How Much Does Server Colocation Really Cost? Understanding What You Pay For

Server colocation costs go beyond simply renting rack space in a data center. The final price depends on several interrelated components: physical space, reserved power capacity and energy consumption, connectivity, IP addresses, redundancy and technical services. Understanding how these components are priced, and distinguishing between upfront and recurring charges, is essential for calculating the true cost of hosting IT infrastructure.

Two companies deploying identical servers may receive significantly different colocation quotes. One may need a single 2U server with a standard internet connection, while the other requires a dedicated rack, redundant power feeds, connectivity through multiple network carriers and guaranteed on-site technical support. The hardware may be similar, but the infrastructure requirements are different.

For IT managers and infrastructure decision-makers, the relevant question is therefore not simply "How much does a rack cost?" but rather "What infrastructure resources am I contracting, and how much will they cost over the entire deployment lifecycle?"

Understanding Colocation Costs: Initial Setup vs. Monthly Operations

Colocation pricing generally consists of two main categories: Non-Recurring Charges (NRC) and Monthly Recurring Charges (MRC).

NRC covers the initial deployment and activation of the infrastructure. Depending on the configuration, these charges may include rack preparation, installation of electrical circuits and Power Distribution Units (PDUs), cabling, network configuration and technical assistance during installation. Larger deployments may require additional work, such as constructing dedicated secure cages or implementing customised power distribution configurations.

MRC represents the ongoing monthly cost of operating infrastructure within the data center. It typically includes occupied rack space, reserved power capacity, electricity consumption, connectivity and any additional contracted services.

Some providers bundle these components into standard packages, while others charge for them separately. Comparing monthly fees without examining the services included can therefore lead to misleading conclusions. A comprehensive assessment should also consider expenses outside the colocation contract, such as equipment transportation, insurance, migration services and application cutover to the new infrastructure.

Rack Space: Physical Capacity Is Only Part of the Equation

Rack space is typically measured in rack units (U), with 1U representing 1.75 inches (44.45 mm) of vertical height. M247 Global's colocation services, for example, support individual 1U-2U servers, quarter racks (approximately 10U), half racks (approximately 20U), full racks (typically 42U-48U) and dedicated secure areas for larger infrastructure deployments.

However, the number of rack units occupied does not determine the price on its own. A 42U rack containing networking equipment may consume only a few kilowatts, while an identical rack populated with high-performance compute servers could require 10-15 kW or more. Both occupy approximately the same floor space, but their power distribution, cooling and redundancy requirements differ considerably. For example, M247 Global offers full-rack colocation configurations with 3 kVA, 6 kVA or 9 kVA power options.

This distinction is particularly relevant for high-density computing environments, where electrical capacity and cooling requirements may have a greater impact on costs than physical space.

Rack space pricing must therefore be evaluated alongside the power density a data center can support. A rack with a relatively low rental fee but insufficient power capacity may not be suitable for the intended workloads.

Power Consumption and Reserved Capacity: Where Costs Add Up

Electricity represents a significant operating expense for data centers, making the power component of a colocation contract particularly important.

A key distinction must be made between reserved power capacity, measured in kilowatts (kW), and actual energy consumption, measured in kilowatt-hours (kWh).

Reserved capacity represents the electrical power allocated to the customer's equipment, while consumption reflects the energy actually used during a given billing period.

Providers may apply a fixed monthly charge for reserved capacity, bill metered electricity consumption or combine both models.

For example, a rack operating at an average IT load of 4 kW consumes approximately 2,880 kWh over a 30-day month. At an illustrative rate of €0.18/kWh, close to the European average reported by Eurostat for non-household consumers in the second half of 2025, the monthly energy cost would be approximately €518.

However, the total power-related charge may also reflect the infrastructure required to deliver electricity reliably, including redundant power distribution, UPS systems and backup generators.

Cooling also contributes to data center operating costs. Infrastructure energy efficiency is commonly assessed using Power Usage Effectiveness (PUE), which compares total facility energy consumption with the energy used by IT equipment. A PUE of 1.4 means that for every 1 kWh consumed by IT equipment, the data center consumes 1.4 kWh in total.

From the customer's perspective, it is important to establish whether cooling and supporting infrastructure energy costs are included in the contracted rate or billed separately. PUE should not automatically be applied as an additional multiplier to the electricity bill.

Power redundancy is another contractual variable. Independent A/B power feeds provide alternative supply paths, but capacity allocation and pricing depend on the data center's architecture and the agreed service configuration.

Connectivity: Why a 10 Gbps Port Does Not Automatically Mean 10 Gbps of Bandwidth

Connectivity is another important component of colocation pricing, particularly for businesses running bandwidth-intensive applications.

A common source of confusion is the distinction between physical network port capacity and contracted bandwidth. A 10 Gbps port defines the maximum interface speed, but it does not necessarily mean that the service includes 10 Gbps of dedicated internet bandwidth.

Connectivity can be priced according to a fixed monthly traffic allowance, contractually committed bandwidth, temporary additional capacity (burstable bandwidth) or unmetered traffic models.

For applications with predictable data transfer volumes, a package with an included traffic allowance can provide good cost visibility. Streaming platforms, content distribution services and applications with significant traffic fluctuations may benefit from committed bandwidth or unmetered traffic models.

Some providers also use the 95th percentile billing method, where traffic is measured at regular intervals and the highest 5% of recorded samples are excluded before determining billable bandwidth. This model accommodates occasional traffic spikes but can result in higher charges when elevated utilisation levels are sustained over longer periods.

Choosing the appropriate connectivity model therefore depends on traffic volume, peak utilisation, application requirements and anticipated growth.

For more information, we also recommend our article: Metered vs. Unmetered Bandwidth: How to Choose the Right Dedicated Server Offer.

Cross-Connects and IP Addresses: Additional Components That Affect the Budget

Connectivity costs may also include cross-connects: physical copper or fiber-optic connections linking customer infrastructure to network carriers, cloud connectivity services or other customers within the same data center.

These connections generally involve an initial installation fee and a recurring monthly charge. The network service delivered over the connection, such as IP transit, may be billed separately.

For organisations using multiple carriers, private interconnections or redundant network paths, these charges can represent a meaningful part of the operating budget. Public IP address allocation may also generate additional costs. Although some colocation packages include a predefined number of IPv4 addresses, additional allocations may incur monthly fees.

Given the limited availability of IPv4 addresses, customers should verify how many usable addresses are included in their contract, the cost of additional allocations and whether they can use their own IP address resources.

Technical Services: The Difference Between Colocation and Managed Infrastructure

Colocation provides a controlled physical environment for customer-owned equipment, but it does not automatically include server or application management.

One of the most commonly contracted additional services is Remote Hands, which allows technicians at the data center to perform physical operations on equipment without requiring the customer's engineers to travel to the facility.

These activities may include server reboots, hardware component replacement, cabling, visual inspections and basic troubleshooting. Depending on the contract, Remote Hands services may be billed hourly, per intervention or through prepaid service packages.

Other optional services may include DDoS protection, managed firewalls, monitoring, backup and disaster recovery solutions. For companies hosting infrastructure in another city or country, on-site technical assistance can significantly reduce travel expenses and incident resolution times. These operational benefits should be considered when comparing service offerings, rather than focusing exclusively on the monthly fee.

Contract Duration and Reserved Capacity Influence the Final Price

The commercial structure of a colocation agreement can significantly influence its total cost.

A single server contracted on a monthly basis will generally have different unit economics from multiple dedicated racks contracted for three or five years. Larger deployments and longer-term commitments may benefit from more favourable commercial terms because they provide data center operators with greater predictability in revenue and capacity utilisation.

However, capacity planning is equally important. Reserving 10 kW when the infrastructure requires only 3 kW can generate unnecessary recurring expenses. Conversely, choosing a configuration with limited expansion capacity may lead to costly upgrades or migrations as demand increases.

Customers should therefore review contract duration, renewal conditions, annual price adjustments, charges for exceeding contracted power capacity and expansion options alongside the initial quotation.

The objective is not necessarily to secure the lowest possible monthly fee, but to contract a configuration that meets current requirements without generating excessive costs or limiting future growth.

How Geography Influences Colocation Pricing

Colocation costs vary significantly across international markets, reflecting differences in electricity prices, grid capacity, real estate costs, connectivity and demand. In North America, particularly in major US data center markets, strong demand for cloud and AI infrastructure has increased pressure on available electrical capacity. In Europe, energy costs, regulatory requirements and grid constraints create substantial differences between established markets such as London, Frankfurt, Amsterdam and Paris and emerging locations in Central and Eastern Europe, including Romania.

The Asia-Pacific region is equally diverse. Limited land availability and energy constraints influence pricing in markets such as Singapore and Tokyo, while other locations have different cost structures. Regardless of region, the lowest rack rental fee or electricity price does not necessarily translate into the lowest total cost. Network performance, infrastructure resilience, regulatory requirements and proximity to end users remain essential considerations.

M247 Global: Colocation Services Tailored to Your Infrastructure Requirements

M247 Global provides colocation services for infrastructure deployments of different sizes, from individual 1U and 2U servers to quarter racks, half racks, full racks and customised configurations.

M247 Global's colocation offering combines redundant power, cooling, physical security, connectivity and access to technical assistance. Depending on the selected configuration, packages may include allocated data traffic and IP addresses, with additional options for Remote Hands services and cross-connects.

By aligning rack space, power capacity and connectivity with actual application requirements, organisations can achieve a more predictable cost structure without investing in their own data center facilities.

Understand your requirements. Contract the right capacity. Pay for infrastructure that supports your business.

Explore M247 Global's colocation services to discover the available configurations and request a quotation tailored to your infrastructure requirements.