BareMetal.com Inc and the Economics of Infrastructure That Refuses to Disappear
BareMetal.com Inc is not important because it is a hyperscale cloud company. It is important because it is the opposite kind of infrastructure firm: small, old, operationally specific, hard to classify, and commercially revealing. Its public record shows a Canadian hosting and domain-registration operator that began in the mid-1990s, incorporated in 2000, served long-tail web and domain customers, relied on upstream datacenter and network providers, and eventually sold the high-value BareMetal.com domain name while continuing under BareMetal.ca. That sequence is a compact case study in the economics of dedicated infrastructure after the first web-hosting era.
The central finding is that BareMetal.com Inc illustrates how hosting economics are shaped less by visible brand scale than by scarce operational assets: IPv4 address assignments, registrar credentials, DNS control, customer billing relationships, long-lived web workloads, upstream datacenter access, and accumulated trust. The company’s name aged into a market category. “Bare metal” now denotes dedicated physical servers consumed with cloud-like APIs, but BareMetal.com Inc’s own public pages describe a broader Canadian web-hosting and domain-registration business whose durability came from support, continuity and low churn rather than from a venture-scale cloud platform. Its sale of the .com name to a successor context connected to i3D.net crystallizes the commercial point: the old host survived, but the most liquid asset became the category-defining domain.
There is also an evidentiary correction. The directory-row clue identifies the target as BareMetal.com Inc, but the delegated-stats ASN 138282 clue does not publicly resolve to BareMetal.com Inc. Public APNIC-derived records identify AS138282 as DMRC-AS, associated with Delhi Metro Rail Corporation Limited in India, not with BareMetal.com Inc. The commercially relevant network trail for BareMetal.com Inc is instead ARIN entity BAREM-2 and a reassigned IPv4 /24, 67.223.102.0/24, inside Priority Colo’s broader 67.223.96.0/20 and AS30176 routing environment. That distinction matters because it changes the company’s economic interpretation: BareMetal appears not as an autonomous network owner, but as a hosting and registrar operator dependent on a specialist colocation and network provider.
The company behind the name
BareMetal’s official history describes a company that came out of consulting work around Internet connectivity and UNIX support, then transitioned into web and hosting services in 1995. The business says its first clients were friends, and it frames its operating ethic around treating customers as friends. The proprietorship was converted into BareMetal.com Inc in 2000. That is consistent with the Better Business Bureau profile, which lists the business start and incorporation date as June 29, 2000, identifies the legal form as a corporation, lists BareMetal.com Inc as an alternate name, and names Tom Brown as owner; BBB also states that it does not verify all third-party information, so its management entry should be treated as secondary evidence rather than a corporate registry record.
The company’s current public presence has moved from the .com to the .ca domain. BareMetal.ca states that the business is a Canadian web-hosting and domain-registration specialist “since 1995.” More unusually, it also states that as of May 2026 the company is “basically done” transitioning from BareMetal.com to BareMetal.ca, that it has “100% Canadian ownership, staff, and physical presence,” and that “the .com has been sold,” with proceeds expected to “pay off some debts.” That one sentence is unusually information-rich. It confirms continuity of the operating company, separation between the company and the premium domain asset, Canadian operational identity, and a financial motivation for asset sale rather than merely a marketing rebrand.
The name itself predates the modern “bare-metal cloud” market. BareMetal’s about page traces the phrase to the Hacker Jargon File and says it refers to work “at the hardware level” and to building systems from scratch. The company interprets the name as implying depth, clean systems and attention to detail. In the 1990s this was a credibility signal for technical web hosting: the provider understood UNIX, servers, connectivity and hands-on infrastructure. In the 2020s the same phrase has become a product category for dedicated physical servers consumed like cloud infrastructure. That semantic shift is central to the company’s commercial story.
The evidence does not support treating BareMetal.com Inc as a modern bare-metal cloud operator in the same sense as i3D.net, OVHcloud, Equinix Metal’s former category, or automated dedicated-server platforms. BareMetal’s own product pages emphasize web hosting, virtual servers, secure web services, domain registration, DNS, mail services, SQL databases, logs and related hosting features. Its commercial promise is not “global API-provisioned metal,” but support, affordability, registrar services and continuity. That makes the company more analytically useful, not less. It is a survivorship case from the first hosting era, when customers bought websites, domains, email, CGI/PHP/MySQL-style hosting and trust in a small operator, not elastic compute primitives.
A business model built from small frictions
BareMetal’s web-services page says the company specializes in web hosting, customer service and affordability, explicitly contrasting its approach with “fast/easy profit.” It lists virtual servers, domain registration, secure web services, redundant servers, virtual FTP, mailing lists, SQL databases, logs and related services. It also states that its rate structure is competitive, that bulk rates are available, and that surcharges apply only to very busy sites. This is a classic long-tail hosting model: low average revenue per account, low-touch infrastructure, support-heavy retention, and profitability dependent on packing many small customers onto shared systems while avoiding catastrophic support load.
The domain-registration side adds a second revenue layer. BareMetal says it has been a CIRA-certified .ca registrar since 2000 and an OpenSRS reseller for other domains since 1999 or 2000. It offers .CA, .US, .COM, .NET, .ORG, .INFO and .BIZ registration, optional DNS, web forwarding, parking, email forwarding and related domain services. CIRA’s registrar model places registrars between registrants and the .CA registry, which means the commercial relationship belongs to the registrar even when the registry controls the authoritative zone. BareMetal’s domain page also emphasizes .CA TBR, or drop-catching/backorder participation, which points to a niche customer base beyond ordinary small-business hosting: domain investors, Canadian registrants and customers who value a registrar with operational knowledge of expiry cycles.
This bundling matters economically. A customer with a domain registered through BareMetal, DNS hosted by BareMetal, web hosting on BareMetal, email forwarding or mail services configured through BareMetal, and billing records in BareMetal’s customer portal faces more than a price comparison when considering migration. Moving the account requires domain transfer authorization, DNS-zone export or reconstruction, MX and SPF/DKIM/DMARC review, web-content transfer, database dump and import, SSL changes, mail migration, cron/script compatibility checks and a billing cutover. None of these frictions is insurmountable, but together they produce inertia. For a small business, club, charity, legacy content site or owner-operated domain portfolio, the rational decision may be to keep paying a familiar provider even if a cheaper commodity host exists.
BareMetal’s terms allocate operational responsibility in a way typical of hosting. Customers must maintain accurate contact information and are responsible for backups and acceptable use. BareMetal reserves immediate termination rights for security, stability or integrity concerns. The service-credit language is narrow: credits for uptime failure are subject to exceptions and depend on BareMetal’s equipment or mistake being the direct cause. The registration agreement also contains liability limits, including language that BareMetal is not liable for substitute services and caps certain liability at $500. The commercial implication is clear: the provider sells continuity and support, but it does not underwrite the full economic value of a customer’s website, domain, email archive or business interruption risk.
This is one reason old hosting relationships persist. The customer is formally responsible for its own backups, security choices and account information, but informally depends on the host’s institutional memory. The host may know which customers have ancient scripts, which zones are fragile, which email forwarders are still used, which billing contacts are stale and which domains are business-critical. That knowledge is not visible on a balance sheet, but it is operational capital. It also has an asymmetric value profile: it is worth little to outsiders until something breaks, and then it can be decisive.
The thin public footprint is itself evidence
BareMetal.com Inc does not have the public footprint of a high-growth infrastructure company. There is no visible investor-relations presence, no hyperscale datacenter map, no public cloud marketplace listing, no broad executive bench on the public web, and no dense stream of engineering job postings discovered in the ordinary public record. Instead, the footprint consists of official company pages, registrar records, hosting terms, an account portal, WHOIS/RDAP traces, customer and forum traces, and third-party hosting directories.
That absence should not be read as failure. In infrastructure economics, opacity often reflects company type. A small profitable or near-profitable hosting operator can exist for decades without press releases. Its market is not public narrative; it is renewals. Its core question is not whether it can win net-new enterprise procurement against Amazon Web Services or Microsoft Azure. It is whether enough customers keep renewing domains and hosting plans, whether the servers remain patched and paid for, whether upstream suppliers remain stable, and whether support burden stays below the contribution margin of the account base.
Public reputation traces are sparse. The BBB profile shows zero customer reviews and zero complaints on the pages checked, but that is not proof of either quality or scale; it is simply evidence that this channel has not accumulated visible disputes. A domain-forum profile identifies a BareMetal.com service representative active in Canadian domain discussions, and forum discussions show BareMetal being mentioned by users as a Canadian hosting or registrar option. These are unofficial traces, but commercially they suggest a business that has remained visible in the Canadian domain community even as the broader hosting market consolidated.
Hosting-review traces are also thin and should be weighted lightly. One hosting-review listing described BareMetal.com as offering shared hosting and domain registration, with an old review referencing “cloud” and “bare-metal” language, but the sample is too small to infer customer satisfaction or revenue. A NamePros forum discussion contains user comments about BareMetal as a registrar/hosting provider, but again this is anecdotal. The information gain is not statistical; it is categorical. BareMetal existed in the community memory of domain and hosting users, not only in official corporate pages.
Address resources and the scarcity premium
The most concrete infrastructure evidence is the IP assignment. Public ARIN-derived records identify 67.223.102.0/24 as a reassigned netblock for BareMetal.com Inc, with organization ID BAREM-2 and a Victoria, British Columbia address. The parent block is Priority Colo’s 67.223.96.0/20. IPinfo-style third-party records identify hostnames such as redir.baremetal.ca within this range and show the larger route 67.223.96.0/20 originated by AS30176, Priority Colo. The RPKI status is reported as valid for the routed aggregate.
Commercially, this is a different profile from a company that owns and originates its own portable IP space. A reassigned /24 gives BareMetal operational address capacity, but the routing, RPKI and upstream dependencies appear tied to Priority Colo’s network. That reduces the burden of operating an autonomous system, maintaining transit contracts and handling global routing policy. It also reduces strategic autonomy. If BareMetal wants to move the entire service stack to another facility or network, address portability may be constrained by contract and by the fact that the public route is Priority Colo’s aggregate, not necessarily BareMetal’s own independently originated prefix.
IPv4 scarcity changes the economics of even small hosting firms. A /24 contains 256 addresses, fewer after network, gateway, broadcast or operational reservations. In a shared-hosting model, many domains can sit behind a single IPv4 address using name-based virtual hosting. In SSL’s early years, dedicated IP addresses were often needed for separate certificates; Server Name Indication reduced that pressure, but not every legacy stack or customer assumption disappeared. Mail reputation, custom DNS, reverse DNS, reseller isolation, abuse containment and customer expectations can still create demand for distinct addresses. A provider with a clean, historically used block has an asset-like operating advantage, even if it does not own the block in a transferable capital-market sense.
Address scarcity also affects customer discipline. Dedicated hosting customers tend to ask for static addresses, reverse DNS, separate test environments, VPN endpoints, monitoring addresses and sometimes multiple IPs for reasons that are technically weak but commercially familiar. The provider must ration. Every additional customer IP has an opportunity cost. In the post-IPv4-exhaustion world, the address pool is no longer an incidental technical input; it is a gating factor for product design and abuse risk.
The AS138282 clue is important because it demonstrates how easy it is to misread infrastructure evidence. Public APNIC-derived records place AS138282 with Delhi Metro Rail Corporation Limited in India. If one mechanically matched the directory row to the delegated ASN, one would infer a non-existent BareMetal autonomous network in Asia. The correct economic conclusion is the opposite: BareMetal’s visible address-resource trail points to a Canadian hosting operator using reassigned space under Priority Colo. The false lead increases confidence in the need for entity-level reconciliation across RIR, RDAP, routing and company evidence before drawing commercial conclusions.
Datacenter dependence and the Toronto anchor
BareMetal’s corporate and support identity is Canadian and Victoria-based, but the infrastructure trail points toward Priority Colo and Toronto network infrastructure. Priority Colo describes itself as offering colocation, dedicated servers and related services on redundant infrastructure, with operations in Greater Toronto Area facilities since 2002. Its public materials emphasize a small team, colocation, datacenter economics, network administration experience and scale economies.
Priority Colo’s network page describes the relevant datacenter inputs: HVAC, fire systems, floor loading, security, multiple paths, transit and transport. It lists a BGP mix including Tata Communications, Level 3, TiNET, Cogeco Data Services and TorIX, with additional transit via Zayo. It also references redundant Layer 2 paths back to 151 Front Street and a building meet-me-room ecosystem with more than 150 carriers. PeeringDB identifies Priority Colo as AS30176, with North American traffic, open peering policy, IPv4 and IPv6 prefix counts, and public peering details. BGP.tools lists AS30176 as an ARIN-allocated network registered in 2003, with upstreams including Lumen, Tata and Hurricane Electric, and shows 67.223.96.0/20 among valid originated IPv4 space.
This supplier relationship is economically rational. A small host does not need to own a datacenter to sell hosting. It needs reliable rack space, remote hands, power, cooling, IP transit, DDoS posture, routing competence and emergency response. Buying those inputs from a colocation/network specialist converts fixed capital expenditure into a recurring supplier cost. It allows the hosting company to focus on customer accounts, registrar operations, server administration and support. But it also creates dependency. If the upstream facility, routing provider, or datacenter contract changes, the host’s customer-facing continuity is exposed.
Toronto is a natural infrastructure anchor for such a Canadian operator. 151 Front Street West is one of Canada’s key carrier-hotel locations, and the Cologix TOR1 PeeringDB facility page lists the 151 Front Street address and a dense exchange ecosystem, including TorIX. Priority Colo is listed among networks at that facility. For a Victoria-based company serving Canadian and international web customers, this topology makes sense: business administration and support can be on Vancouver Island, while servers sit near dense interconnection in Toronto.
The separation between corporate location and server location is not a footnote. It is the business model. Infrastructure firms arbitrage geography. The customer may buy “Canadian hosting” from a Victoria company, while the packets flow through Toronto, transit providers and exchange fabrics. The customer’s legal comfort, support relationship and currency may be local; the infrastructure’s efficient physical location may be wherever power, cooling and network density are best.
Dedicated-server supply and the old hosting cost curve
Bare-metal hosting, in the literal modern sense, means a physical server dedicated to one tenant. OVHcloud and other providers describe bare metal as single-tenant physical infrastructure rented to customers without the virtualization layer that characterizes much cloud computing. i3D.net’s current bare-metal cloud page says customers run directly on dedicated servers with no hypervisor and no noisy neighbors, while consuming capacity through cloud-like tooling such as APIs, Terraform, BMC access and on-demand or committed models.
The economics are straightforward but unforgiving. A dedicated server has high fixed cost and lumpy capacity. The provider buys or leases hardware, installs it, powers it, cools it, monitors it, connects it, replaces failed drives, handles abuse, and eventually retires or redeploys the asset. Unlike a virtualized cloud node, a dedicated box cannot be efficiently sliced among many unrelated tenants unless the provider itself becomes a virtualization provider. If a customer leaves after three months, the provider may be left with stranded hardware configured for a workload that no longer exists. If the customer stays for years, the server can become highly profitable after capital recovery, provided power, support and failure rates remain controlled.
This is why old hosting companies often prefer simple products and long customer tenure. A shared-hosting server can support many low-traffic sites. A domain-registration account can renew for years with minimal intervention. DNS, email forwarding and parking services are lightweight. The economics improve when the provider avoids bespoke enterprise obligations, aggressive service-level penalties and high-touch migrations. A small host can survive by being boring: stable prices, stable nameservers, stable support, and few forced platform changes.
The same mechanics explain the modern bare-metal cloud market. The customer wants physical isolation, predictable performance, direct hardware access and network quality. The provider wants committed capacity so hardware is not stranded. i3D’s page makes this explicit by distinguishing committed capacity from on-demand hourly usage. It says committed bare-metal capacity is equivalent to traditional dedicated bare metal for predictable long-term workloads, while on-demand servers support rapid deployment and burst use. It also says each server includes one public IPv4 address, traffic bundle, redundant uplink and premium network. The inclusion of only one IPv4 address is not incidental; it is an economic signal about address scarcity.
BareMetal.com Inc’s public footprint sits on the older side of this curve. Its pages do not present a global automated dedicated-server fabric. They present a hosting and registrar operator with long-lived services. But the commercial mechanisms are continuous with the modern market. Physical servers are capital assets. IP addresses are scarce. Customer migrations are risky. Datacenters are specialized suppliers. Network quality is purchased through upstream relationships. Support and trust determine retention. The difference is scale and packaging.
Customer lock-in without monopoly power
BareMetal’s apparent customer lock-in is not monopoly lock-in. Customers can transfer domains, change DNS, move websites and switch email providers. The lock-in is procedural and risk-based. The smaller and older the customer, the stronger the friction.
A domain may be the login identity for a bank account, a government portal, a nonprofit donation system, an old e-commerce checkout, a Google Workspace tenant or a mailing list. DNS records may have been added over many years by different vendors. The website may run on legacy PHP, old Apache behavior, hardcoded paths or an abandoned CMS. Mail may depend on forwarding rules nobody has documented. The business owner may not know which account controls the registrar, which account controls DNS, and which account controls hosting. In that environment, paying a familiar host is a rational insurance premium.
BareMetal’s account portal underscores the billing relationship. It exposes account login, billing ID concepts and invoice examples associated with BareMetal.com Inc. The company’s contact page lists support email, phone numbers and business hours, while also noting the mailing address is not a business office. This is the support model of a small infrastructure operator: identifiable, reachable, but not built around walk-in retail or enterprise field sales.
Domain registration adds a second lock-in mechanism: procedural timing. Expired domains, auto-renew grace periods, redemption periods, transfer locks and registry rules all create moments when the registrar’s competence matters. In Canadian domain forums, a BareMetal service-representative account has discussed .CA expiry and auto-renew-period behavior. This is unofficial forum material, not a formal policy source, but commercially it shows why niche registrar knowledge can retain customers. A registrar that understands edge cases around expiry, TBR and registry states can be valuable even without broad consumer brand recognition.
The lock-in is also emotional in the narrow business sense. BareMetal’s about page explicitly says the company’s first clients were friends and that its ethic is to treat clients that way. In a mass-market hosting environment this might sound quaint. In the long tail, it is a retention strategy. Small customers do not run formal RFPs for a $15 domain or modest hosting plan. They renew with the provider that answers the phone, remembers the account, and does not break the old site.
Ownership ambiguity and what the .com sale means
The official evidence supports a narrow ownership conclusion: BareMetal.ca presents the continuing operator as Canadian-owned, Canadian-staffed and physically Canadian, while saying the BareMetal.com domain has been sold. It does not say the company was sold. It does not disclose the buyer, price, debt amount, asset-purchase agreement, transition obligations or whether any customers were transferred. BBB’s secondary profile identifies Tom Brown as owner, but that should not be treated as a current corporate registry substitute.
The present commercial use of BareMetal.com points to i3D.net. The BareMetal.com URL redirects to i3D.net’s Bare Metal Cloud page. i3D describes a global bare-metal cloud offering with committed and on-demand capacity, direct server access, automation, private VLANs, anti-DDoS, global points of presence and custom BGP-session support. i3D’s own corporate page states that Ubisoft acquired i3D.net in 2018, that i3D continued to operate independently, that it kept serving external customers, and that the acquisition produced cost savings for Ubisoft workloads that otherwise would have been hosted by public cloud providers.
This is a clean example of domain-asset repricing. For BareMetal.com Inc, the .com was the legacy identity of a Canadian host. For i3D, “baremetal.com” is a category domain aligned with a global product. The same string can have low operating utility for a small incumbent after a transition to .ca, but high customer-acquisition utility for a global infrastructure seller trying to capture search and direct-navigation demand for bare-metal servers. The buyer does not need the old company’s servers to value the asset. It needs the generic term, the type-in traffic, the search credibility, and the semantic fit.
The sale also reveals financial pressure. BareMetal.ca says the .com sale will “pay off some debts.” That disclosure is rare among small hosting firms and commercially meaningful. It suggests the domain was monetized not merely because it was surplus but because it could improve the balance sheet. In small hosting economics, debt can come from hardware refresh cycles, accumulated tax or supplier obligations, operating losses, owner financing, or acquisition of intangible assets. The public evidence does not identify the debt source, so the prudent conclusion is limited: the domain sale was financially material to the continuing company.
For the successor context, i3D’s network scale changes the economics. i3D’s public pages describe bare-metal capacity as part of a global low-latency network with automation, private networking and anti-DDoS. PeeringDB and i3D’s peering page present AS49544 as an actively peered network with formal public policy, NOC requirements and multi-geography peering expectations. This is not the same business as small shared hosting. It is infrastructure for latency-sensitive workloads such as gaming, interactive services and distributed applications.
Market structure: from web hosts to infrastructure platforms
BareMetal.com Inc belongs to a generation of hosting companies that emerged when domain names, DNS, mail, FTP, UNIX support and web hosting were bundled. In that market, the provider’s advantage came from competence and continuity. The customer did not buy primitives; the customer bought “my website works,” “my domain renews,” and “someone answers when something breaks.”
The modern infrastructure market has separated and recombined those layers. Registrars became a scale business and a domain-investor niche. DNS became both a commodity and a premium security service. Email moved to Google, Microsoft and specialized providers. Web hosting split into shared hosting, managed WordPress, VPS, cloud infrastructure, serverless and managed platforms. Dedicated servers became “bare metal,” then “bare metal cloud,” then part of hybrid cloud, AI infrastructure, private cloud and gaming infrastructure narratives.
This transition creates pressure on small hosts. They lose the high-growth customer cohorts to cloud platforms and SaaS builders. They face security and compliance burdens that did not exist in the 1990s. Hardware refreshes become less forgiving. IPv4 addresses become more expensive. Customers expect HTTPS, modern control panels, spam filtering, two-factor authentication, automatic backups and instant provisioning. Meanwhile, the host cannot easily raise prices on old customers without increasing churn or support tickets.
But small hosts retain advantages where trust, jurisdiction and continuity matter. A Canadian customer may prefer a Canadian registrar and host. A domain investor may value a registrar that understands .CA TBR. A legacy site owner may value non-disruptive service over modern features. A small organization may prefer phone support to a global cloud ticket queue. This is not a high-growth market, but it can be a cash-flow market if managed conservatively.
BareMetal’s official pages show exactly this posture. The company emphasizes affordability, customer service and a long history. It does not present itself as the cheapest possible mass host or the most advanced global cloud. The operating thesis appears to be continuity. In infrastructure, continuity can be a strategy when the installed base is sticky enough and the cost base is disciplined.
Infrastructure survivorship
The word “survivorship” is not decorative here. Hosting companies die in several ways. They can be acquired and absorbed. They can lose routing or datacenter access. They can fail to patch systems and suffer security collapse. They can lose registrar accreditation or upstream reseller relationships. They can lose key personnel. They can become uneconomic when too many customers leave but too much infrastructure remains. They can sell the domain and disappear.
BareMetal has not disappeared. The .com brand asset has been sold, but BareMetal.ca continues to present the operating business, account systems, hosting services, domain services and Canadian identity. That is a different outcome: asset monetization plus operational continuation.
Infrastructure survivorship depends on reducing the number of things that can kill the company at once. A host that owns a datacenter carries real-estate, power and facility risk. A host that outsources colocation carries supplier risk. A host that owns portable address space has routing autonomy but must manage routing complexity. A host that uses reassigned space has less routing burden but more provider dependence. A registrar has recurring revenue but must comply with registry obligations. A domain reseller has less compliance burden but less control. Each choice trades autonomy for cost structure.
BareMetal’s observed choices lean toward survivorship through specialization and dependency management. It appears to rely on Priority Colo for core network/datacenter routing inputs, while retaining customer, registrar and hosting relationships. It sells domain services and web hosting rather than trying to build a global cloud. It moved from .com to .ca rather than abandoning the brand entirely. It monetized the generic category domain when the market value of that name likely exceeded its operating value to the legacy host.
That is an economically coherent path for an old infrastructure firm. It is not the same as growth. It is adaptation under constraint.
What would change the commercial view
The public evidence is thin enough that several unresolved facts would materially change the valuation or interpretation.
The first is active customer count. A registrar/hosting company with a few hundred active accounts is a lifestyle or wind-down business. One with tens of thousands of domains under management is a more valuable renewal base. Public pages show service offerings and continuity, but not active domains under management, hosting accounts, revenue or churn.
The second is the composition of revenue. Domain registrations generate predictable but low-margin renewals. Shared hosting can be higher margin if support is low and servers are depreciated. Dedicated or colocated servers can produce higher monthly revenue but require more operational work and capital discipline. The public evidence supports domain and web-hosting activity, but not the current mix of shared hosting, VPS, dedicated servers, colocation resale or pure registrar revenue.
The third is the Priority Colo relationship. If BareMetal owns hardware in Priority Colo space, the economics differ from reselling a managed service. If BareMetal leases servers, margins and migration flexibility differ again. If the 67.223.102.0/24 reassignment is contractually stable, the host has more continuity. If it is easily revoked or non-transferable, the migration risk is higher.
The fourth is the .com sale structure. A simple domain-name sale is one thing. A sale that included traffic redirection, customer referrals, transition assistance, trademark covenants or non-compete terms is another. BareMetal.ca’s statement confirms the .com was sold and debt repayment was a use of proceeds, but not the transaction structure.
The fifth is security and platform modernization. A legacy hosting company’s value can be impaired quickly by outdated control panels, old operating systems, unpatched web stacks, weak account security or mail-reputation problems. Third-party traces of hosted sites and old server headers can suggest legacy infrastructure, but they are not enough to judge current internal patching or security. The key question is whether the company can keep old customers stable without letting old software become systemic risk.
The economics revealed by BareMetal
BareMetal.com Inc reveals five mechanisms that are easy to miss in larger cloud-company analysis.
First, infrastructure brands can outlive their original product meaning. BareMetal chose its name to signal low-level technical competence. The market later turned “bare metal” into a standardized product category. The company’s domain then became more valuable to a global bare-metal seller than to the original small host. That is brand-option value created by semantic drift.
Second, IPv4 scarcity creates hidden balance-sheet effects. BareMetal’s visible /24 reassignment is not a massive asset in hyperscale terms, but it is operationally meaningful. It enables hosting continuity, mail, DNS and customer isolation. Whether owned, reassigned or bundled with colocation, address capacity affects product design and bargaining power.
Third, customer lock-in in hosting is mostly procedural. It comes from DNS, email, registrar status, old scripts, billing relationships and fear of downtime, not from formal exclusivity. This kind of lock-in is weaker than monopoly power but stronger than a commodity-price comparison.
Fourth, datacenter dependence is a rational substitute for vertical integration. A small host can buy network and facility quality from a provider such as Priority Colo rather than building it. The cost is dependence on the supplier’s routing, facility and commercial continuity.
Fifth, infrastructure survivorship can be economically valuable without being newsworthy. A company that keeps renewing customers, answering support, maintaining domains and paying suppliers may produce little public evidence. That quietness is not the absence of economics. It is the form taken by a mature, long-tail infrastructure business.
BareMetal.com Inc therefore matters less as a standalone investment target than as a diagnostic case. It shows how the first-generation web-hosting stack decomposed into assets: customer relationships, registrar channels, IP assignments, supplier contracts, domain names and operational know-how. Some of those assets remain with the continuing Canadian operator. One, the BareMetal.com name, appears to have moved into a global bare-metal cloud context. The economics of dedicated infrastructure are visible in that split.
Evidence ledger
BareMetal official company pages: BareMetal.ca identifies the business as Canadian web-hosting and domain-registration specialists since 1995, states that the transition from BareMetal.com to BareMetal.ca was essentially complete by May 2026, says the .com was sold, and says sale proceeds would help pay debts. BareMetal’s about page describes the company’s origin in Internet connectivity and UNIX support, its move into web/hosting in 1995, and incorporation as BareMetal.com Inc in 2000.
BareMetal services and terms: The company’s web-services page lists virtual servers, domain registration, secure web services, redundant servers, FTP, mailing lists, SQL databases and logs. The domain page says BareMetal has been a CIRA-certified .ca registrar since 2000 and an OpenSRS reseller for other domains since 1999 or 2000. The terms and registration agreement assign backup responsibility to customers, reserve termination rights for security and stability concerns, and limit certain liability.
Contact and account evidence: BareMetal’s contact page lists support email, phone numbers, weekday Pacific support hours and a Victoria mailing address that is not a business office. The account portal shows continuing account-login and billing-ID infrastructure associated with BareMetal.com Inc.
Corporate secondary evidence: The BBB profile lists BareMetal.com Inc as a corporation, gives a June 29, 2000 start/incorporation date and names Tom Brown as owner, while BBB’s own disclaimers mean this should be treated as secondary, not definitive corporate-registry proof. BBB review and complaint pages showed no visible reviews or complaints in that channel.
RIR, WHOIS, BGP and address evidence: Public ARIN-derived WHOIS/RDAP records identify BareMetal.com Inc, org ID BAREM-2, as the reassigned organization for 67.223.102.0/24 inside Priority Colo’s parent 67.223.96.0/20. Third-party IP intelligence records connect BareMetal hostnames to that range and show routing through AS30176/Priority Colo with valid RPKI at the aggregate. BGP.tools identifies Priority Colo’s AS30176, upstreams and originated routes.
ASN 138282 correction: Public APNIC-derived and BGP directory records identify AS138282 as DMRC-AS, Delhi Metro Rail Corporation Limited, India. This does not support attributing AS138282 to BareMetal.com Inc.
Priority Colo and datacenter evidence: Priority Colo’s official pages describe colocation, dedicated servers, related services, GTA facilities and network infrastructure. Its network page lists transit, transport and peering elements including TorIX and major carriers. PeeringDB identifies AS30176 as Priority Colo with public network attributes. PeeringDB’s Cologix TOR1 facility page places Priority Colo in the 151 Front Street ecosystem.
Successor and category-domain evidence: BareMetal.com redirects to i3D.net’s Bare Metal Cloud page, which describes dedicated physical servers consumed with cloud-like automation, committed and on-demand models, one included public IPv4 address per server, redundant uplinks and global network features. i3D states that Ubisoft acquired i3D.net in 2018 while i3D continued operating independently and serving external customers. i3D’s peering material and PeeringDB profile describe a much larger global network context than BareMetal’s visible footprint.
Unofficial market traces: Domain and hosting forums, hosting-review pages and user comments mention BareMetal as a Canadian hosting or registrar provider. These are low-weight sources and not reliable for revenue, customer satisfaction or current operating scale. Their commercial value is limited to showing community visibility and category perception.
Watchpoints
The first watchpoint is whether BareMetal.ca remains the active operating identity after the .com sale. Continued account-portal availability, nameserver continuity, support responsiveness and domain-renewal operations would indicate a stable post-sale transition. Any prolonged outage, nameserver change or registrar-status change would be commercially significant.
The second watchpoint is the fate of 67.223.102.0/24. If the reassigned BareMetal range is withdrawn, renumbered, transferred, or absorbed into a different hosting stack, that would indicate either migration, contraction or supplier restructuring. If it remains stable behind Priority Colo, the continuity thesis strengthens.
The third watchpoint is CIRA registrar status and .CA TBR activity. BareMetal’s registrar role is a core differentiator from a generic legacy web host. Loss of registrar status, reduced TBR visibility or migration to pure reseller status would lower the strategic value of the customer base.
The fourth watchpoint is customer migration after the BareMetal.com sale. The old .com domain had identity value, search value and trust value. A smooth move to BareMetal.ca would show that customer relationships are stronger than the domain. Customer confusion, lost renewals or support escalation would show that the .com was more operationally important than the company’s transition language implies.
The fifth watchpoint is i3D’s use of BareMetal.com. If i3D continues to use the domain as a category landing page, the transaction should be read as customer-acquisition and search-intent capture. If i3D later builds BareMetal.com into a separate brand, marketplace or self-serve dedicated-server portal, the acquired domain’s value would be higher than a simple redirect suggests.
The sixth watchpoint is ownership disclosure. BareMetal.ca’s public statement establishes Canadian ownership and the sale of the .com, but not the company’s current shareholding, the domain-sale buyer, the price, or any covenants. Corporate-registry filings, litigation, secured-creditor filings or domain-transaction disclosures would materially improve the commercial picture.
The seventh watchpoint is modernization risk. Legacy hosting firms can survive for decades, but the cost of security, mail deliverability, SSL automation, two-factor authentication, backup expectations and software patching rises over time. A visible platform refresh would support long-term survivorship. Stagnation would raise the probability that the remaining value is mainly domains, renewals and goodwill rather than durable hosting infrastructure.