Summary
- MapMakers Group Ltd is best understood as the English-facing identity of a Moscow weather-software, data and web-portal operator, not as a conventional retail ISP. Matching address, tax number and contact details tie the RIPE NCC member to Scientific and Production Centre Map Maker LLC, the operator behind GIS Meteo and Gismeteo.
- The company controls a /22 of IPv4 space, a /32 of IPv6 space and AS203444. That is meaningful evidence of network capability, but not proof of a telecom product. RIPE observations showed no globally visible routes from the autonomous system on 10 July 2026 and no observed IPv6 announcement.
- The commercial model is broader than connectivity: professional software licences, implementation and support, weather-data subscriptions priced by request volume and service level, consulting, government and enterprise contracts, and the audience economics of Gismeteo. Reliability can support all of those revenues, but the company publishes no tariff card or service-level performance record.
- Filed-account aggregations show 2025 revenue of RUB 562.79 million, down 20.5% from RUB 707.55 million in 2024. Net profit fell 97.1%, from RUB 65.52 million to RUB 1.88 million, while headcount rose from 69 to 70. Revenue growth had already been volatile, and the 2025 figures expose substantial operating leverage.
- The gross margin remained 65.2% in 2025, but administrative costs absorbed RUB 333.53 million of RUB 367.05 million in gross profit. The problem was therefore not a business with no gross value; it was an overhead base and revenue mix that left almost no return for shareholders when sales weakened.
- Historic routing policy named two upstreams, while current public web endpoints sit in networks operated by other providers. Independence is therefore conditional: address rights and routing expertise increase bargaining power and portability, but power, facilities, transit, equipment, domain infrastructure and external data remain supplier dependencies.
- Gismeteo gives the company unusual consumer reach: Google Play showed more than five million downloads, while Similarweb estimated gismeteo.ru as the fourth most visited weather website globally in May 2026. Those are market signals, not audited revenue, yet they demonstrate a distribution asset that should be monetizable.
- The judgment is unfavourable but not terminal. MapMakers has credible products, intellectual property, customer relevance and enough gross profit to recover. It has not publicly demonstrated that owning network capability produces a return above the cheaper alternatives of multihoming through hosting providers, using content-delivery services and concentrating capital on software and data quality.
Independence is an economic choice, not a badge
The incentive to avoid pure resale is obvious in this business. A firm that merely buys a hosted server, republishes third-party forecasts and resells access has little control over latency, restoration priority or unit cost. It can change vendors, but it cannot credibly sell operational accountability. MapMakers serves use cases where the customer does not buy weather as entertainment. Aviation staff, hydrologists, agricultural operators, emergency planners and logistics teams buy a decision input. If that input is unavailable when a flight is being prepared, a river is rising or a route is being planned, the value of the service falls much faster than the length of the outage.
Owning an autonomous system and address space can improve that position. Portable addressing can reduce migration friction. Multiple upstream contracts can lower the risk that one carrier failure removes a service from the internet. Direct control of routing policy can shorten response time during an incident. A company can separate public traffic, data ingestion, customer access and administrative services instead of accepting the architecture chosen by a hosting reseller. Those are real benefits.
They are not free, and they are not the same as full independence. The company still buys physical connectivity, power, rack space, equipment and often distributed protection from larger suppliers. Engineers have to maintain routing configuration, security, capacity and address records. Hardware needs spares and refresh. Customers may demand support beyond office hours. A Russian operator with cross-border vendors also carries currency, payment and export-control friction. The correct comparison is not ownership versus zero cost. It is the total cost of controlled network capability versus a managed hosting alternative that can provide two carriers, two facilities, geographic failover and a contracted service level.
That comparison matters more after a weak year. When revenue is rising, additional network control can be justified as preparation for scale. When revenue falls by one fifth and almost all net profit disappears, every fixed layer has to defend its return. A resource that exists in a registry but does not carry visible routes is especially difficult to defend as a current resilience asset. It may preserve optionality. Optionality has value, but only if the company can use it quickly, at a lower expected cost than buying equivalent capability when needed.
The economic question is therefore precise. Do enterprise customers pay a premium for MapMakers to take responsibility for availability, local deployment and support? Does Gismeteo generate enough advertising, data or subscription income to finance high-traffic infrastructure? Do owned resources reduce supplier cost or merely sit alongside outsourced hosting? Public evidence supplies pieces of the answer, but not a clean yes.
The operating company is a weather business, not a retail ISP
The identity boundary must be established before the economics can be judged. RIPE NCC lists "MapMakers Group" Ltd as a Russian Local Internet Registry at Novovagankovsky Lane, building 5, in Moscow, with a Gismeteo email address. The RIPE Database identifies the organisation as ORG-GL201-RIPE, gives Russian registration number 1037739071204 and associates it with AS203444. The company website uses the English name MapMakers Group Ltd, the same Moscow address and tax number 7703209129. Russian company records identify the legal person as Scientific and Production Centre Map Maker LLC, founded in 1999.
That alignment is stronger than a shared brand. The registration number in the RIPE record matches the public corporate record; the address, tax number, telephone range and Gismeteo contacts point in the same direction. The English name is therefore reasonably treated as the international label of the operating company, rather than a separate carrier that happens to share an office.
There is one clerical warning. The footer on the company site displays registration number 1027739575971, while the RIPE record and multiple corporate-data services show 1037739071204. The latter is also the number attached to the active legal person and the 1999 incorporation date. The shared tax number makes confusion with a different taxpayer unlikely, but a company selling trusted information should correct an identifier inconsistency on its own site. Reliability begins with accurate identity data.
The operating boundary is also clear from the company's description of its work. It names three areas: professional meteorological and geographic-information software; integration and distribution of weather datasets to companies, public bodies and civil aviation; and operation of the high-traffic Gismeteo portal. Its declared activities include software development, databases and information resources, web portals, hydrometeorological information, forecasting and cartography. This is a vertically integrated information business with a network requirement. It is not public evidence that the company sells broadband, IP transit, colocation or managed corporate networks.
That distinction prevents a common analytical mistake. RIPE membership is an institutional and operational fact. It indicates that an organisation can hold and administer internet number resources under the regional registry's rules. It does not certify retail service quality, prove an active customer network or establish telecom revenue. Large websites, banks, universities, industrial groups and data-intensive businesses become registry members because address control serves their own operations. MapMakers fits that pattern more naturally than it fits the profile of a regional access provider.
The realistic perimeter is thus narrower and more interesting. MapMakers controls products, software, weather processing, customer relationships and a major consumer-facing brand. It appears to control some network policy and address rights. It does not control the physical carriers, all facilities, the global routing system, app stores, public weather observations or the customer's own connection. Any reliability promise spans that whole chain, while the company's direct control covers only part of it.
Three revenue engines share one reliability burden
The first revenue engine is professional software. GIS Meteo is presented as a comprehensive environment for receiving, processing, visualising and documenting meteorological information. Its applications cover synoptic forecasting, aviation, hydrology, ocean conditions, agriculture and pollution analysis. The product supports weather maps, numerical-model outputs, scheduled processing, specialised meteorological formats and server-side work. The software can be installed for a professional user, supplied with hardware, deployed inside a customer's corporate network or accessed online.
This model can produce licence revenue, implementation fees, maintenance and support income. It can also create switching costs. A professional weather office that has trained staff, established map templates, connected data feeds and embedded output into procedures will not replace the system as casually as a consumer changes a mobile app. That is where local accountability can command a price. The vendor understands Russian institutional requirements, can support an on-premises deployment and can combine meteorological expertise with software work.
The second engine is weather data as a service. The company's weather interface offers forecasts, archives, radar information, maps and reanalysis data. Its public description says customers receive a trial, connect the interface to their own IT environment and then sign a subscription or service agreement. Charges depend on daily requests, the weather variables selected and the required availability level. The customer account records requests, areas and locations. This is a recognisable usage-based model with an availability premium layered on top.
This is the business line most directly connected to network economics. Each extra request has a low distribution cost once data processing and capacity are in place, but the platform has to be ready for peak demand and incidents. Forecast generation, observation ingestion, storage, database queries, maps and customer authentication all have different load patterns. The firm can earn attractive incremental margins if volume grows faster than the fixed cost base. It can also suffer badly if demand falls while engineers, facilities and support commitments remain.
The third engine is Gismeteo. The company says it develops and supports the portal, serves millions of people daily and processes heavy request volumes. Public app and traffic signals support the claim that the brand has substantial reach, although they do not verify the exact daily audience. A consumer weather portal can generate advertising income, commercial distribution, brand value and a low-cost acquisition path for business data products. It also imposes a reliability burden that free users do not directly fund. Traffic spikes during severe weather, exactly when availability matters most and infrastructure can be most expensive.
Consulting and public-sector work sit across the three engines. MapMakers offers meteorological consulting during office hours or around the clock. Its site displays endorsements or partnership references involving aviation, rail, hydrometeorological bodies, universities and emergency-response institutions. Procurement aggregators show completed contracts for software licences and data-processing equipment. These signals support the existence of paying institutional demand, but the company does not disclose a customer list with revenue shares, contract duration or renewal rates.
The combination can be powerful. Consumer reach provides brand and data demand; professional software produces higher-value contracts; subscriptions produce recurring usage revenue; consulting helps define the customer's need and pull through a larger contract. It can also conceal weak economics. Advertising may subsidise free traffic. Bespoke deployments may require costly engineering. Government work may be lumpy. Support can become open-ended. Without segment accounts, revenue growth cannot be separated from value creation. A ruble earned from a repeatable data subscription is not economically equivalent to a ruble earned from a labour-heavy custom installation.
The network record proves capability, then raises a harder question
The RIPE record is specific. MapMakers holds AS203444, an IPv4 allocation covering 185.134.200.0/22 and an IPv6 allocation covering 2a03:a4a0::/32. A /22 contains 1,024 IPv4 addresses. The IPv6 block is vastly larger in address count, though address count alone says little about use. Route records exist for portions of the IPv4 block. The autonomous-system record, created in January 2016, historically named AS49063 and AS9002 as upstream paths.
This is more than a rented virtual server. It shows that the company invested in registry membership, addressing and routing capability. It also provides a useful date: the resources were obtained when the consumer portal and professional data services were already established, suggesting the decision was made to support an operating business rather than an untested idea.
The current observations are less reassuring. RIPEstat reported no announced prefixes for AS203444 over its recent observation window on 10 July 2026. Its routing-status view showed zero of 327 IPv4 observation peers seeing the autonomous system and zero announced IPv4 or IPv6 space. The last visible route in that view was recorded in April 2026. The allocated IPv6 /32 had no observed route history. Public PeeringDB data returned no network entry for the autonomous system.
Those negatives must be interpreted carefully. A missing public announcement does not mean the company has no servers, no private network and no service. It may use provider-assigned addresses, private interconnection, cloud or hosting addresses, or a temporarily withdrawn autonomous system. Public route collectors do not observe every private arrangement. Nor does a missing PeeringDB profile prove the absence of peering; participation in that database is voluntary.
What the evidence does mean is that the owned autonomous system was not providing visible internet reachability at the observation time. That weakens any claim that it was then delivering active multihomed independence. Historic policy named two upstreams, but a policy statement last modified years earlier is not proof of two live contracts in 2026. Registered route objects are authorisations and records, not measurements of current traffic.
The public endpoints reinforce the distinction. A DNS check on 10 July placed mapmakers.ru at 92.242.36.124, inside a prefix originated by AS49063. The Gismeteo site resolved to four addresses in 31.172.76.0/24, a prefix originated by AS48018, and used Cloudflare name servers. These observations can change, and they reveal only public entry points. Still, they show the visible services depending on provider networks rather than being announced from AS203444.
That arrangement may be entirely rational. A specialist hosting company can spread equipment, attack mitigation, facilities and network staff across many customers. Cloud-based authoritative DNS can improve resilience. Separating the corporate site from the high-volume consumer service limits common failure. The mistake would be to call that arrangement independence. It is a mixed model: owned resources preserve control and options, while external networks carry the publicly visible load.
Routing security is another gap. RIPEstat returned no validating route-origin authorisation for the tested IPv4 announcements, producing an unknown status rather than valid or invalid. There was no visible route at the observation time, so this is not evidence of an active hijack or misconfiguration. It is nevertheless a governance watchpoint. If the company restores public announcements, signed origin authorisations should be part of the basic control set. A weather business that sells dependable delivery should not leave route origin unverifiable when a straightforward registry mechanism is available.
Pricing must convert avoided loss into recurring cash
The company's pricing logic is visible even though its prices are not. The weather interface is sold by volume, data selection and required availability. Consulting is priced according to working-hours or round-the-clock coverage. Software is quoted individually. This tells us the company understands that customers differ in load and consequence of failure. It does not tell us whether price captures those differences.
For an agricultural customer, a basic forecast feed may improve planning but tolerate a short interruption. For aviation or emergency-response work, the consequence of stale data can be far higher. The vendor should therefore charge not only for requests but for reserved capacity, recovery objectives, support coverage, geographic duplication and contractual service credits. Otherwise the most demanding customers consume the most resilience while paying only for data volume.
The economic value can be framed as avoided loss. A logistics company does not care how many routers MapMakers owns; it cares whether weather intelligence reduces delays, damage, fuel use or missed delivery windows. An airport weather office cares whether a system supports safe and timely decisions. A renewable-energy operator cares whether forecast quality improves generation planning. Reliability commands a premium only when it protects a measurable operating outcome.
This favours a tiered contract design. A low tier can offer standard access with best-effort support. A business tier can reserve capacity and define response times. A critical tier can include 24-hour support, duplicate delivery paths, customer-side deployment options and a tested recovery process. Each tier should have a clear gross-margin target. Bespoke exceptions should be priced as engineering work rather than hidden inside a subscription.
The public offer hints at this architecture but stops before the decisive numbers. There is no list price per thousand requests, no published premium for a higher service level, no standard credit regime and no disclosed renewal rate. Flexible pricing is useful for complex buyers; it can also become a euphemism for weak price discipline. A sales team can win revenue by accepting custom requirements that the operations team must support indefinitely.
MapMakers has a second monetisation problem in Gismeteo. A huge free audience creates bargaining power with advertisers and commercial partners, but it also trains users to expect the service at no direct charge. Consumer traffic is valuable if advertising yield, data partnerships or paid features exceed the cost of serving it. Download counts and website rankings do not answer that question. The 2025 accounts suggest that scale alone did not protect earnings.
The burden of proof lies with retention and margin, not with audience. A high-traffic portal that sends business customers to profitable subscriptions is a strategic asset. A high-traffic portal that requires constant capacity, support and product work while advertising weakens can destroy value despite impressive reach. Segment disclosure would settle the point; the company does not provide it.
The 2025 accounts reveal the price of fixed ambition
Public aggregations of filed accounts provide the strongest economic evidence. Revenue was RUB 562.79 million in 2025, down 20.46% from RUB 707.55 million in 2024. Gross profit fell from RUB 490.07 million to RUB 367.05 million, a decline of 25.1%. Operating profit fell from RUB 135.81 million to RUB 33.52 million. Net profit fell from RUB 65.52 million to RUB 1.875 million. Headcount moved from 69 to 70.
The margins make the contradiction clearer. Gross margin declined from 69.26% to 65.22%, still a strong level for many businesses. Operating margin dropped from 19.19% to 5.96%. Net margin collapsed from 9.26% to 0.33%. Administrative expenses declined only 5.85%, from RUB 354.26 million to RUB 333.53 million, while gross profit fell by a quarter.
This is operating leverage in its unpleasant form. The company retained a large amount of gross value after direct cost, but its overhead base did not adjust nearly as fast as revenue. Just 9.6% of 2025 gross profit remained as operating profit, compared with 27.7% a year earlier. Revenue per employee fell from roughly RUB 10.25 million to RUB 8.04 million even though staffing was almost unchanged.
The figures do not identify which business line weakened or which expense rose. One should not assign the decline to network spending without evidence. Administrative costs include much more than connectivity, and cost of sales may contain data, infrastructure, licences, staff or contractors depending on accounting policy. Other income and expense also affected the final result. The accounts establish pressure, not its precise cause.
They do, however, set a strict test for capital allocation. At a 0.33% net margin, there is almost no room for an unnecessary layer of fixed cost. Every owned server, spare, facility commitment, support shift and registry obligation must either reduce another cost, preserve revenue that would otherwise be lost or enable a price premium. Strategy without that resource allocation is marketing.
The longer history warns against reading a single year as a stable trend. Revenue rose from RUB 341.42 million in 2019 to RUB 454.20 million in 2020, RUB 642.79 million in 2021, fell to RUB 460.18 million in 2022, then rose to RUB 671.28 million in 2023 and RUB 707.55 million in 2024 before the 2025 decline. Net profit was negative in 2019, strong in 2021, much lower in 2022, recovered in 2023, then fell in both 2024 and 2025. This is not a smooth subscription compounding story.
Volatility could reflect public contracts, advertising conditions, one-off licences, data agreements, currency, customer timing or cost recognition. Without segment information, none can be confirmed. The strategic implication is the same: the company needs more recurring contracted revenue and less dependence on a small number of large decisions. Reliability investments are easiest to finance when customers commit for multiple years and pay separately for service obligations.
The balance-sheet evidence offers some protection. One public financial analysis reported assets of about RUB 322 million and net assets of about RUB 258 million at the end of 2025. That is not the profile of an insolvent shell. It gives MapMakers room to repair the model. But a near-zero net margin is not acceptable simply because the company remains solvent. The purpose of controlled infrastructure is to improve durable cash generation, not merely to survive a weak year.
Reliability has four cost layers, only one of which is address space
The first cost layer is computing and storage. Weather services ingest observations, model outputs, radar data and archives; transform them; store time series and maps; and answer user requests. Professional products may also require customer-side servers or hardware. Capacity must be sized for peak load and for parallel operation during maintenance. Storage grows even when revenue does not. Archives create customer value, but they also create a permanent cost unless older data can be moved to cheaper media.
The second layer is connectivity and facilities. An autonomous system needs upstream access to be useful. Redundancy needs more than two logical sessions in one building: it requires path, power and facility diversity. If both carriers enter through the same duct, or both services depend on one data hall, the second contract may provide little resilience. MapMakers does not publish topology, facility count, committed bandwidth, traffic, outage history or capacity utilisation. Those omissions prevent an external calculation of cost per delivered request.
The third layer is labour. The public accounts show 70 employees in 2025. The company describes a broad technology set, professional meteorological work, data integration, consumer web operations, support and bespoke deployments. That breadth is expensive for a small enterprise. It requires software engineers, database skills, infrastructure operations, meteorologists, customer support, sales and administrative functions. A single product team can share components across business lines, but a wide portfolio can also fragment scarce expertise.
Field support is particularly important. On-premises software and hardware create travel, spare-parts and compatibility obligations. A customer may run an older operating environment because replacing certified equipment is costly. The vendor then supports old and new versions simultaneously. Revenue from the original licence can look highly profitable, while the long tail of support consumes cash later. Maintenance contracts must fund that burden explicitly.
The fourth layer is compliance and continuity. RIPE membership has a visible annual service fee of EUR 1,800 per LIR account in 2026, plus an ASN fee. Those amounts are small relative to MapMakers' revenue. The costly part is the surrounding work: accurate records, sanctions checks, information security, data handling, Russian infrastructure requirements, procurement documentation and the ability to support customers in regulated fields. External equipment and software restrictions add vendor-screening and replacement work.
Equipment refresh is where geopolitical cost becomes operational. Servers, storage, network devices and support contracts have finite lives. A company can extend life with spares and local expertise, but deferred replacement eventually increases failure risk and power cost. Substituting vendors may require testing, software changes and staff training. If customers demand domestic technology and local hosting, MapMakers may win contracts, but it also inherits the cost of proving that the alternative works.
The company claims its weather interface is supported by its own import-independent infrastructure in Russia. That claim matters commercially because local control can appeal to public and regulated buyers. It is not independently quantified. The visible web endpoints on external networks also show that "own" cannot mean free of suppliers. The useful interpretation is domestic operational control over selected systems and data, not physical self-sufficiency.
Upstream diversity exists on paper; supplier concentration remains
The historic autonomous-system record lists AS49063 and AS9002. That is consistent with an intention to buy connectivity from more than one network. AS49063 is associated with DataLine, while AS9002 is associated with RETN. If both were live, physically diverse and able to carry the company's portable addresses, the arrangement could reduce dependence on either one.
Current route observation does not confirm that arrangement. AS203444 was not visible, and the corporate site itself was reached through AS49063 address space. Gismeteo used addresses in AS48018. The visible architecture therefore distributes services across suppliers, but not through the company's own public routing domain. That can still be resilient. It shifts the control mechanism from routing under MapMakers' own address space to contracts and deployment across provider networks.
The downside is migration and accountability. A provider-assigned address can tie a service to a host. Moving a database-heavy application is harder than changing DNS. During an outage, the customer of the weather service calls MapMakers, while MapMakers may depend on a facility or carrier to restore service. The company carries reputational loss even if the immediate fault sits upstream.
Supplier concentration also extends beyond carriers. GIS Meteo processes data from Russian and international meteorological centres and supports WMO formats. Gismeteo depends on mobile platforms and search distribution. The website uses external DNS infrastructure. Email relies on outside services. A professional deployment may depend on operating systems, database products or specialist hardware. The public technology list contains both domestic-operable open technologies and components whose commercial support can be affected by cross-border restrictions.
This is why redundancy must be priced as an end-to-end outcome. Two carriers do not compensate for one database. Two sites do not compensate for one irreplaceable data licence. Two copies do not help if recovery has never been tested. MapMakers should sell and measure the customer's usable service, not the number of components it owns.
The realistic alternative is selective ownership. Keep the internet number resources and routing competence because they preserve leverage. Use specialist providers where their scale is superior. Place critical services across genuinely independent failure domains. Maintain the ability to move high-value business endpoints onto owned addresses when that improves recovery. Do not run commodity infrastructure merely to claim independence.
Customer strength is real, but concentration is hidden
MapMakers' market position is stronger than that of a small generic software vendor. Its site points to aviation, rail, hydrometeorology, emergency response, universities, agriculture and energy. Its products address specialised workflows rather than generic weather widgets. The Russian software register includes GIS Meteo, according to the company and the public registry entry. Procurement data aggregators show completed public contracts, including software-use rights. A long operating history and professional intellectual property raise the cost of replacement.
The Gismeteo brand adds a second form of power. Similarweb estimated gismeteo.ru as the fourth most visited weather site globally in May 2026, with 2.71 pages per visit and an average visit of two minutes and nineteen seconds. Google Play showed more than five million app downloads and over 61,000 reviews in July 2026. These figures are not accounts, and Similarweb uses estimates. They nevertheless indicate that the portal is not a marginal side project.
Scale can lower the unit cost of acquiring a business customer. It can produce real-world load experience and a large feedback base. It can make the brand familiar to buyers before a formal procurement begins. It can also generate data about demand patterns and product performance. These benefits may explain why MapMakers maintains broad infrastructure and a large team despite being a small legal enterprise.
The concentration risk is not disclosed. The company does not state how much revenue comes from Gismeteo advertising, any large commercial partner, public procurement, aviation, rail or hydrometeorological institutions. It does not provide top-customer shares. The revenue volatility and the sharp 2025 decline are consistent with concentration or lumpy contracts, but they do not prove either.
Public-sector and regulated customers can be attractive because contracts may be sticky and local accreditation matters. They can also impose slow procurement, fixed pricing, extensive documentation and payment timing risk. Bespoke requirements can turn a product company into a project contractor. If one or two large agreements end, a 70-person cost base does not disappear with them.
The consumer audience has a different concentration risk: platform and advertising dependence. Google Play lists the app under Winner Satellite and identifies Roman Shmelkin in the developer information, rather than presenting the legal company name prominently. That does not invalidate the relationship, but it complicates brand-to-entity attribution for an outside buyer. Search engines, mobile stores and ad intermediaries can change distribution or terms without negotiating with MapMakers.
The company needs a revenue-quality view that the public cannot see: recurring contracted revenue, renewal rate, revenue by product, gross margin by service tier, top-ten customer share, public-sector exposure, consumer advertising yield and cost to serve each request class. Without those figures, the strongest market signals remain evidence of relevance rather than proof of pricing power.
Competitors can attack from above, below and beside
At the top of the market, global weather-data and cloud companies can offer broad coverage, mature interfaces and infrastructure spread across many regions. Their scale can lower unit cost and accelerate product development. For a Russian public body or regulated customer, however, cross-border procurement, payment, data location and service continuity can make those alternatives impractical. MapMakers' local presence, Russian software status and domain knowledge are meaningful advantages.
At the bottom, free public forecasts and commodity weather interfaces anchor customer expectations near zero. Many users do not need a professional meteorologist's workstation or a specialised archive. They need a temperature, precipitation probability and wind forecast. If MapMakers does not demonstrate the economic gain from better data, integration and support, customers will compare it with a cheap interface rather than with the cost of a bad operational decision.
Beside it are Russian software firms, public hydrometeorological services, internal customer teams and specialist integrators. A large airline, rail operator or state institution can buy data from one party, infrastructure from another and build its own display. An integrator can combine public data with general geographic software. The alternative to MapMakers is not always another weather company; it can be internal development backed by commodity hosting.
The response should not be more product breadth for its own sake. The portfolio already spans numerous professional domains, web tools, hardware-linked systems, consulting and consumer services. Each additional variant increases support and testing. MapMakers should concentrate investment where it owns defensible meteorological methods, customer workflow knowledge or distribution. Commodity display and infrastructure should be standardised or bought.
The best competitive position is local accountable intelligence. That means a customer can buy data, software, integration and support from one party that understands the operating consequence. It also means MapMakers must accept contractual responsibility and price it. A vague claim of reliability without measured delivery will not beat a cheaper alternative.
The network assets help only if they reinforce that position. Portable addressing can support controlled migration. Routing knowledge can reduce incident time. Multiple providers can improve availability. None of these creates a customer moat by itself. The moat is the combination of weather expertise, embedded workflow, trusted delivery and switching cost.
Regulation protects local demand while raising the cost of failure
Russian policy creates both advantage and burden. Domestic software registration and IT accreditation can improve eligibility and credibility in local procurement. Data localisation and infrastructure expectations can favour a supplier operating in Russia. The 2019 law on the stable and secure functioning of the Russian internet also reflects a policy environment in which network control and local continuity matter.
The same environment increases compliance work. A company that handles user accounts, corporate data and critical operational services must maintain appropriate records, security and local processing. Public-sector buyers can demand certifications, documentation and long support commitments. Regulatory change can require implementation without a matching price increase under an existing contract.
Geopolitics increases equipment and vendor risk. European restrictions cover categories of technology and certain business software supplied to Russia. The exact application depends on the product, counterparty and contract; it should not be reduced to a blanket statement that all technology is unavailable. The practical effect is uncertainty: vendors, banks and distributors conduct more checks, support can be withdrawn, and replacement choices narrow.
RIPE membership has its own sanctions context because the registry is based in the Netherlands and follows European rules. Its Q2 2026 transparency report explains that it freezes registration, not use, when applicable sanctions affect a holder; frozen parties cannot acquire or transfer resources, but their addresses are not simply deregistered. The public report does not identify MapMakers as affected, and nothing reviewed here establishes that it is sanctioned. The relevant risk is procedural and banking friction, not an allegation against the company.
Financing conditions magnify all of this. The Bank of Russia's key rate was 14.25% in early July 2026. A high nominal cost of capital raises the hurdle for spare equipment, second sites and long-payback infrastructure. Even a company financing investment from cash should compare the return with that opportunity cost. Redundancy that preserves a high-margin contract may clear the hurdle; vanity ownership probably will not.
Currency is another asymmetry. Registry fees, imported equipment or external services may be priced in euros or other currencies, while most revenue is likely in rubles. A weaker ruble raises cost before contracts can be repriced. The annual RIPE fee itself is immaterial next to more than half a billion rubles of revenue, but it illustrates the mismatch. Hardware and specialist support can make it material.
These pressures strengthen the case for local competence and weaken the case for unnecessary complexity. MapMakers can benefit as customers seek a Russian vendor able to deploy locally. It must use that demand to raise recurring prices and standardise delivery. Otherwise regulation becomes another fixed burden absorbed by a margin that already fell close to zero.
Market signals show reach, not satisfaction with the economics
The strongest non-company signal is audience. Similarweb's May 2026 category estimate placed Gismeteo behind weather.com, accuweather.com and weathercn.com, and ahead of many national and commercial weather services. The estimate may be wrong in absolute terms, but a top position across a broad category supports the conclusion that the portal has substantial reach.
Google Play supplies a different signal. More than five million downloads and tens of thousands of reviews show a long-lived user base. The displayed rating of 3.9 is solid but not dominant. Individual reviews praise clarity and local nowcasting, while others report stale updates, differences between app and website readings, or historic battery use. Individual comments are anecdotes, not representative measurements. They are useful because the complaints map directly to the commercial promise: consistency, freshness and efficient delivery.
The app was updated in April 2026 with notes referring to stability, performance and critical reliability issues. That wording does not quantify incident frequency, but it indicates active maintenance and acknowledges that reliability is an ongoing product issue. For a company selling weather delivery to business customers, consumer reliability work can improve shared components. It can also consume engineering capacity without producing direct payment.
Employment-market evidence is thin. A technology-career profile listed no open roles and too few employee reviews for a meaningful rating. That says little about retention or capability. It does reinforce the broader disclosure gap: outsiders cannot tell whether the stable headcount reflects deliberate retention of scarce expertise, difficulty resizing or a shift toward contractors.
The market signals therefore support two conclusions and no more. Gismeteo is a valuable distribution asset, and users notice freshness and reliability. They do not show the portal's advertising yield, subscription conversion, cost per visit or contribution margin. Reach is an input to value creation, not value creation itself.
What would change the judgment
First, live network evidence would matter. If AS203444 resumes stable announcements from two physically independent upstreams, with valid route-origin authorisations and a documented recovery test, the owned resources become an active resilience asset rather than an option. Traffic volumes and the share of business services carried on owned addresses would show whether the design is economically important.
Second, service economics would matter. Revenue by segment, recurring contract value, renewal rates, gross margin by service tier and top-customer concentration would reveal whether the business is compounding or living from periodic deals. A disclosed price premium for higher availability, supported by measured uptime and renewal, would directly answer who pays for independence.
Third, 2026 accounts could reverse the view. If revenue recovers while administrative expense remains controlled, the 2025 collapse may prove temporary. If operating margin returns above the mid-teens and cash generation funds infrastructure without expensive borrowing, the company can justify retaining broad capability. If revenue remains near 2025 levels and margin stays close to zero, the cost base needs structural reduction.
Fourth, clarity on Gismeteo would help. Audited audience, advertising and commercial-data revenue, together with direct infrastructure cost, would show whether the portal subsidises the professional business or consumes its profit. A top-four traffic estimate is strategically impressive; it is financially incomplete.
Fifth, customer evidence would matter. Multi-year agreements in aviation, rail, energy or public weather services that separately price 24-hour support, local deployment and recovery would demonstrate pricing power. A broad logo page without contract values cannot do that.
Finally, identity and governance housekeeping should improve. Correcting the inconsistent registration number, maintaining current routing records and publishing a concise security and availability statement would cost little relative to the trust gained. The company sells dependable information. Its own public records should meet the same standard.
The conclusion: keep the option, stop pretending it proves independence
MapMakers has more substance than its small routing footprint suggests. It operates specialised software, weather-data services and a consumer brand with considerable reach. It has existed for more than two decades, owns intellectual property, serves professional use cases and retained a strong gross margin even in a poor year. There is a viable business to defend.
The public evidence does not show that network ownership is currently earning an adequate return. The autonomous system was not visibly announcing routes in July 2026. The public services sat on provider networks. The company still depended on facilities, carriers, DNS, platforms and weather inputs. Revenue fell by one fifth in 2025, operating margin fell to 6% and net margin to 0.33%. Independence did not protect the accounts.
The rational strategy is not to surrender the resources. A /22, a /32, an autonomous-system number and the competence to use them preserve bargaining power and migration options that may be difficult to recreate. The rational strategy is to stop treating those holdings as proof of reliability. Activate them where they lower expected outage cost or improve contract value; otherwise use scaled providers and hold them as contingency.
Customers who benefit from redundancy should pay for it through explicit service tiers. Customers demanding local accountability should fund local support. Bespoke deployment should carry a bespoke price. Free consumer reach should either generate advertising and commercial conversion or face a disciplined cost ceiling. Suppliers should carry contractual consequences where they control failure, while MapMakers retains end-to-end responsibility to its own customer.
The downside currently sits too heavily with the company. It carries staff, product breadth, support, infrastructure and compliance through weak revenue, while customers see no public tariff for the resilience they consume. That is why revenue scale did not become durable value in 2025. Until MapMakers can show recurring reliability premiums and restored network evidence, the cold conclusion is unavoidable: it owns useful independence options, but it has not shown that customers pay enough to make independence itself a profitable product.