DataCenter.BZ and the price of local control in Columbus colocation

The local buyer's first question is control

Imagine a managed-service provider in Ohio with a familiar problem. A healthcare client wants predictable disaster recovery. A school district wants servers outside its own building but still close enough for physical visits. A manufacturer has a line-of-business application that does not deserve a national cloud migration project, but also cannot live under a desk or in a closet. The provider can resell capacity from a hyperscale cloud region, or it can put the customer's equipment in a local data center where the provider knows the engineers, the cross-connect path, the loading dock, the power rules, and the real escalation chain.

That is the practical frame for DataCenter.BZ, LLC. The company is not important because it became a global brand. It is important because it shows how a regional colocation operator could become a control layer for smaller enterprises, hosters, public-sector bodies, schools, healthcare systems, network carriers, and service providers that needed a place where rack space, local support, power density, and carrier access met in one Columbus facility. The public record now has two faces. The current DataCenter.BZ website at http://www.datacenter.bz/ no longer behaves like an active operating site; it resolves to a parking-style lander at https://www.datacenter.bz/lander. At the same time, PeeringDB still preserves a DataCenter.BZ, LLC organization record at https://www.peeringdb.com/org/7026, and Cologix's current Columbus materials identify the Scherers Court locations as active pieces of a much larger Columbus data-center platform.

That split matters. A buyer looking only at the old domain would see uncertainty. A buyer looking at the acquired facilities would see a continuing Columbus interconnection story. DataCenter.BZ's economic significance sits between those two signals: the original company built a local colocation asset valuable enough to be acquired, and the acquired asset helped seed a Columbus platform that is now marketed around network choice, cloud access, and high-density capacity. The company should therefore be read as a legacy operating business whose public identity has faded while its facility logic has lived on in a larger platform.

The original value proposition was not mysterious. DataCenter.BZ sold the things that make local colocation attractive when a customer is too operationally sensitive for bargain hosting and too hands-on for pure cloud: cabinet space, cages, high-density power, carrier diversity, dark fiber, compliance-sensitive environments, virtual infrastructure, disaster recovery workspace, and 24-hour support. Its own 2013 release described headquarters in Columbus, high-density colocation, virtual data-center services, a campus at the convergence of regional, municipal, and long-haul fiber, and 32,000 square feet of raised floor, with a second 90,000-square-foot raised-floor complex then under development; the release remains available at https://www.prnewswire.com/news-releases/datacenterbz-is-the-fastest-growing-data-center-provider-in-columbus-ohio-228316851.html. That is not a hyperscale narrative. It is a regional infrastructure narrative: local businesses wanted capital expense turned into monthly service, but they still wanted proximity, control, and people they could call.

The resulting judgment is straightforward. DataCenter.BZ was economically strongest where Columbus customers valued local operating control more than brand scale. Its weakness was that the same control layer required capital, power, network relationships, and credibility that became easier to fund inside a larger platform. The acquisition by Cologix in 2014 was therefore not a footnote. It was the expected outcome for a facility that had outgrown a small local-provider balance sheet but retained strategic value as an interconnection point.

The company is clearer after the acquisition than before it

DataCenter.BZ, LLC appears in public infrastructure records as a company tied to Columbus-area colocation and network infrastructure. PeeringDB's organization page at https://www.peeringdb.com/org/7026 lists DataCenter.BZ, LLC, two facilities, and a network entry associated with AS40715. Two facility pages provide the more concrete location evidence. The Worthington-Columbus facility page at https://www.peeringdb.com/fac/1328 lists DataCenter.BZ - Worthington-Columbus at 535-575 Scherers Ct., Worthington, Ohio 43085, with five listed networks and a last-updated timestamp in September 2025. The Datacenter.BZ facility page at https://www.peeringdb.com/fac/1420 lists 535 Scherers Court, Columbus, Ohio 43085, with no listed local exchanges and no listed networks on that facility page.

Those records should not be overread as a complete current operating profile. PeeringDB is a public interconnection database, and records can preserve legacy naming, operator history, and community-maintained facility context. The current website no longer offers the old service narrative. The safer reading is that DataCenter.BZ remains a useful label for the legacy Columbus facility and network evidence, while the operating commercial platform is now Cologix. That is consistent with Cologix's own 2014 acquisition release, preserved as a PDF at https://www.cologix.com/pdf/PressReleases/2014-02-04-Cologix-Acquires-DataCenter.BZ-In-Columbus-Ohio.pdf. The release said Cologix closed a transaction to acquire DataCenter.BZ in Columbus, described two network-neutral data centers across 32,000 square feet of raised floor, and identified the asset as a leading Ohio interconnect point with 30 network choices and the OHIO-IX inside the meet-me rooms.

The acquisition economics were unusually specific. Cologix said the transaction included five acres of land, buildings, mission-critical facilities, more than 100 customers, metro conduit and dark fiber assets, and more than $5 million of run-rate EBITDA. For a regional colocation business, that package explains the price logic better than any generic "data center" label. EBITDA showed recurring service revenue. More than 100 customers showed diversification. Dark fiber and conduit showed control over routes, not just rooms. Land and buildings gave Cologix the option to expand. The team continuity mattered because support quality was part of the product.

The pre-acquisition technical claims were also harder than a brochure slogan. A 7x24 Exchange Spring 2012 proposal for an Emerson Network Power and DataCenter.BZ case study, available at https://www.7x24exchange.org/downloads/7x24SP12_pres.pdf, described DataCenter.BZ's 535 Scherers Court headquarters, Gordon Scherer as president, a second carrier-neutral data-center expansion in 2011, facility-wide PUE of 1.25 or less, 2(N+1) redundancy on critical power and cooling infrastructure, and a 32,000-square-foot raised floor engineered for about 500 watts per square foot. It also described average rack density of 5-10 kW and high-density cabinet deployments up to 90 kW, with 50 kW in standard rack deployments. Those figures make the company less like a small hosting shop and more like a regional facility trying to sell high-density enterprise infrastructure before Columbus became a national data-center growth story.

Data Center Knowledge's acquisition coverage at https://www.datacenterknowledge.com/next-gen-data-centers/cologix-acquires-datacenter-bz-in-columbus-ohio gives the same shape from an industry-news angle: Cologix acquired DataCenter.BZ and its 32,000-square-foot Columbus facility, with customers gaining access to 30 network choices and OHIO-IX inside meet-me rooms. Fierce Network's report at https://www.fierce-network.com/telecom/cologix-snaps-up-datacenter-bz-establishes-ohio-data-center-foothold adds the strategic framing. Cologix was entering Ohio, Columbus was treated as a tier-two market with edge potential, and the acquired facility was presented as highly connected for its region.

That means the company identity is not best understood through the old domain alone. It is best understood through a sequence. DataCenter.BZ was founded and promoted as a Columbus high-density colocation and virtual infrastructure provider. It built enough local density to attract carriers and regional enterprise customers. Cologix bought it as an interconnection foothold. Then Cologix expanded the Columbus platform far beyond the original footprint. The name DataCenter.BZ remains relevant because it marks the origin of the acquired Columbus control layer, not because it still appears to be the active retail brand.

Scherers Court made the thesis physical

The strongest public evidence points to a specific place, not a vague metropolitan claim. Scherers Court in the Worthington-Columbus area is the spine of the story. Hurricane Electric's 2012 announcement at https://www.he.net/releases/2012-10-03.html said it established a point of presence at DataCenter.BZ's facility at 555 Scherers Court in Columbus. The same release described DataCenter.BZ as a Columbus company offering high-density colocation and virtual infrastructure from carrier-neutral, purpose-built facilities, with access to regional, municipal, and private fiber networks. It also explained why the point of presence mattered: Hurricane Electric could reach Ohio and the wider Midwest more efficiently, while DataCenter.BZ customers could access another global backbone through cross-connects and transport providers.

The address evidence then lines up with Cologix's current Columbus pages. Cologix's Columbus overview at https://cologix.com/data-centers/columbus/ lists COL1 at 535 Scherers Court and COL2 at 555 Scherers Court. COL1 is described in more detail at https://cologix.com/data-centers/columbus/col1/ as an interconnection hub with enterprise-grade infrastructure, dense regional connectivity, a 44,000-square-foot purpose-built data center, custom cabinets, cages and private suites, 30 MW of utility power available onsite, 45-plus unique networks in a Cologix-controlled meet-me room, and local support. COL2 at https://cologix.com/data-centers/columbus/col2/ is described as an enterprise-grade data center at the same Columbus campus, also framed around long-haul fiber crossroads, regional carrier fiber, uptime, network choice, custom cabinets, secure cages, and disaster recovery use.

The facility arithmetic is messy in a useful way. PeeringDB uses 535-575 Scherers Ct. for one facility and 535 Scherers Court for another. Hurricane Electric named 555 Scherers Court. The Cologix pages separate 535 Scherers Court as COL1 and 555 Scherers Court as COL2. The acquisition release used 32,000 square feet of raised floor, while current Cologix pages speak in 44,000-square-foot purpose-built facilities on the campus. The 7x24 proposal described a 66,000-square-foot headquarters and a second carrier-neutral data-center addition. This is not a reason to discard the evidence. It is a reason to read the numbers as different cuts of a campus: raised floor, building shell, old/new phases, and later Cologix packaging are not always the same measurement.

The continuity is not just geographic. The old DataCenter.BZ materials emphasized fiber convergence, high-density cabinets, compliance, local support, and disaster recovery seating. The current Cologix pages for COL1 and COL2 emphasize dense interconnection, Scherers Court locations, high power per cabinet, multiple fiber entrances, emergency workspace, 24-hour engineering and support, and the ability to back up primary sites in Chicago and New York. The language has changed from local founder-led provider to North American platform, but the operating thesis is recognizable.

That physical evidence also helps explain why a Columbus facility could serve a buyer choosing between local colocation and a hyperscale region. The buyer is not merely renting space. The buyer is choosing a control position. A rack in a carrier-dense facility allows private circuits, internet transit diversity, direct peering, remote hands, cloud onramps through the larger Cologix platform, and the option to keep physical ownership of hardware. A virtual machine in a remote cloud region gives speed and service breadth, but the customer gives up some physical control, some procurement clarity, and often some ability to isolate network path economics. For many mid-market customers, the question is not cloud versus colocation. It is which workloads deserve which operating model.

Scherers Court solved the part of that question that cloud does not automatically solve: where to place shared infrastructure for customers that need both proximity and network choice. A Columbus managed-service provider could put backups, firewalls, storage arrays, customer-owned servers, and private interconnects in a local data center, then decide workload by workload what should go into public cloud. That was the commercial niche DataCenter.BZ occupied.

The business model sold certainty in monthly units

The most important economic claim in the 2013 DataCenter.BZ release was not the growth award. It was the conversion of customer capital expense into predictable operating expense. The release at https://www.prnewswire.com/news-releases/datacenterbz-is-the-fastest-growing-data-center-provider-in-columbus-ohio-228316851.html described customers reducing operating and capital costs, meeting uptime and compliance demands, and replacing persistent individual capital investment with scalable service spending. That is the core colocation bargain.

For smaller enterprise and hosting buyers, the full cost of an internal server room is rarely visible at first. The first rack looks cheap. Then come redundant power circuits, UPS refresh, generator maintenance, cooling, fire suppression, physical security, monitoring, carrier contracts, emergency access, after-hours staffing, audits, and the uncomfortable question of whether a building designed for office work should also serve as critical infrastructure. Colocation converts those shared burdens into cabinet, cage, power, bandwidth, remote-hands, cross-connect, and support charges.

The revenue logic follows the same structure. DataCenter.BZ could monetize space by the cabinet or cage, power by capacity, connectivity by cross-connects and carrier relationships, virtual infrastructure by service layer, disaster recovery by workspace or continuity package, and customer stickiness through operational dependence. Once equipment is installed, circuits are ordered, policies are written, and backups are pointed at the facility, churn is costly for the customer. The customer may negotiate price, but it does not move racks casually.

That explains why more than $5 million of run-rate EBITDA was an acquisition-relevant number in the Cologix PDF. EBITDA in this context implies a base of recurring colocation, connectivity, and managed infrastructure revenue after facility operating costs. It is not the same as high-margin software revenue, because data centers carry real estate, power, cooling, maintenance, staffing, and capital costs. But it is attractive when the asset also controls scarce local interconnection density. The margin is not just rent. It is rent plus reliability plus network optionality plus customer inertia.

The public pricing market gives a useful floor for the smaller-buyer economics. CeraNet's Ohio colocation pricing page at https://www.cera.net/home/ohio-server-rack-cage-colocation-cost/ advertises individual 1U server colocation at $139.95 per month with a standard power supply and a full dedicated rack at $950 per month with one 20 amp power supply, plus charges for additional power. CeraNet's broader site at https://www.cera.net/services/colocation/features-specs/ emphasizes no-startup-cost options, predictable pricing, flexible contracts, technical guidance, and local support. Those are not DataCenter.BZ prices, but they show the price grammar that a local buyer understands: rack units, full racks, power circuits, ports, uptime, support, and annual discounts.

For a Cologix-scale facility, many contracts will be custom rather than menu-priced. Still, the menu-priced local market helps explain why DataCenter.BZ's niche existed. A small customer may start with one or two servers, then move to a rack, then to a cage, then to blended cloud and colocation. The facility that hosts that customer's early deployment has a chance to become the default location for new firewalls, storage, backup appliances, carrier handoffs, and customer-specific hardware. In that sense, local colocation can become the physical operating surface for a managed-service provider. The provider sells services, but the facility makes those services dependable.

The cloud comparison is not only compute price

For the managed-service provider, the real comparison is not a bare server versus one cloud instance. It is the monthly bill after bandwidth, storage growth, physical access, cross-connects, compliance visits, and support are counted. AWS's EC2 pricing page at https://aws.amazon.com/ec2/pricing/on-demand/ says customers receive 100 GB of free data transfer out to the internet per month across most AWS services and regions, and then pay data-transfer-out rates by tier. The same page lists the first 10 TB per month at $0.09 per GB, the next 40 TB at $0.085 per GB, the next 100 TB at $0.07 per GB, and volumes above 150 TB at $0.05 per GB. That means a stable 10 TB/month outbound workload can add close to a rack-sized monthly line item before the buyer has paid for compute, storage, support, monitoring, backup, or private connectivity.

AWS Direct Connect changes the comparison, but it does not make bandwidth free. AWS's Direct Connect pricing page at https://aws.amazon.com/directconnect/pricing/ says pricing depends on capacity, port hours, and data transfer out through the Direct Connect location; it also says data transfer in over Direct Connect is charged at 0.00 USD per GB. Cologix's Columbus page says its Columbus data center hosts an AWS Direct Connect node. That is an advantage for a Columbus hybrid buyer because private connectivity can reduce latency and make traffic routing more predictable. It is still a paid design choice. A customer needs a port or hosted connection, a local provider or cross-connect path, routing skill, and enough traffic value to justify the fixed and variable pieces.

This is where DataCenter.BZ's old value proposition still reads as economically precise. A local MSP could use a Scherers Court rack for appliances, backup targets, customer-owned servers, licensing-constrained workloads, and private circuits, then use AWS for elastic front ends or managed services. The MSP could keep high-volume east-west or backup traffic inside the local facility, use cloud selectively, and avoid turning every restore, replica, or customer download into an egress event. The local rack did not eliminate cloud spending. It gave the MSP a place to decide which bytes deserved cloud pricing and which bytes belonged on equipment under local control.

The rough buyer arithmetic looks like this. A CeraNet-style local full rack at $950 per month with a 100 Mbps dedicated port is not equivalent to a Cologix contract, but it gives the lower end of the Columbus price grammar. An AWS egress tier at $0.09 per GB makes 10 TB of outbound traffic worth about $900 before other cloud line items if the traffic is billed in that tier. A managed colocation buyer with owned hardware, stable traffic, and a predictable support burden may prefer fixed facility and transit charges. A software team that values managed databases, autoscaling, and global deployment may still prefer cloud. DataCenter.BZ mattered because it gave Central Ohio buyers a credible middle position: local hardware economics without the burden of owning the data-center plant.

PeeringDB changes the reading from real estate to control surface

Without PeeringDB, DataCenter.BZ could look like a real estate and hosting story. With PeeringDB, it becomes an interconnection story. The organization page at https://www.peeringdb.com/org/7026 still lists DataCenter.BZ, LLC with facilities and a network entry. The Worthington-Columbus facility page at https://www.peeringdb.com/fac/1328 lists five networks: Amplex Electric, DataCenter.BZ, Everstream, Fidelity Voice & Data, and Horizon Telcom. The Columbus facility page at https://www.peeringdb.com/fac/1420 lists the same organization and address family but currently shows no peer networks on that page.

The five-network signal on the Worthington-Columbus page is modest compared with a major carrier hotel, but it is meaningful for a local colocation facility. It shows that the location is not merely a powered room; it has enough interconnection relevance to be represented in a public peering database. The names are also regionally logical. Everstream and Horizon Telcom fit a Central Ohio connectivity frame, while the DataCenter.BZ network entry points to the facility's own network history. This is precisely the kind of evidence that matters to a managed-service provider: which networks can be reached, which carriers are physically or operationally close, and whether the data center can support private connections rather than just generic internet transit.

The BGP evidence adds another layer, but it needs careful handling. Hurricane Electric's BGP Toolkit page at https://bgp.he.net/AS40715 currently identifies AS40715 as Cologix, Inc., still shows the company website field as http://www.datacenter.bz/, and lists prefixes including one described as DataCenter.BZ, LLC. BGP.tools at https://bgp.tools/as/40715 identifies AS40715 as Cologix, Inc., shows the website as http://www.datacenter.bz/, marks the network active and allocated under ARIN, lists 10 IPv4 prefixes and one IPv6 prefix originated, and shows 67.154.188.0/22 with a DataCenter.BZ, LLC description. ARIN RDAP at https://rdap.arin.net/registry/autnum/40715 gives the registry view: AS40715, name COLOGIX-COL, registration date 2008-03-03, last changed 2022-08-18. These pages do not make AS40715 the subject of the article. They show that routing records preserve the transition from DataCenter.BZ to Cologix and that the old name remains visible in network-resource traces.

The same caveat applies to RADB-style route registry records. They are operational evidence, not corporate narrative. They can reveal who maintains routing objects, what names still appear in route descriptions, and how legacy assets are folded into a larger network. They do not tell a buyer about customer service quality, uptime history, pricing, or contractual terms. For this article, their role is narrower: they confirm that DataCenter.BZ's footprint extended beyond a building name into network administration and interconnection records.

PeeringDB also highlights the reputational uncertainty around DataCenter.BZ's standalone identity. The organization record exists, the facility records are recently updated, and the address trail is concrete. But the current website is parked, the public brand is no longer active in the way a retail colocation buyer would expect, and Cologix's own current pages are the more useful source for operating details. The evidence therefore supports a legacy-company thesis: DataCenter.BZ built the control point; Cologix now commercializes and expands it.

Columbus fiber made local colocation more than local storage

The DataCenter.BZ thesis depends on Columbus being more than a cheaper place to put racks. Ohio IX's website at https://ohioix.net/ describes Columbus as a strategic location for high-capacity infrastructure and says Ohio IX helps service providers, content networks, and enterprises exchange internet traffic, reduce costs, improve performance, and keep local traffic closer to local users. It also lists Cologix Data Center among Central Ohio points of presence. The language is promotional, but the mechanism is real: local exchange and regional carrier density can reduce backhaul dependence and improve routing options.

Cologix's current Columbus page at https://cologix.com/data-centers/columbus/ makes the same argument at platform scale. It markets the Columbus data centers as the most connected facilities in Ohio, says customers can use an AWS Direct Connect node in Columbus, and lists 45-plus network providers, 35-plus cloud onramps, and a 99.999 percent uptime SLA. The exact numbers are Cologix's current claims, not DataCenter.BZ's original claims, but they show what the acquired foundation became. The old local control layer has been absorbed into a broader interconnection product.

COL3 deepens the comparison. Cologix's COL3 page at https://cologix.com/data-centers/columbus/col3/ describes a Tier III Certified Constructed Facility, a 160,000-square-foot, 18-plus MW data center directly linked to COL1 and COL2, with 45-plus unique networks in the meet-me room, 16-plus fiber entrances, Columbus FiberNet connectivity, metro dark fiber availability, and fiber feeds reaching all 88 Ohio counties. COL4 at https://cologix.com/data-centers/columbus/col4/ is described as a 256,000-square-foot Scalelogix data center with up to 33 MW and 50-plus unique networks. COL5 at https://cologix.com/data-centers/columbus/col5/ is expected to be ready for service in Q3 2026 with 60,000 square feet, 25 MW of utility power, direct access to 50-plus networks, cloud providers, AWS Direct Connect, Google Cloud Platform, and Ohio IX. COL7 at https://cologix.com/data-centers/columbus/col7/ is positioned on the Johnstown campus with 36 MW site capacity and direct fiber access to 50-plus networks.

These later facilities can obscure the smaller original story. DataCenter.BZ did not begin as COL4 or COL7. Its importance was that it gave Cologix a connected Columbus base. The 2014 acquisition did not simply add square footage; it added a market position in a region where enterprises wanted alternatives to Chicago and New York, where disaster recovery logic favored inland geography, and where regional fiber could support local and national routes. That is why the original acquisition language focused on meet-me rooms, OHIO-IX, dark fiber, customers, and land.

The result is a two-tier reading of Columbus. For hyperscalers and large cloud buyers, Columbus now competes as a power-and-land market with growing data-center campuses. For smaller managed-service and enterprise buyers, Columbus still competes as a local interconnection and support market. DataCenter.BZ's original value sat in the second tier. Cologix's current platform tries to capture both.

Rack economics favor buyers that know their own workload

The managed-service provider in the opening scenario needs to answer a practical question: which workloads belong in a local rack? The answer is not "everything." Commodity web applications, bursty compute, globally distributed services, and software that can be rebuilt with managed cloud services often belong in public cloud. But workloads with stable demand, expensive egress, specialized hardware, compliance-sensitive physical controls, predictable storage growth, or legacy licensing can be strong colocation candidates.

DataCenter.BZ's historical offering fit that mix. It described high-density colocation, virtual data centers, storage, servers, network systems, demand-based infrastructure, private cages, custom cabinets, and disaster recovery support. The company was selling a bridge between customer-owned infrastructure and provider-managed infrastructure. That bridge can be attractive to customers that want a known place for hardware but do not want to operate the building.

The cost comparison is easiest to see in power and network increments. A single 1U server at a low local price can be cheaper than a cloud instance if the workload is steady, bandwidth is predictable, and the customer already owns the hardware. A full rack can be cheaper than a cloud migration if the customer's applications are not being modernized and the main need is uptime outside the office. But the comparison flips when equipment refresh, spare parts, security operations, backup architecture, licensing, and labor are included. Colocation is economical when the buyer understands the full operating model. It becomes expensive when the buyer treats it as just cheap space.

DataCenter.BZ's strength was that it could sell certainty as a service. A buyer did not need to build a data room, buy redundant UPS systems, negotiate multiple carriers, or staff the facility around the clock. It could place equipment in a facility built for the purpose and use cross-connects, remote hands, and local support. The more the buyer valued specific physical control, the more attractive the model became. The less the buyer cared about physical control, the more public cloud competed.

The revenue mix also explains why the local support reputation mattered. A facility with equivalent power and fiber can lose business if customers do not trust remote hands, access procedures, ticket response, or escalation quality. The 2014 acquisition coverage repeatedly emphasized high-touch local support. That support was not sentimental. It was a monetizable layer. When a customer's server is down at 2 a.m., the difference between an engineer in the building and a distant support queue is part of the product.

Non-official market signals support that reading without proving service quality. A 2014 LowEndTalk discussion at https://lowendtalk.com/discussion/30169/need-stable-data-center-for-solocation-in-us shows buyers comparing U.S. colocation options for custom network configuration, BGP sessions, redundant internet connectivity, and thoughtful data-center staff; one participant suggested Cologix in Columbus, and another referred to Datacenter.BZ as now Cologix. That is only forum chatter, not verified customer evidence. Its value is that it shows the buying vocabulary of the segment: custom routing, redundancy, local support, and practical facility selection.

The cost base was power, cooling, staff, security, and reinvestment

The attractive revenue model comes with a heavy cost base. Data centers consume capital before they produce reputation. Power infrastructure, generators, UPS systems, switchgear, cooling plant, security systems, fire protection, building maintenance, insurance, compliance audits, remote-hands labor, and network operations all have to be funded before the customer sees a simple monthly invoice.

DataCenter.BZ's public materials stressed 30 MW of utility power, multiple substation feeds, backup diesel generation, and 24-hour onsite support. Cologix's current COL1 and COL2 pages continue to emphasize 30 MW of utility power available onsite, redundant configurations, multiple utility feeds, onsite fuel capacity, round-the-clock monitoring, biometric access, and local engineers. Those features are valuable because they are expensive. A buyer pays for the right not to own that complexity.

Power is now the biggest swing factor in the Columbus market. Cologix's 2024 COL4 announcement at https://cologix.com/news/cologix-first-colocation-provider-to-complete-ai-ready-data-center-columbus/ said its Columbus portfolio then spanned four data centers, 500,000 square feet, and 80 MW, all connected by a diverse fiber ring. It also described 50-plus unique network and cloud service providers, AWS Direct Connect, Google Cloud Interconnect, and Ohio IX. Its 2024 land-acquisition announcement at https://cologix.com/news/cologix-expands-central-ohio-footprint-with-land-acquisition-for-new-ai-ready-800mw-data-center-campus/ said Cologix acquired approximately 154 acres in Johnstown for a campus that could reach 800 MW across 2.0 million square feet at full buildout.

Those figures are far beyond the DataCenter.BZ era. They show how the cost base of Columbus colocation has changed. The early bargain was regional control at enterprise scale. The new bargain is high-density capacity in a power-constrained and politically visible market. A smaller buyer still cares about racks, remote hands, and cross-connects. But the facility operator now competes for utility capacity, tax incentives, construction labor, and long-term power commitments alongside hyperscale demand.

AEP Ohio's February 2026 update at https://www.aepohio.com/company/news/view?releaseID=10753 shows the scale of that pressure. AEP said data centers or developers had signed binding contracts for 5,642 MW under its data-center tariff, in addition to 12,219 MW of data-center contracts signed before the tariff, for 17,861 MW of contracted projects scheduled through 2035. It also said earlier requests had exceeded 30,000 MW before tariff filtering. Those numbers are not specific to DataCenter.BZ, but they change the Columbus operating environment. Power is no longer merely an input cost. It is a gating asset.

The Office of the Ohio Consumers' Counsel adds the public-bill side of the same story at https://www.occ.ohio.gov/factsheet/quick-facts-data-centers-ohio. It says Ohio has more than 200 data centers, with most in Central Ohio; that data-center growth can require transmission and substation upgrades; and that AEP Ohio's data-center tariff requires new large data centers to pay for at least 85 percent of contracted capacity for up to 12 years even if they use less. For a legacy Scherers Court buyer, that does not mean a cabinet contract suddenly becomes a hyperscale utility negotiation. It means the local facility's price is influenced by a power market where unused capacity, long-term commitment, and infrastructure recovery are now political issues.

The U.S. Energy Information Administration's monthly electricity update at https://www.eia.gov/electricity/monthly/update/end-use.php adds macro context: average retail electricity revenues per kilowatt-hour were rising year over year in 2026. For a data-center operator, even small changes in power cost, capacity charges, or minimum-demand obligations can materially affect pricing and margin. For a customer, that means the old local-colocation bargain can become less predictable when grid costs and capacity rules change.

Supplier dependence sat in the fiber room and the utility yard

DataCenter.BZ's supplier dependence was not the same as a software company's cloud-vendor dependence. It depended on utilities, fuel suppliers, generator and UPS maintenance, cooling contractors, carriers, fiber providers, equipment vendors, auditors, security systems, and skilled technicians. The facility's public value came from offering carrier choice, but carrier choice itself required commercial and physical relationships. The meet-me room was a marketplace only because carriers and networks had reason to be there.

Hurricane Electric's 2012 point of presence shows how one supplier relationship could also become a sales asset. When a global backbone puts equipment in a facility, existing customers get another route option, and outside customers get another reason to consider the facility. The announcement at https://www.he.net/releases/2012-10-03.html said DataCenter.BZ customers could connect directly through a cross-connect, while non-colocated businesses could connect through transport providers available at the facility. That is the interconnection flywheel in plain language: more carriers attract more customers; more customers attract more carriers.

The Cologix era amplified that flywheel. Current pages for COL1, COL3, COL4, COL5, and COL7 all stress meet-me rooms, fiber entrances, network-neutral status, cloud onramps, and Ohio IX. But the supplier dependence remains. If a facility loses carrier density, remote-hands quality, power availability, or utility confidence, its control-layer value weakens. A buyer does not colocate merely because concrete exists. It colocates because the facility provides reliable access to networks, people, and power.

Upstream dependence is also visible in the routing evidence. AS40715 now appears as Cologix in BGP tools, with Cologix as observed peer/upstream on public BGP pages. That is expected after acquisition. It suggests the legacy network is no longer an independent strategic center in the same way the old DataCenter.BZ brand may have presented itself. For customers, this can be positive: a larger parent can bring capital, broader network reach, cloud connectivity, and operational maturity. It can also reduce the old local-provider feel if service becomes more standardized.

The supplier-risk lesson is that local control is not independence from large systems. It is a different bundle of dependence. Instead of depending mainly on a hyperscale region and its service abstractions, the customer depends on a local facility, utility capacity, specific carriers, remote-hands quality, cross-connect delivery, and the owner's reinvestment cycle. DataCenter.BZ's success came from making that bundle attractive to Central Ohio buyers.

Customers bought proximity and proof

DataCenter.BZ's named customer list is not fully public, but the available sources indicate the customer profile. The company's own releases referred to telecommunications carriers, government entities, healthcare and education systems, technology service providers, and Fortune 1000 enterprises. LinkedIn's public company page at https://www.linkedin.com/company/datacenter.bz describes enterprise-grade data center and telecom solutions, high-density colocation, virtualization and cloud resources, metro-area dark fiber, virtual data centers and storage-area networks, disaster recovery, and an onsite 150-seat hot site. The page also says the company was founded in 2007 and had 11-50 employees. Those claims map to the 7x24 case-study outline, which listed government entities, educational institutions, healthcare systems, and Fortune 1000 companies among the customer categories.

Those claims are self-presented, so they should be weighed accordingly. Still, they match the acquisition evidence. A facility with more than 100 customers, dark fiber assets, meet-me rooms, and regional network choices would logically serve a mix of carriers, managed-service providers, institutions, and mid-market enterprises. It would also make sense for government, healthcare, and education customers that needed local control, audit support, and continuity planning.

One public financial-reporting trace gives a concrete example of the service category. An Ohio Auditor report for the Electronic Classroom of Tomorrow, available at https://ohioauditor.gov/auditsearch/Reports/2013/Electronic_Classroom_of_Tomorrow_12-Franklin.pdf, includes DataCenter.BZ in operating lease notes for server equipment space and property/equipment/security of server equipment. That does not establish service quality or a current relationship. It does show that DataCenter.BZ's business included real institutional server-storage arrangements, not just abstract hosting claims.

For customers, the practical value was proof. A local provider could tour the facility, meet staff, inspect cages, test remote-hands responsiveness, map carrier routes, and negotiate physical access. A public cloud region does not offer that kind of proof. It offers different proof: global scale, service breadth, certifications, APIs, and standardized operations. Many mid-market customers need both. DataCenter.BZ was positioned for the part that still needed a facility.

Customer dependence ran both ways. DataCenter.BZ depended on customers that would pay for quality rather than chase the lowest hosting price. The customers depended on a facility that could stay capitalized, maintain certifications, keep carriers, and preserve support. That balance helps explain why acquisition was rational. Customers gained access to Cologix's broader platform while keeping the local asset. Cologix gained a sticky customer base and a Columbus interconnection base.

Competition turned a local bargain into a platform contest

The Columbus data-center market has changed dramatically since the DataCenter.BZ acquisition. Cologix is no longer the only visible story. Cologix's own expansion created a larger internal platform. Expedient's Columbus data-center page at https://expedient.com/data-centers/columbus/ markets three facilities, 13.4 MW of critical IT load, 152,800 square feet of total size, and 59,600 square feet of raised floor. CenterSquare's Columbus page at https://www.csquare.com/data-centers/columbus markets modern colocation, power availability, 100 percent uptime SLA language, and built-to-suit options. Iron Mountain's Ohio data-center page at https://www.ironmountain.com/data-centers/locations/ohio-data-center markets an Ohio facility in Miamisburg serving Cincinnati, Columbus, and Dayton with 44,000 square feet and 1.4 MW of power. Google lists Central Ohio data-center communities at https://datacenters.google/locations/ohio, including New Albany, Lancaster, and Columbus.

The benchmark spread is the point. Cologix COL1/COL2 sell a carrier-dense campus story at Scherers Court. Expedient sells a Columbus managed-infrastructure and cloud-services footprint with published size and IT-load figures. CenterSquare sells colocation and build-to-suit flexibility. Iron Mountain sells a southwest Ohio facility with enterprise continuity language. Google is not chasing one-rack managed-service buyers; it changes the regional power, labor, land, and talent market around them. DataCenter.BZ's old niche was valuable because it sat below hyperscale procurement and above bargain server hosting. That niche survives only if the Scherers Court platform can keep enough local intimacy while Cologix scales the market.

Those competitors do not all target the same buyer. A hyperscaler building owned campuses is not selling the same product as a local 1U colocation provider. But they shape the same regional labor market, power market, land market, and perception of Central Ohio as digital infrastructure territory. DataCenter.BZ once competed by being local, carrier-neutral, and high-touch. Cologix now competes by pairing that local base with a North American platform and expanding capacity.

The current Cologix platform gives former DataCenter.BZ assets more value, but it also makes the old company harder to evaluate as a standalone entity. A buyer no longer asks, "Should I buy from DataCenter.BZ?" It asks, "Does the Cologix Columbus platform, rooted partly in the old DataCenter.BZ footprint, offer better economics and control than alternatives?" That is a different commercial question.

Competitor signals also change pricing pressure. Low-end local providers can advertise simple rack pricing. Larger platforms can emphasize network density, compliance, cloud onramps, support, and power scalability. A customer with one server may compare against CeraNet-style pricing. A customer with multiple cabinets, compliance needs, and cloud interconnect requirements may compare Cologix against Expedient, CenterSquare, or a direct cloud strategy. A customer with megawatt-scale needs is in a different market altogether.

DataCenter.BZ's historical sweet spot was between bargain hosting and hyperscale procurement. That middle has not disappeared. It has become more contested. The winner is the provider that can make the hybrid decision easy: local hardware where it makes sense, cloud connectivity where it is useful, support when things break, and pricing that does not surprise the customer after power and bandwidth are added.

Market signals point to lasting relevance and real uncertainty

The market signals around DataCenter.BZ are mixed in a way that is common for acquired infrastructure companies. The positive signals are strong. Cologix bought the business. Cologix publicly attributed more than 100 customers, dark fiber, land, buildings, and run-rate EBITDA to the transaction. PeeringDB still contains organization and facility records. Cologix's current pages map Scherers Court locations into the Columbus platform. Ohio IX lists Cologix among Central Ohio points of presence. BGP records still show the old website and legacy DataCenter.BZ naming traces.

The uncertain signals are also real. The DataCenter.BZ domain is parked. The brand's public site no longer explains services. PeeringDB records may preserve legacy labels rather than current commercial branding. Some public business-index pages suggest the original Ohio LLC may no longer be active, although those unofficial pages are not a substitute for direct state-record verification. The standalone DataCenter.BZ LinkedIn page remains visible, but it reads like a historical company profile rather than an active sales channel.

For a research judgment, the uncertainty does not destroy the thesis. It clarifies it. DataCenter.BZ is not being evaluated as a current independent operator with a fresh website and an active sales motion. It is being evaluated as a directory company whose public evidence ties it to a Columbus colocation asset, a network/interconnection footprint, and a 2014 acquisition that helped build Cologix's Columbus platform. The entity's relevance is therefore historical-operational rather than standalone-retail.

The market chatter is consistent with that. LowEndTalk references from 2014 treat Datacenter.BZ as "now Cologix" and place Cologix Columbus in the set of U.S. colocation options for buyers needing BGP, redundant connectivity, and thoughtful facility support. A WebHostingTalk thread at https://www.webhostingtalk.com/showthread.php?t=1469793 similarly contains a market-memory reference to Datacenter.bz becoming part of Cologix while buyers debated Midwest colocation options. Those are not proof of current service quality. They are useful because they show how technically aware buyers remembered the transition: the old local facility became part of a larger colocation provider while retaining Columbus relevance.

The bigger reputational risk is not that DataCenter.BZ was unimportant. It is that the old name can mislead readers if treated as a current independent brand. Any public profile should distinguish the legacy company, the acquired facility evidence, and Cologix's current operating platform. That distinction protects against two mistakes: overstating DataCenter.BZ's present independence and understating the role its assets played in Columbus colocation.

Regulation and geopolitics now price the same facility differently

The regulatory and geopolitical risk around a Columbus colocation facility used to be relatively local: zoning, utility reliability, diesel fuel, fire codes, compliance audits, customer data rules, and carrier availability. Those risks still matter. But the rise of large AI and cloud campuses has pulled Central Ohio into a broader fight over power, tax incentives, land use, and public cost allocation.

Ohio Revised Code section 122.175 at https://codes.ohio.gov/ohio-revised-code/section-122.175 provides the statutory basis for Ohio's computer data center equipment tax exemption. That kind of policy helped make Ohio attractive for data-center investment. But by 2026 the politics had shifted. The Ohio governor's site at https://governor.ohio.gov/media/news-and-media/governor-dewine-announces-pause-of-data-center-tax-exemption announced a pause in consideration of new data-center tax exemption requests while lawmakers reviewed the sector's growth. Signal Cleveland's reporting at https://signalcleveland.org/ohio-approves-last-data-center-exemption-before-moratorium/ said the Ohio Tax Credit Authority approved tax exemptions for two projects, including Cologix, around the time of the pause.

For DataCenter.BZ's original customer segment, this policy fight may feel distant. A small managed-service provider is not asking for 800 MW. But the indirect effect is meaningful. If power interconnection queues tighten, if capacity charges rise, if tax treatment changes, or if local communities resist new projects, the cost and availability of colocation can change for everyone. A customer renting a cabinet may not negotiate with the utility, but the facility's power economics flow into pricing and expansion options.

The geopolitical angle is mostly domestic. Columbus sits inside a U.S. infrastructure race shaped by cloud regions, enterprise data gravity, AI workloads, tax policy, grid planning, and regional competition. Cologix's Johnstown campus plans and the presence of Amazon, Google, Microsoft, Meta, Vantage, CyrusOne, and others in Central Ohio, as noted in Data Center Dynamics coverage at https://www.datacenterdynamics.com/en/news/cologix-buys-land-in-ohio-for-800MW-campus/, show that the region has moved from secondary-market story to national infrastructure battleground. DataCenter.BZ's legacy is part of that arc: a local colocation operator helped demonstrate that Columbus could support dense interconnection before the market scaled up.

Operational risk remains more concrete. A local-colocation customer still needs to ask whether the facility has enough power headroom, whether remote hands are responsive, whether fiber diversity is real, whether cloud onramps are priced sensibly, whether carrier contracts are flexible, whether equipment deliveries and access windows fit the customer's needs, and whether support is local enough for the customer's risk tolerance. Those questions are where the old DataCenter.BZ value proposition either lives on inside Cologix or loses relevance to alternatives.

What would change the judgment

The current judgment is that DataCenter.BZ mattered as a Columbus colocation control layer whose value was validated by acquisition and expanded through Cologix. Several facts could change that view.

First, verified state corporate records could clarify the present legal status of DataCenter.BZ, LLC and whether the original entity remains active, merged, dissolved, or otherwise maintained. That would not erase the facility history, but it would refine the company-status language. Second, current operator confirmation could explain why PeeringDB continues to show DataCenter.BZ, LLC records and how those records map to Cologix's facility management. Third, current customer or service data could show whether any services are still sold under the DataCenter.BZ name, although the parked website makes that unlikely.

Fourth, more detailed facility history could reconcile square-footage figures across older releases, acquisition coverage, PeeringDB addresses, and current Cologix COL1/COL2 pages. The public record includes 32,000 square feet of raised floor in acquisition materials, 44,000-square-foot descriptions on current COL1/COL2 pages, and larger later-campus figures for COL3 and beyond. The right answer may be that different sources count raised floor, building size, campus, or facility phases differently. Fifth, updated routing and PeeringDB evidence could show whether AS40715 and the related facility records remain operationally meaningful or are mostly legacy labels.

Sixth, pricing evidence specific to Cologix Columbus would sharpen the economics. Public local comparators and AWS egress tiers help frame the buyer's alternatives, but custom enterprise colocation pricing depends on cabinet density, power commitment, cross-connects, bandwidth, cloud connectivity, support, term length, and build requirements. Seventh, customer evidence would clarify whether the old high-touch support reputation survived inside the larger platform. The acquisition sources present that support as a strength; current customers would determine whether it remains a differentiator.

Finally, Ohio's power and tax regime could alter the whole Columbus calculation. If data-center tariff rules, tax exemptions, utility capacity, or community opposition materially raise costs, the value of existing connected facilities may rise because they already have power and network position. Alternatively, if those costs flow through too aggressively, smaller buyers may choose public cloud or lower-cost secondary facilities. DataCenter.BZ's original bargain was local control at a rational price. The durability of that bargain now depends on whether Cologix can preserve local-control economics while operating in a market increasingly shaped by megawatt-scale demand.

For the managed-service provider in the opening scenario, the answer is therefore conditional rather than nostalgic. Local colocation beats an anonymous hyperscale region when the customer needs physical control, predictable steady-state economics, carrier choice, remote hands, and a place close enough to understand. A hyperscale region wins when service breadth, elasticity, and abstraction matter more than proximity. DataCenter.BZ's contribution to Columbus was to prove that the local option could be more than a server room. It could be a control layer. Cologix's task has been to keep that layer useful as Columbus becomes a much larger and more expensive data-center market.