The sanctioned workload chooses first
Start with the buyer, not the vendor. A Tehran software team with a payment page, a media archive, a database-backed public service or a high-traffic education product has to decide where the workload can actually live. The tidy answer in most markets is to buy a global cloud account, add a CDN, put the database in a managed service, pay with an international card, and negotiate compliance later. In Iran the order is reversed. The team first asks which provider can take payment, keep latency low for domestic users, keep the service alive when international routes degrade, satisfy local institutional expectations, and still look credible enough that engineers do not feel trapped.
That choice makes Softqloud and ArvanCloud economically interesting before one reads a single company slogan. A pure host rents servers. A sovereign-cloud substitute sells a bundle of availability, local procurement, routing, compliance posture and political risk absorption. ArvanCloud's public pricing page turns that abstraction into a visible commercial anchor. Its CDN page shows a Basic plan at free, a Growth plan at EUR 20 monthly, and a Professional plan at EUR 200 monthly, with six-month and twelve-month discounts and an Enterprise tier priced by request (https://www.arvancloud.ir/en/pricing/cdn). The same site says cloud-server cost is calculated hourly according to resources, with examples visible in search snippets such as a one-CPU, 2GB RAM, 25GB storage cloud server around EUR 10 and a two-CPU, 4GB RAM, 25GB storage configuration around EUR 20 (https://www.arvancloud.ir/en/products/cloud-server). Support pricing adds another anchor: Basic is free, Growth is EUR 5 monthly, Professional is EUR 150 monthly, and Enterprise is customized with a claimed 59-second ticket response target (https://www.arvancloud.ir/en/pricing/support). Managed database pricing separately shows SSD disk at EUR 0.19 per GB monthly, local SSD at EUR 0.27 per GB monthly and an IPv4 address at EUR 1.33 monthly (https://www.arvancloud.ir/en/pricing/databases).
Those numbers do not prove revenue. They prove the shape of the product. ArvanCloud is not merely offering a rack and a login. It is pricing a local stack: CDN, compute, database, IP address, support, APIs, Terraform, status reporting, network map and industry packages. A domestic buyer comparing it with a foreign cloud does not compare only EUR 20 with a Cloudflare or AWS line item. It compares the full friction of paying, routing, support language, regulatory exposure and continuity.
The first hard risk anchor is regulatory, not technical. On June 2, 2023, the U.S. Treasury's Office of Foreign Assets Control designated Navyan Abr Arvan Private Limited Company, known as Arvan Cloud, under Executive Order 13846, along with two senior employees and ArvanCloud Global Technologies L.L.C. in Dubai (https://home.treasury.gov/news/press-releases/jy1518). The OFAC recent action page lists the Iranian entity, the Dubai affiliate, business-registration identifiers and aliases including Arvan Cloud, ArvanCloud and Abr Arvan (https://ofac.treasury.gov/recent-actions/20230602). Treasury alleged that ArvanCloud helped lay groundwork for Iran's National Information Network and had contract language involving lawful interception. ArvanCloud and Softqloud have disputed allegations around their roles; Softqloud's own German-language statement says its work with Abr Arvan was commercially, spatially and financially separated and that it terminated the joint offering in 2022 (https://softqloud.com/statement.html).
The buyer therefore faces an asymmetric decision. For an Iran-facing workload that must take local payment, serve domestic users and remain useful when foreign services are hard to buy, the public evidence changes the buy-versus-foreign-cloud decision toward a local provider for non-sensitive, operationally necessary workloads. For a cross-border workload, a company with foreign investors, a user base outside Iran or a compliance department exposed to U.S. or U.K. sanctions, the same evidence pushes the decision the other way. Foreign cloud becomes more valuable not because it is cheaper, but because it avoids the sanctions, reputational and payment frictions attached to ArvanCloud.
The document that would most change this judgement is specific: either an OFAC delisting or specific-license/payment-processing document that lets banks, card networks and non-Iranian suppliers clear ArvanCloud-related transactions without fear, or a signed customer-payment document showing an Iranian buyer can lawfully and reliably pay a foreign cloud/CDN provider for the same workload under 31 CFR section 560.540 and OFAC FAQ 441/1088 (https://ofac.treasury.gov/faqs/441 and https://ofac.treasury.gov/faqs/1088). Without that payments proof, local cloud has a scarcity premium. With it, ArvanCloud has to compete more like an ordinary cloud vendor.
The identity is a stack of traces, not a clean company card
Softqloud's public identity is not as simple as the directory name suggests. PeeringDB lists AS208006 as "Softqloud", with organization "Softqloud GmbH", also known as Softqloud, and a company website override pointing to https://www.arvancloud.com (https://www.peeringdb.com/net/26232). The same PeeringDB record describes the network type as Content, traffic as 20-50Gbps, traffic ratio as heavy outbound, geographic scope as global, an open peering policy, two operational 1G exchange connections at DE-CIX Mumbai and DE-CIX Delhi, and one interconnection facility at COMNET Datacenter Istanbul. The structured PeeringDB API gives the same public facts, including organization id 28850, network id 26232 and facility/exchange identifiers (https://www.peeringdb.com/api/net/26232).
That record is useful because it is not written like a marketing page. It shows a small but real interconnection footprint tied to a German-named organization and an ArvanCloud website. It does not show customer revenue, ownership, full traffic flows or current control. It also does not make AS208006 the Iranian internet. It is evidence of a public network resource associated with the ArvanCloud commercial surface.
RIPE RDAP adds a second layer. AS208006 is named ARVANCLOUD-CDN and active, with registration on October 18, 2019 and last change on November 28, 2022 (https://rdap.db.ripe.net/autnum/208006). The RDAP record lists ARVANCLOUD GLOBAL TECHNOLOGIES L.L.C. in Dubai as an organization entity, with UAE addresses and an abuse contact at arvancloud.ae. A related RIPE RDAP record for 185.215.232.0/22 names the allocation AE-ARVANCLOUD-20170802 and lists ARVANCLOUD GLOBAL TECHNOLOGIES L.L.C. (https://rdap.db.ripe.net/ip/185.215.232.0). Another RDAP record for 185.143.232.0/24 is assigned in Iran, names AbrArvan and Arvan Cloud roles, and includes the remark "AbrArvan BGP Anycast" (https://rdap.db.ripe.net/ip/185.143.232.0).
The public-resource chain therefore runs from Softqloud GmbH in PeeringDB to ArvanCloud website branding, to AS208006 in RIPE as ARVANCLOUD-CDN, to a Dubai affiliate that Treasury later designated, to Iranian and UAE address resources. That is exactly why a plain hosting-company reading is inadequate. The asset is not just a server. It is a cross-border operating surface that tried to connect Iranian cloud demand with international content delivery and supplier/customer access.
The historical relationship is disputed in its implications but not invisible. ArvanCloud's own newsroom posted on September 30, 2022 that Softqloud GmbH had sent a termination notice and that, under the termination terms, ArvanCloud would migrate services by December 31, 2022; the same note says Softqloud had been a commercial partner responsible for ArvanCloud's international customers and suppliers (https://news.arvancloud.ir/en/contract-termination/). Softqloud's own statement says the joint service offering with Abr Arvan began at the end of 2019, was terminated on September 30, 2022, and involved cloud server space and CDN services, while denying ownership ties, data-center operation in Dronten and involvement in Iranian internet controls (https://softqloud.com/statement.html).
Those statements are not identical in tone, but together they answer the basic economic question. Softqloud was not a random name attached to an IP record. It was part of an international commercial arrangement around ArvanCloud. If that arrangement has ended, Softqloud's value is now mostly evidentiary: it marks the international bridge that made ArvanCloud more than a domestic host. If parts of the resource chain continued under the Dubai affiliate, the bridge did not disappear; it was repapered or re-registered. That is why sanctions and payments documents matter as much as network maps.
The business model is local-cloud substitution
ArvanCloud's own website now presents a broad cloud platform. The English homepage and product menus list Cloud Server, CDN, Video Platform, Object Storage, Edge Computing, Cloud Container, Cloud Logs, Managed Database, DNS, Load Balancer, smart routing, DDoS protection, WAF, network firewall, floating IP, API, CLI, SDK, Cloud Shell and Terraform (https://www.arvancloud.ir/en). The pricing overview says the company uses pay-as-you-go, cost-effective plans, enterprise contracts and free services; it says CDN, object storage and video can be used on the Basic plan with additional usage charged by hour, day or month, while large businesses can use annual enterprise contracts (https://www.arvancloud.ir/en/pricing).
That is a materially richer model than "Iranian hosting". Hosting revenue is usually sold by server, disk, bandwidth and support ticket. ArvanCloud sells an operating environment. A developer can start with CDN DNS delegation, add a cloud server, attach an object store, place a database, use Terraform or API provisioning, buy support, monitor status and route traffic through an anycast CDN. The revenue logic is cumulative: cheap entry products reduce switching friction, then storage, support, traffic, enterprise features, dedicated IPs, managed database and reliability needs add spend.
The free tier is not generosity in isolation. It is acquisition cost. ArvanCloud's CDN page offers a Basic plan for free, including free SSL, DNSSEC support, webpage acceleration, DDoS protection and five page rules (https://www.arvancloud.ir/en/pricing/cdn). The public pricing FAQ says free services are available in four products: CDN, Object Storage, Video Platform and Support Services (https://www.arvancloud.ir/en/pricing). That structure makes sense in a market where developer trust is expensive. A small team that is worried about sanctions, reliability or political reputation may not begin with an enterprise contract. It may begin by pointing a low-risk domain at the CDN, watching latency and support, then deciding whether to move the database or the main application.
Support is a revenue line and a trust signal. ArvanCloud's support pricing says Growth includes telephone technical support and a two-hour ticket response, Professional includes a dedicated messenger group, 30-minute ticket response, 24/7 direct technical support and an SLA, while Enterprise includes a dedicated account manager and a 59-second ticket response target (https://www.arvancloud.ir/en/pricing/support). Those promises should be read as commercial claims, not audited performance. But the economic logic is clear. In a restricted market, support is not a cost center hidden behind self-service documentation. It is part of the substitution case against foreign cloud: a local engineer can call, message or ticket a local provider when an AWS billing card fails, a sanctions filter blocks a service, or international traffic is unstable.
Managed database pricing shows how the model moves up the stack. ArvanCloud's database product page promises automatic recovery, high availability through backup nodes, read-only nodes, automatic backup, point-in-time recovery within seven days, firewalling, monitoring and user management (https://www.arvancloud.ir/en/products/databases). Its database pricing page charges by disk, IP address and resource use, with backups up to seven days shown as free and database send/receive traffic shown as free (https://www.arvancloud.ir/en/pricing/databases). That is not a bare virtual machine. It is the kind of sticky service that makes migration away more difficult. Once a customer places operational data, firewall rules, backup habits and support expectations inside ArvanCloud, the provider's value comes from continuity as much as price.
The risk is that the same stickiness can become a liability. A buyer that cannot easily move because the database, CDN, object store and support relationships are bundled is exposed if sanctions deepen, if international routes degrade, if an upstream supplier exits, or if public controversy makes ArvanCloud unacceptable to customers or investors. That is the central economic tension: local cloud is more valuable when foreign cloud is hard to use, but the very conditions that create that value also raise the cost of trust.
The network evidence is visible but incomplete
A cloud platform's market claim depends on network reach. ArvanCloud's CDN network page says its content delivery network covers seven continents, more than 30 countries and 40 cities, with 40 PoPs, 2Tbps of DDoS mitigation and anycast architecture (https://www.arvancloud.ir/en/products/cdn/network). It lists locations across the Americas, Europe, MENA and Asia/Oceania, including New York, Dallas, Los Angeles, Miami, Frankfurt, Amsterdam, Rotterdam, Paris, Istanbul, Tehran, Dubai, Johannesburg, Sydney, Singapore, Tokyo, Hong Kong, Mumbai, Delhi, Pune, Chennai, Seoul and Kuala Lumpur. Some city names are marked as having multiple PoPs.
ArvanCloud's IP ranges page lists CDN edge ranges and tells origin operators to whitelist them to avoid disruptions when PoP sites change (https://www.arvancloud.ir/en/dev/ips). The ranges include 185.143.232.0/22, 188.229.116.16/30, 94.101.182.0/27, 2.144.3.128/28, 37.32.16.0/27 through 37.32.19.0/27, 185.215.232.0/22, 178.131.120.48/28, 94.101.183.0/28 and 78.157.36.112/28. IP lists are not glossy. They are operational evidence. If a customer needs to whitelist the provider's CDN edges, the provider is part of the customer's production surface.
The Looking Glass page adds another operational clue. It lets users run BGP, ping and traceroute checks from selected ArvanCloud routers and says the service is designed to reveal routing paths, round-trip time and packet journeys (https://lg.arvancloud.ir/). A looking glass is not capacity proof. It is a public accountability instrument. Providers publish it because customers, peers and engineers need to test the network rather than trust a map.
The public status page is more useful than a perfect uptime claim because it shows the kind of failure a customer actually prices. On the retrieved page, ArvanCloud showed global services such as DNS, CDN, CDN PoPs, Edge Computing, Panel and API, IAM and Website as operational, and listed named regions such as Bamdad, Simin, Forough, Shahriar and Goethe (https://www.arvancloudstatus.ir/). It also showed recent incidents: a Goethe Datacenter connectivity and cluster availability issue; cloud-server create delays; a statement that an upstream internet provider had performed unannounced maintenance; a note that the issue was limited to an upstream uplink layer; and separate network-bandwidth and DNS-resolution issues affecting services and international PoPs.
That status evidence is commercially important. It shows ArvanCloud is running something more complex than a single hosting shelf. It also shows supplier dependency. A local cloud can only sell sovereignty up to the point where its upstream, facility power, routing, hardware, software and support dependencies hold. When an upstream uplink layer fails, the customer's local-cloud bargain is tested. The provider can recover, communicate and retain trust, or it can turn a price advantage into a migration trigger.
PeeringDB's Softqloud record sits beside this operating evidence. AS208006 has only two listed exchange connections and one facility in PeeringDB, while ArvanCloud's own CDN map claims a far broader PoP surface (https://www.peeringdb.com/net/26232 and https://www.arvancloud.ir/en/products/cdn/network). That mismatch is not necessarily a contradiction. PeeringDB is voluntary and may show only a subset of public interconnection. The mismatch does mean one should not infer the entire network from one registry page. The safest conclusion is narrower: Softqloud/AS208006 is a real public network-resource trace attached to ArvanCloud's international CDN story, while ArvanCloud's complete production network has to be judged by multiple records, public IP ranges, status incidents, looking glass behavior and customer evidence.
Costs are hidden in procurement, power and people
The appeal of ArvanCloud's local-cloud model is obvious: domestic customers get lower latency, local language and support, local payment, a familiar legal environment and some resilience when international services are difficult. The cost base is less visible. Compute requires servers, disks, networking gear, virtualization software, monitoring, backup capacity, facilities, power, cooling, spares and engineers. CDN requires PoPs, transit, exchange ports, cache servers, DDoS capacity, WAF engineering, DNS operations and incident response. Managed databases require storage redundancy, backup media, replication, update discipline and specialist support. Every one of these costs is harder in a sanctioned economy.
Foreign-cloud providers can buy hardware, software, support contracts and insurance through deep global supply chains. An Iranian provider has to handle currency weakness, import restrictions, supplier caution, payment screening and technology availability. It may get a local demand premium because foreign rivals are hard to use; it also pays a local cost penalty because foreign suppliers are hard to buy from. That makes gross margin less predictable than a price page suggests.
The sanctions environment creates a two-sided moat. On the demand side, Iranian firms that cannot easily use AWS, Azure, Google Cloud or Cloudflare have a stronger reason to pay a local provider. On the supply side, a provider whose entity, affiliate or executives are designated has fewer clean options for capital, software, bank accounts, foreign customers, foreign suppliers and insurance. Treasury's June 2023 action also issued a time-limited general license for winding down transactions with Navyan Abr Arvan and ArvanCloud Global Technologies L.L.C., which underscores the practical payment issue: counterparties do not merely ask whether the cloud works; they ask whether a transaction can clear (https://ofac.treasury.gov/recent-actions/20230602).
OFAC's communications guidance complicates the picture. FAQ 441 says section 560.540(a)(1) authorizes fee-based cloud computing services to Iran that support the exchange of communications over the internet (https://ofac.treasury.gov/faqs/441). FAQ 1088 says a cloud-based service provider whose non-Iranian customers provide authorized communications tools into Iran can rely on the authorization if it conducts ordinary-course due diligence and does not generally need to evaluate ultimate end use in certain circumstances (https://ofac.treasury.gov/faqs/1088). That helps explain why not all Iran-related cloud activity is prohibited. But it does not make a designated provider bankable. A general authorization for categories of communications services is not the same as permission to deal with a blocked counterparty.
This is why payments documentation is so valuable. In normal cloud economics, a customer underwrites uptime, price, support and lock-in. In ArvanCloud's case, the customer also underwrites whether a payment rail will remain available, whether a foreign supplier will continue service, whether an account will be flagged by a sanctions-screening tool, whether a public association with ArvanCloud creates reputational cost, and whether a local alternative can meet the same technical standard. The cheapest VM can become expensive if it narrows future options.
Labor is another hidden cost. ArvanCloud's Farsi careers page listed open roles including Cloud Support Technician, DataCenter's Network Engineer, Site Reliability Engineer, Senior Software Engineer, Product Designer, CFO and business roles (https://www.arvancloud.ir/fa/jobs). Job listings are not a headcount audit, but they show the operating functions needed to keep the platform alive: support, data-center networking, reliability engineering, product, finance and partner acquisition. The same page describes the company as a platform for integrated cloud and AI services in Iran and other parts of the world, with dozens of products. That ambition requires expensive people. A local cloud's moat is partly the ability to retain engineers who can operate under constraint.
The demand is domestic, but not only nationalist
Iranian cloud demand is often described as a sovereignty story, and that is partly correct. The state has long promoted domestic digital infrastructure, and U.S. Treasury describes the National Information Network as a countrywide intranet under Iranian authority (https://home.treasury.gov/news/press-releases/jy1518). The Center for Human Rights in Iran has described national data centers as important infrastructure for the NIN and for domestic storage, hosting and communications (https://iranhumanrights.org/2018/01/ir2017-data-centers/). For any provider that can host domestic workloads at scale, state policy creates demand even where private developers would prefer a more open internet.
But ArvanCloud's market is not only a government-policy artifact. Its website targets startups, enterprises, financial services, online shopping, online education, transport, gaming, medical and health, tourism, disaster recovery, banking CDN, cloud security and private cloud (https://www.arvancloud.ir/en). Those categories are ordinary commercial demand. Iranian companies need websites, databases, object storage, video, security, logs, APIs and support whether or not they like the politics of the network around them. A retail site wants checkout uptime. A school wants video lessons to load. A gaming service wants latency. A hospital software company wants backups and access control. A bank wants secure domestic delivery.
That practical demand is why ArvanCloud can be more valuable than a simple host. A host sells capacity. A cloud substitute reduces the number of separate procurement problems a customer has to solve. If a customer can buy CDN, server, database, storage, support and security from one Iranian provider, paid through workable local rails and supported by local engineers, the provider captures integration value. The customer's alternative might be a mix of local VPS hosts, self-managed servers in an Iranian data center, a foreign CDN reached through a difficult account, and an in-house engineer carrying pager duty. ArvanCloud's commercial pitch is to make that alternative look costly.
Competitors and substitutes still matter. TechSci Research's public market page names Afranet and ArvanCloud among players in Iran's cloud-computing market, alongside global and software firms (https://www.techsciresearch.com/report/iran-cloud-computing-market/7880.html). Local substitutes also include Iranian hosting companies, ISP-affiliated data centers, cloud specialists, self-managed colocation and smaller developer platforms. International substitutes include Cloudflare for CDN/security, AWS/Azure/Google where available, and regional cloud or bare-metal providers reached through foreign entities. The exact competitive set changes by workload. For an Iranian news site or e-commerce store, local latency and payment may dominate. For a SaaS product selling abroad, sanctions and investor trust may dominate.
ArvanCloud's pricing suggests it understands the ladder. Free CDN entry fights Cloudflare's habit-forming free tier. Pay-as-you-go cloud servers fight VPS hosts. Managed database and object storage fight in-house operations. Support plans fight the idea that a local provider is less professional than a foreign one. Enterprise contracts fight the procurement department's fear that a self-service account is not enough for critical infrastructure. This is not a trivial hosting strategy. It is a full-stack substitution strategy.
The weakness is developer trust. Trust is not only uptime. It includes the belief that the provider will not become a political liability, will not lose upstream access, will not mishandle legal requests, will not surprise customers with downtime, and will not lock customers into a stack that becomes hard to explain to auditors. ArvanCloud's transparency report tries to address part of this. It says the company reviews requests from governmental and law-enforcement entities, rejects illegitimate requests, notifies customers before disclosure unless a judicial decision requires confidentiality, says no law-enforcement spyware is installed on the network, and says it does not disclose encryption or authentication keys (https://www.arvancloud.ir/en/legal/transparency-report). It then lists quarterly counts of abuse, ownership-rights, judicial-oversight and rejected requests, including high request volumes in 2023 and 2024.
The report is a confidence-building device. It is also an admission that legal and political requests are part of the product environment. A foreign cloud buyer worries about subpoenas too; an Iranian cloud buyer worries about them in a much more contested political setting. The market will discount ArvanCloud if developers believe the transparency report is decorative. It will reward ArvanCloud if developers believe the company is one of the few local providers capable of handling pressure with process, documentation and continuity.
Sanctions can increase local value while lowering external value
Sanctions usually sound like a simple negative for a business. For ArvanCloud they are more complicated. Treasury's designation, the U.K. listing and prior EU action reduce international counterparties' willingness to deal with the company. The UK Sanctions List entry for Arvan Cloud says it is an involved person under the Iran (Sanctions) Regulations 2023 (https://search-uk-sanctions-list.service.gov.uk/designations/IRN0165/Entity). Radio Free Europe/Radio Liberty reported that the EU removed sanctions on ArvanCloud in April 2024, while the company remained under U.S. and British sanctions and Washington had no intention of delisting it (https://www.rferl.org/a/iran-eu-sanctions-arvancloud-internet-censorship/32897604.html). Carnegie's Mahsa Alimardani criticized the EU delisting as setting a concerning precedent (https://carnegieendowment.org/emissary/2024/05/hard-questions-about-the-eu-lifting-sanctions-on-an-iranian-tech-company).
For foreign valuation, this is bad. It blocks clean acquisition interest, reduces supplier optionality, makes bank diligence harder and limits foreign enterprise demand. A global buyer does not need to decide whether ArvanCloud's product is technically good if the payments team cannot approve the vendor. A foreign cloud competitor can also use the designation as a sales argument: even if ArvanCloud is cheaper or closer to Iranian users, the compliance cost may outweigh the infrastructure benefit.
For domestic substitution, however, sanctions can increase scarcity value. If global providers are hard to buy, if cards fail, if accounts close, if routes are unpredictable, and if procurement teams want domestic accountability, a local provider becomes more valuable. Sanctions do not create trust, but they create dependence. The economic question is whether ArvanCloud can convert dependence into loyalty rather than resentment.
The answer varies by workload. A public-interest tool, foreign-facing application or investor-backed startup may treat ArvanCloud as a risk even if the service is technically strong. A domestic e-commerce catalog, local video platform, school portal or bank-facing CDN may treat ArvanCloud as rational infrastructure because the alternatives are operationally worse. A government-adjacent workload may value domestic hosting for political and procurement reasons. A developer community may split: some praising low price and local performance, others viewing the brand as too controversial.
Unofficial and semi-public signals fit that split. G2's public ArvanCloud review snippets praise uptime, customer support, fast downloads, cheap pricing and usability, but the sample is small and platform-hosted (https://www.g2.com/products/arvancloud/reviews and https://www.g2.com/sellers/arvancloud). The Terraform Registry shows an ArvanCloud provider with hundreds of thousands of all-time downloads and latest version 0.6.4 published in November 2022, while GitHub shows a small open-source repository with releases and issues around provider configuration and documentation (https://registry.terraform.io/providers/arvancloud/arvan/latest and https://github.com/arvancloud/terraform-provider-arvan). These are not proof of broad enterprise adoption. They are evidence that ArvanCloud has tried to meet developers where infrastructure decisions are now made: in code, not only in a sales meeting.
The status page provides a less flattering but more useful signal. Recent incidents around Goethe network connectivity, cloud-server creation delay, DNS resolution and bandwidth show that ArvanCloud has operational problems like any cloud provider (https://www.arvancloudstatus.ir/). The important question is not whether incidents exist. Incidents always exist. The question is whether the provider communicates clearly, restores service quickly, explains supplier dependency, and keeps the customer from feeling that local-cloud lock-in has become a trap.
Customer concentration is the missing valuation fact
The most important unavailable business fact is not AS number, PoP count or product list. It is customer concentration by type and revenue. ArvanCloud's about page claims more than 200,000 customers from 38 countries, more than 755 million requests per hour and more than 150 features in six product categories (https://www.arvancloud.ir/en/about). These are large claims, but they are not audited financial disclosures. A buyer would ask how much of the revenue comes from private Iranian startups, government-linked bodies, banks, media, education, retail, foreign customers, free-tier users and enterprise contracts.
The answer changes valuation. If the revenue base is diversified across private domestic customers who pay reliably and use multiple products, the business is a serious local-cloud platform. If a small group of government-adjacent or politically exposed customers dominate, the revenue may be sticky but heavily discounted. If many users are free-tier CDN accounts, the platform may have reach but weaker monetization. If foreign customers were important before Softqloud's termination and sanctions, the international-growth story may have already been damaged.
Customer dependency also changes the buy decision. Suppose a domestic online education company uses ArvanCloud for CDN, database, object storage and support. If all students are in Iran, the local provider's value is high: lower latency, local payment, local support and better alignment with domestic network conditions. Suppose the same company sells courses abroad and needs foreign investors, a global app store, foreign payment processors and international CDN performance. The ArvanCloud relationship becomes harder to justify. The same infrastructure can be an asset or a liability depending on the customer's capital market and user geography.
This is why a local cloud can be worth more than a simple hosting business without being universally attractive. The value is conditional. ArvanCloud is most valuable where the customer's world is domestic, payments are local, latency matters, sanctions are already a fact of life, and developer trust can be won through performance and support. It is least valuable where the customer needs clean foreign payments, global investor confidence, cross-border enterprise sales or a sanctions-neutral supplier chain.
Facility dependency is another missing fact. ArvanCloud's status page lists named regions and products, and its network map lists many CDN cities. But public pages do not fully disclose which facilities are owned, leased, partner-operated or dependent on third-party data-center and upstream contracts. Softqloud's statement explicitly denied operating data centers and said it rented cloud server space for customers, especially from third parties in the Netherlands (https://softqloud.com/statement.html). That detail matters because a provider's margin and resilience are different if it owns core facilities, leases cages, rents servers, uses partner PoPs or relies on upstream carriers for critical regions.
The cost of trust is therefore embedded in contracts the public cannot see: facility leases, upstream transit, exchange ports, hardware procurement, software licenses, bank accounts, enterprise customer agreements, legal-request handling and support staffing. Public evidence can show the contours. It cannot underwrite the balance sheet.
Evidence register
The identity evidence rests on PeeringDB, RIPE RDAP and company statements. PeeringDB lists Softqloud GmbH, AS208006, traffic of 20-50Gbps, two DE-CIX India exchange connections and a COMNET Istanbul facility, with ArvanCloud as the website (https://www.peeringdb.com/net/26232 and https://www.peeringdb.com/api/net/26232). RIPE RDAP names AS208006 as ARVANCLOUD-CDN and links it to ARVANCLOUD GLOBAL TECHNOLOGIES L.L.C. in Dubai (https://rdap.db.ripe.net/autnum/208006). ArvanCloud's own termination notice and Softqloud's statement describe a terminated commercial partnership around international customers, suppliers, cloud server space and CDN services (https://news.arvancloud.ir/en/contract-termination/ and https://softqloud.com/statement.html).
The product and pricing evidence comes from ArvanCloud's public pages: CDN pricing, support pricing, general pricing method, cloud server, managed database, network map, IP ranges, looking glass and status page (https://www.arvancloud.ir/en/pricing/cdn, https://www.arvancloud.ir/en/pricing/support, https://www.arvancloud.ir/en/pricing, https://www.arvancloud.ir/en/products/cloud-server, https://www.arvancloud.ir/en/products/databases, https://www.arvancloud.ir/en/pricing/databases, https://www.arvancloud.ir/en/products/cdn/network, https://www.arvancloud.ir/en/dev/ips, https://lg.arvancloud.ir/ and https://www.arvancloudstatus.ir/).
The sanctions and legal-access evidence comes from U.S. Treasury, OFAC, the UK Sanctions List, RFE/RL and Carnegie. Treasury's June 2, 2023 press release and OFAC recent action describe the U.S. designation and wind-down license (https://home.treasury.gov/news/press-releases/jy1518 and https://ofac.treasury.gov/recent-actions/20230602). OFAC FAQs 441 and 1088 define part of the communications-services authorization context (https://ofac.treasury.gov/faqs/441 and https://ofac.treasury.gov/faqs/1088). The UK listing and EU-delisting coverage show the sanction landscape is not uniform (https://search-uk-sanctions-list.service.gov.uk/designations/IRN0165/Entity, https://www.rferl.org/a/iran-eu-sanctions-arvancloud-internet-censorship/32897604.html and https://carnegieendowment.org/emissary/2024/05/hard-questions-about-the-eu-lifting-sanctions-on-an-iranian-tech-company).
The market and trust evidence comes from ArvanCloud's about page, careers page, transparency report, Terraform Registry/GitHub and limited review-platform snippets (https://www.arvancloud.ir/en/about, https://www.arvancloud.ir/fa/jobs, https://www.arvancloud.ir/en/legal/transparency-report, https://registry.terraform.io/providers/arvancloud/arvan/latest, https://github.com/arvancloud/terraform-provider-arvan, https://www.g2.com/products/arvancloud/reviews and https://www.g2.com/sellers/arvancloud). These signals are useful for product maturity and developer posture, but they are weaker than contracts, audited accounts or customer-retention data.
What would change the judgement
The judgement today is conditional but clear. Softqloud/ArvanCloud is more economically important than a simple hosting business because it sits at the point where Iranian domestic demand, sanctions, payment access, CDN reach, local support and developer trust meet. For domestic workloads, that intersection creates a real premium. For foreign-facing workloads, the same intersection creates a discount.
Several facts would change the judgement quickly. The first is sanctions and payments. A current OFAC delisting, specific license, bank comfort letter or payment-processor approval covering ArvanCloud-related transactions would raise external value. A new designation, enforcement action or bank de-risking notice would lower it. A clear document showing Iranian customers can reliably buy foreign cloud/CDN services under the communications authorization without account closure would also lower ArvanCloud's local scarcity premium.
The second is customer mix. If ArvanCloud can show recurring private-sector revenue, low churn, limited customer concentration, strong enterprise attach, clean abuse metrics and reliable collection, the business looks like a durable domestic cloud platform. If revenue depends heavily on politically exposed accounts, free-tier traffic or subsidized demand, it looks less like a cloud company and more like strategic infrastructure with uncertain commercial margin.
The third is facility and supplier control. Owned or long-term controlled capacity in key Iranian regions, documented redundancy, stable upstream contracts and transparent incident postmortems would support the sovereign-cloud thesis. Heavy reliance on fragile third-party facilities, hard-to-renew upstreams or sanctioned suppliers would make the platform more vulnerable than its marketing suggests.
The fourth is developer trust. ArvanCloud's public tooling, status page, support plans and transparency report are all attempts to build it. Trust would improve if engineers saw consistent uptime, quick postmortems, active Terraform/API maintenance, credible legal-request handling and a product roadmap that survives politics. It would weaken if the provider's tooling stagnates, incidents become opaque, or market chatter turns from skepticism into migration.
The economic answer is therefore not that Softqloud is "good" or "bad", nor that ArvanCloud is just another host. The better answer is that a local cloud in a sanctioned market can be valuable precisely because the outside option is impaired. ArvanCloud's public evidence shows enough product breadth, network surface and operational machinery to matter. It also shows enough sanctions, resource-chain and trust risk to prevent a clean underwriting case. The buyer's decision is not simply whether the server works. It is whether the workload is better off inside a constrained local cloud than outside it, fighting for foreign access that may not clear payment, policy or politics.