stock supply consumption, EUR 67,176 of personnel costs and EUR 11,103 of after-tax profit. The assets side showed EUR 264,783 of total assets and EUR 115,751 of cash and bank balances at year-end 2025. The liabilities side showed EUR 118,686 of equity and EUR 93,310 of long-term bank loans.

The most useful reading is not that any one line reveals the broadband margin. The accounts do not split access, hosting, software, voice and other services. The useful reading is that ActiveNet has enough revenue to be an operating technology business, but not enough scale for the access service to stand apart from the rest of the firm. If a House of Technology customer buys a symmetric line and later adds a server, webhosting, VoIP or support work, that is not an accidental upsell. It is the business model.

ActiveNet's own "about" page reinforces that interpretation. It says the company specialises in websites, domain registration, webhosting and information systems, and that in the House of Technology it provides internet services, internet telephony, virtual servers and dedicated servers through ProfiServer. It also describes software work, database work, testing and administration either on customer servers or in ActiveNet's hosting. That is not a pure access-provider story. It is a local IT services story with a network layer.

Why symmetry changes the comparison

The immediate objection to ActiveNet's tariff is that the speeds look small beside national broadband marketing. Slovanet's current public consumer internet page, for example, lists optical packages such as Optik 100 at EUR 17 per month, Optik 500 at a promotional EUR 10 then EUR 20, and Optik 800 at a promotional EUR 11.50 then EUR 23, with upload speeds of 20, 100 and 160 Mbps respectively. Orange's business internet page says its Internet pre firmy starts from EUR 13 per month and emphasises backup connection, DDoS/security options, account support and selection of the best available access method for small and medium firms. A customer looking only at megabits per euro could conclude the national offers are cheaper.

That conclusion may be right for some customers. It is also incomplete.

First, availability is local. A published national tariff does not guarantee that the exact office in the exact building can order the advertised product on the same terms. Buildings have risers, access agreements, existing cabling, landlord rules, power constraints and installation lead times. ActiveNet's claim is more specific: it is already selling in the House of Technology and nearby buildings. A tenant's switching cost may include not only the monthly price but the risk and time of bringing a different provider through the building.

Second, symmetric upload can matter more than headline download. A 30/30 Mbps circuit is not impressive for streaming video, but it may be more useful than a much faster asymmetric line for a firm that sends files, hosts a private service, runs VPN sessions, backs up local machines, or keeps call quality stable while multiple people upload. The value is not the number 30 by itself. It is the predictability of the upload side relative to the office's actual workload.

Third, a building operator can package support in a way that a mass product cannot. ActiveNet's page advertises 24/7 customer support with a phone number and e-mail contact, while the general terms for the internet service set out written contractual mechanics for fault reporting, complaints, invoices, payment and service limits. The terms say faults are reported by phone or e-mail, that ActiveNet confirms receipt electronically, and that if a fault is attributable to ActiveNet it should begin work no later than ten hours after reporting and remove the fault within twenty-four hours from the start of work if the repair lies within its personnel and technical capacity, unless the customer contract says otherwise.

Those conditions are not a gold-plated service-level agreement. The same terms limit liability, exclude lost profit and consequential damages, require written complaints within a deadline and reserve operational rights for misuse or non-payment. They also provide that ActiveNet may apply a fair-use policy of 1,000 GB per month for downloaded or uploaded internet data, with speed reduction to 128 kbit/s after breach. That fair-use term sits uneasily beside the marketing page's language about data and time being unlimited. A careful buyer should ask whether the FUP is still applied to the specific business tariff, whether a different contractual term supersedes it, and how any guarantee is measured.

But the presence of these terms is still useful. They show ActiveNet is selling a regulated public electronic communications service, not only informal office Wi-Fi. They define the risk allocation. For a small office, that may be enough if the value is rapid local contact and practical knowledge of the building. For a company with critical uptime requirements, it should trigger tougher questions.

The network record says "real but small"

ActiveNet's network evidence is meaningful, but it must be read narrowly. The RIPE aut-num record for AS209998 lists the AS name as "activenet," the organisation as ORG-AS719-RIPE, and routing policy remarks for VNET and GOLEMTECH. The RIPE search record for 152.89.16.0 shows the IPv4 allocation 152.89.16.0 - 152.89.19.255, netname SK-ACTIVENET-20190123, country SK, organisation ORG-AS719-RIPE, status ALLOCATED PA, and a route object for 152.89.16.0/22 originated by AS209998. The route object was created in May 2019.

Public BGP views support the same operating identity. Hurricane Electric's BGP Toolkit page for AS209998 ActiveNet shows one IPv4 prefix originated, zero IPv6 prefixes, 1,024 IPv4 addresses and one observed IPv4 peer, VNET a.s. BGP.tools describes AS209998 as a small RIPE network with one IPv4 prefix, zero IPv6 prefixes, one upstream and the same 152.89.16.0/22 prefix. IPinfo's AS page reports 1,024 IPv4 addresses, zero IPv6 addresses, 67 hosted domains, one upstream and one peer, and lists VNET a.s. as the upstream. RIPEstat's announced-prefixes API for AS209998 reported 152.89.16.0/22 as the visible prefix through the period ending 10 July 2026.

That is strong evidence that ActiveNet has a live public routing footprint. It is also evidence of small scale. A single visible upstream in public BGP data means the route record should not be converted into a claim of diverse global interconnection. PeeringDB adds a useful complication: ActiveNet's PeeringDB profile self-reports an open policy, 10-20 Gbps traffic level, heavy outbound traffic, IPv4 and IPv6 support, and five IPv4 and five IPv6 prefixes, but it shows no public peering exchange-point rows and no interconnection-facility rows to unauthenticated readers. The self-reported prefix count also conflicts with the BGP views that show one IPv4 prefix and zero IPv6 originated.

The article's conclusion should therefore be precise. AS209998 and 152.89.16.0/22 support ActiveNet's identity as the network operator behind the service. They do not by themselves prove that a House of Technology circuit has low latency, high uptime, multi-homed resilience or enough spare capacity under all conditions. Those are contract and measurement questions. The public record is enough to say ActiveNet is not merely reselling an office router under a brochure name. It is not enough to say it is operationally equivalent to a larger multi-homed carrier.

Upstream dependence is the central technical risk

The most visible technical risk in the public data is upstream dependence. Hurricane Electric, BGP.tools, IPinfo and the RIPE routing policy all point to VNET as the clearest upstream for AS209998. The RIPE aut-num policy also includes GOLEMTECH, but the live public views consulted here did not show that as a current observed peer. That distinction matters.

For a small operator, buying transit from a stronger local network is normal. It can be efficient. VNET is a Slovak carrier and data-centre operator with a much larger routing base. ActiveNet does not need to build every layer of the internet itself. It needs enough upstream capacity and commercial terms to support the customers it actually serves. If VNET provides reliable transit, the arrangement can be rational.

But the customer should separate "two optical routes into the building" from "two independent internet upstreams visible in BGP." The House of Technology page says the building is connected by two independent optical routes, which addresses the risk of a physical cable cut near the building. Public BGP views, however, still show a very small upstream picture for the AS. Physical route diversity and autonomous-system transit diversity are different forms of resilience. A buyer that needs high availability should ask for the precise access design, upstream design, failover behaviour, maintenance windows, service credits and any measured history of outages.

The same applies to IPv6. PeeringDB's profile indicates IPv6 support, but Hurricane Electric, BGP.tools and IPinfo report zero IPv6 prefixes originated for AS209998, and RIPEstat's visible announced-prefix data is IPv4-only for the consulted endpoint. If a customer requires IPv6, the question is not whether a website field contains an IPv6 icon. It is whether the ordered service includes routed IPv6, how it is delivered, and whether it is supported by the customer's equipment and contract.

Those caveats do not make ActiveNet unattractive. They define the appropriate buyer. A small local office may care more about getting a human on the phone than about seeing multiple global upstreams. A SaaS company, payment processor, call centre or other high-availability customer may need a larger carrier circuit, a second independent line or a cloud architecture that assumes local access can fail.

Hosting is not an add-on; it is part of the margin story

ActiveNet's connectivity economics become clearer when the hosting portfolio is included. The company's ProfiServer site advertises dedicated servers, virtual servers and server housing. The ProfiServer page lists dedicated-server examples from EUR 60 to EUR 90 on the top cards, then a wider comparison table with systems from small devices and older desktops up to a Dell PowerEdge T430. It describes non-stop server operation, technical support, one public IP address, transferred-data statistics, backed-up power, backup, administration, licensing and monitoring options. It also lists server housing from a base configuration of EUR 18, with VAT shown separately, and a EUR 20 setup fee.

The hardware list is a market signal. Some systems are clearly older. That does not make them useless: small business workloads, legacy applications, test environments, low-traffic websites and local databases can run economically on modest equipment. But it does place ProfiServer closer to pragmatic low-cost hosting and housing than to hyperscale cloud. The customer is not buying an abstract global elastic compute platform. The customer is buying a concrete server or virtual-server relationship with a local operator.

ABChosting adds the lower end of the stack. The ABChosting page advertises domain registration and webhosting packages, including a parking plan at EUR 0.50 per month, Start at EUR 2 per month, Standard at EUR 4 per month and Profi at EUR 10 per month, with prices shown before VAT. Billing.sk shows the software side: its Billing system page presents a pawnshop CRM with hosted and local options, including Basic Billing on shared hosting at EUR 20 per month, Profi Billing on a private virtual server at EUR 40 per month and Local Billing with a one-time licence model. ActiveCall presents the voice layer, with VoIP/PBX features such as extensions, call transfer, groups, voicemail to e-mail, IVR, call recording, online call records and failover routing.

Together, those pages explain why a 10/10 or 30/30 circuit may be more valuable to ActiveNet than its standalone monthly price suggests. The access line is a relationship anchor. It gives the operator a reason to speak with tenants in the building. Once there, ActiveNet can sell hosting, server housing, web work, domain management, voice, security certificates, backups, monitoring or custom software. Conversely, a customer that already buys a server or VoIP service from ActiveNet may prefer to keep access with the same supplier because troubleshooting becomes simpler. If voice, server and access fail together, the customer has one contact. If they fail separately, the customer avoids a three-provider blame chain.

That convenience has a price and a risk. It can reduce coordination cost for a small firm, but it can also concentrate dependency. If one small supplier manages access, hosting, voice and software, the customer should be comfortable with that supplier's capacity, documentation, backups and exit path. A larger organisation may prefer separation: one carrier for access, a different cloud provider for compute, another provider for voice, and internal documentation that allows migration. A small tenant may rationally prefer one local team.

National carriers sell a different bundle

The larger Slovak providers put pressure on ActiveNet in two ways. They offer higher nominal speeds where their networks reach the address, and they surround those speeds with broader service portfolios.

Orange Slovensko's business internet page is especially useful because it speaks directly to the small and medium business customer ActiveNet may meet in the House of Technology. Orange says Internet pre firmy is intended for offices, operations and warehouses, starts from EUR 13 per month, and can include backup internet that automatically starts within seconds, DDoS and online protection options, a fixed IPv4 address and a sales representative who assesses the customer's needs. Orange also distinguishes home internet, Internet pre firmy and fixed dedicated internet for larger businesses. That framing is a direct challenge: Orange is telling small companies they can buy something more robust than home broadband without jumping to a corporate dedicated line.

SWAN, by contrast, presents itself more as a business-telecom and systems partner. Its current home page says it has more than 25 years of experience and the second largest share in the Slovak business telecom market, and offers internet, voice, data, security and cloud under one roof. The service menu includes business internet, managed WiFi, SD-WAN, IP VPN, Ethernet line, Ethernet VPN, cloud services, backup storage, network storage, data centre and DDoS protection. That is a broader enterprise catalogue than ActiveNet's building tariff, although SWAN's prices for specific comparable Bratislava circuits are not visible from the homepage alone.

Deutsche Telekom's Slovakia profile says Slovak Telekom is the leading player in the Slovak telecommunications market, with fixed network services over fiber optic and metal networks for households, and business services including internet access, IoT, cloud, data storage, computer networking, cybersecurity and smart-city solutions. For a Bratislava SME, the Telekom brand has procurement comfort, national coverage and a broad product catalogue. It may also have less appetite for a small bespoke building job unless the address is already well covered.

Slovanet's current public internet page shows the consumer and small-user price pressure from another angle. Its optical offers have much higher nominal speeds than ActiveNet's 10/10 to 30/30 packages, with lower monthly prices on promoted tiers. But Slovanet's page itself says offers and properties depend on technology available in the specific town or address, and that users should verify availability. It also separates ordinary internet choices from more demanding business users, who may choose from business offers. Again, the lesson is not that ActiveNet is universally cheaper or more expensive. The lesson is that public headline speed is the wrong single metric.

ActiveNet's defence against larger carriers is specificity. It has the address, the local platform, the small support relationship and the service adjacency. Its vulnerability is that larger carriers can add backup, security, account management and national infrastructure around a small-business product, while public cloud vendors can remove part of the server need altogether.

Public cloud is a substitute for servers, not for the access line

Public cloud pricing gives another benchmark. AWS Lightsail lists simple Linux bundles starting at USD 5 per month for 0.5 GB memory, 20 GB SSD and 1 TB transfer, USD 7 for 1 GB and 40 GB, and USD 12 for 2 GB and 60 GB when using public IPv4 bundles. DigitalOcean's Droplet page lists basic virtual machines starting at USD 4 per month for 512 MiB memory, one vCPU, 10 GiB SSD and 500 GiB transfer, with larger basic plans at USD 6, USD 12 and above. Those vendors make small servers look cheap and instantly available.

For ActiveNet, this cuts both ways. Public cloud is a clear substitute for a small hosted server, test environment or low-traffic web application. A tenant that used to rent a local virtual server may now run a small application in Frankfurt, Amsterdam or another European region and pay by card. It does not need to visit a data centre. It does not need local hardware. It can scale beyond one operator's rack.

But public cloud does not replace the office access line. The office still needs connectivity to reach the cloud. If anything, cloud adoption can make upload and stable latency more important. When accounting, backups, documents, customer systems and voice move off-premise, the access line becomes the dependency through which everything else flows. A cheap cloud virtual machine is not useful when the office cannot connect to it.

That is why ActiveNet's strategy can still make sense if it leans into the access side and treats hosting as a convenience product for customers that want local support. It should not try to beat hyperscalers on global elasticity or developer ecosystems. Its advantage is practical: a customer can call someone near the building, combine a circuit with a simple server or phone system, and keep the relationship in Slovak business terms. Its disadvantage is also practical: sophisticated customers can compare its server prices and hardware list against global cloud vendors in minutes.

The price is also a labour contract

The most easily missed cost in ActiveNet's tariff is labour. A 10 Mbps circuit does not consume much transit capacity in 2026, but it can consume support time. Someone must handle the order, confirm the address, configure the port, install or test customer premises equipment, explain the invoice, answer outage calls, decide whether a fault is inside the customer's router or ActiveNet's network, and coordinate any building access. For a small operator, one difficult customer can absorb a large part of the gross margin of several small circuits.

This is why the 10/10 package is expensive per megabit. It is not only bandwidth. It is the minimum charge for a business relationship. If the customer stays simple, pays on time and never calls support, the line may be profitable. If the customer needs repeated handholding, the margin can disappear. The tariff's upward slope gives ActiveNet more room on larger accounts while still anchoring the smallest office at a monthly price that can justify some human attention.

The general terms show how ActiveNet tries to control the labour risk. The customer must provide identifying data, use approved equipment, prevent third-party misuse of the service, pay on time, report changes and cooperate with installation and maintenance. ActiveNet can refuse a contract if service at the requested location or scope is technically infeasible, if the customer does not accept the terms, or if the customer gives reason to doubt compliance. It can interrupt service for misuse, non-payment or breach. Late payment can trigger penalties and reminder fees.

This is not unusual. Small carriers and IT providers survive by keeping support obligations bounded. The difference is that the bounds are more visible in a micro-operator. Large providers can absorb more bad accounts across a national base. ActiveNet has less room for unfunded support. That makes customer selection important. The best customers are likely stable tenants with predictable needs, willingness to pay for support, and possible demand for hosting or voice. The least attractive customers are those who buy the cheapest tier, demand enterprise-grade guarantees, or generate many support events without buying adjacent services.

What the financials imply about resilience

ActiveNet's 2025 accounts point to a functioning small company, but not a company with unlimited shock absorption. Revenue rose, profit improved and the balance sheet showed more than EUR 100,000 in cash and bank accounts at year-end. That is positive for a micro-operator. The company was not presented in the public sources consulted as insolvent or inactive.

At the same time, the profit margin was modest after costs, and the company carried long-term bank loans. Personnel costs were also lower than in the previous year, according to the 2025 financial statement, while FinStat's size category remained at two employees for 2026. If those records reflect the operating reality, the business depends heavily on a small number of people and suppliers. That can be a strength when customers need flexible decisions, and a weakness when customers need deep staffing, formal escalation ladders or guaranteed response capacity across holidays, illness and concurrent incidents.

The accounts do not reveal customer concentration. That is one of the most important missing facts. A small provider can look healthy while depending on a few software, hosting or building customers. If the House of Technology access base is small, each tenant matters. If one larger software or hosting customer leaves, the effect may be material. Conversely, if the company has many small recurring accounts across webhosting, domains, software, VoIP and building access, the revenue may be more resilient than the headcount suggests.

The most constructive interpretation is that ActiveNet has built a compact recurring-service platform: business internet in a known location, hosted servers, low-cost webhosting, VoIP, a vertical software product and custom IT work. The risk is not that any one component is implausible. The risk is that the components require different forms of competence and availability from a small team.

Market signals are thin, and that matters

Public social and forum signals around ActiveNet are limited. Search results identify a Facebook presence for activenet.sk with a small visible follower count, old IT yearbook listings, local business-directory references and third-party ASN/domain listings. They do not reveal a broad current review base that would independently validate service quality. That absence should not be overstated: many small Slovak B2B providers operate through referrals and direct relationships rather than public reviews. But it means outside readers should not treat customer satisfaction as publicly proven.

The richer market signals are indirect. IPinfo reports 67 hosted domains across AS209998 and shows several pingable IPs in the 152.89.16.0/22 block. BrowserScan-style domain listings and IPinfo hosted-domain data indicate that the address block is used for web and hosting workloads, including profiserver.sk and a mix of Slovak domains. That supports the idea that ActiveNet's network is not only an access label; it also carries hosted services. Still, hosted-domain counts are not revenue, customer retention or performance proof. They are signs of use.

Another signal is the persistence of ActiveNet's web properties. The main site, House of Technology tariff, ProfiServer, ABChosting, Billing and ActiveCall pages are all live. Some designs and hardware references look old, and parts of the ecosystem read like long-running small-business pages rather than polished national-product sites. For a certain buyer, that may be reassuring: the provider is established and practical. For another, it may raise questions about product refresh, security maintenance and formal service management. The same signal cuts in both directions.

What would change the judgement

Several facts would materially change the assessment of ActiveNet's economics.

The first is actual contracted uptime and repair performance. The public terms define process and some time limits, but they do not prove outcomes. A record of ticket times, outage history, service credits or third-party measurements would make the support claim stronger or weaker.

The second is current upstream diversity. If ActiveNet has an active second upstream, private interconnection or failover path not visible in the public views consulted here, the technical risk would be lower. If all production traffic depends on one upstream path despite two building routes, the risk remains concentrated.

The third is customer mix. A building access base of a few tenants is different from a diversified base of access, hosting, software and voice subscriptions. The accounts show total revenue, not churn, average revenue per customer or recurring share.

The fourth is IPv6 delivery. If ActiveNet can provide IPv6 on request even without visible originated IPv6 in public route views, that should be documented in customer offers. If it cannot, the gap will become more visible over time as business networks, security tools and cloud services assume dual-stack capability.

The fifth is cloud migration pressure. If ActiveNet's hosting customers are mostly simple websites and small applications, public cloud and managed SaaS can keep eroding the server side. If customers value local support, data handling, custom legacy environments and integrated access, hosting can remain sticky despite cheaper virtual machines elsewhere.

The sixth is building occupancy and landlord access. Because the strongest tariff evidence is tied to the House of Technology, any change in tenant mix, building ownership, cabling policy or competitive access could change the addressable market quickly.

The right conclusion is modest but useful

ActiveNet's House of Technology product should not be judged by a residential speed race. A 30/30 Mbps circuit at EUR 55.36 per month looks expensive if the only question is how many download megabits a household can buy elsewhere. It looks more coherent when the question is how a small Bratislava office buys symmetric upload, local installation, a support contact, and the option to keep hosting, phone and server issues with the same supplier.

The company behind the product is small, real and specialised. Public records show a Bratislava IT firm founded in 2007, with 2025 service revenue above EUR 400,000 and a small employee category. RIPE, RIPEstat, BGP.tools, Hurricane Electric and IPinfo identify AS209998 and the 152.89.16.0/22 allocation as ActiveNet's public network footprint. Company pages show a portfolio that combines access, webhosting, dedicated and virtual servers, server housing, VoIP and software.

The cautions are equally clear. Public BGP data points to a small single-upstream footprint, not a highly diversified network. PeeringDB self-reported prefix information conflicts with observed public route data. The terms contain a fair-use clause and liability limits that buyers should understand. Public social proof is thin. Larger Slovak operators can offer broader business connectivity portfolios, and public cloud vendors can undercut parts of the hosting proposition.

That leaves ActiveNet in a defensible but narrow position. It is strongest where the customer values a local building-level operator and a bundled relationship more than maximum headline speed. It is weakest where the customer needs formal multi-carrier resilience, national procurement comfort, hyperscale cloud economics or independently verified performance. For the small firm inside the House of Technology, the decision is therefore not "is ActiveNet faster than everyone else?" It is "is the symmetric megabit, supported locally and tied to nearby hosting and IT labour, worth the monthly account price?"

That is why ActiveNet is worth tracking. It shows a regional ISP economy at the smallest scale: not a network built to dominate a country, but a compact service business that turns one Bratislava building, one routed IPv4 block, one support relationship and a set of hosted services into recurring local infrastructure revenue.