Building BuyVPS: Why we chose transparency over volume

Starting a hosting business during present times proves to be a straightforward process. The process of creating something that will endure through time takes more than just construction work.

The company founders Sjoerd and Robert made their first business decision to stay out of price competition with other providers. Our company would deliver measurable results that customers can track, and we would maintain complete transparency in all our operations.

This is our first post since launch. It covers how we build our platform and how our pricing works.

What we actually do

We provide Linux VPS infrastructure in Amsterdam and New York.

That sounds simple. The construction processes of these two buildings differ from each other.

The virtualization system for each VPS operates through KVM-based virtualization technology. No container-style resource sharing. No ambiguous allocation models. The system contains separate instances that operate with their own kernels and maintain specific resource restrictions.

The storage system operates with local NVMe arrays, which focus on maintaining consistency instead of seeking maximum marketing value. Our performance metrics are public and the methodology is documented. Median results are shown instead of "best runs".

The goal is simple: customers must be able to validate our results through evidence, not take our word for it.

How We Build Infrastructure

The first decisions we make will establish the path that all subsequent developments will follow.

Low-density node policy

Nodes are not filled to theoretical maximum capacity. The system has a fixed limit for available memory, which developers have established. The system achieves better stability because it minimizes the influence of noisy neighbors, which results in enhanced performance during times of peak usage.

Measured, not claimed

The published benchmarks include complete details about their system configurations. The system performs multiple tests of CPU, disk, network, and real database operations throughout different time periods.

Local NVMe architecture

Storage is not abstracted through oversubscribed network layers. The system enables users to monitor both I/O performance metrics and latency measurements, which they can track.

Consistent stack across regions

Amsterdam and New York run the same virtualization model, control layer, and operational policy.

These decisions limit margin in the short term. They protect customers in the long term.

Who we are building for

The main purpose of BuyVPS is to provide services to experts who require its capabilities for their professional activities instead of offering basic virtual server testing to budget-conscious users.

It is built for:

• SaaS applications
• Agencies managing client workloads
• High-concurrency APIs
• Production databases

The development teams, which focus on achieving accurate project forecasts, operate as their main priority.

These users care about p95 latency more than marketing slogans. The team focuses their efforts on I/O consistency rather than creating dashboard interfaces that have an attractive design.

That is who we serve.

Why we publish benchmarks

The hosting industry makes performance statements that lack specific details.

• "High performance."
• "Ultra-fast NVMe."
• "Enterprise-grade."

These phrases say nothing.

We publish:

• CPU benchmark results
• 4K random IOPS
• Average and percentile latency
• Network throughput
• Database workload results
• All with configuration details.

The testing of infrastructure stands as a requirement because it needs to function properly rather than existing for promotional purposes.

Why we are not the cheapest

The market contains providers who base their competition solely on offering lower prices. That is a valid strategy.

It is not ours.

The implementation of low-density architecture together with reserved capacity buffers and quality hardware components and published testing requires financial investment.

• The process of filling a node at maximum capacity leads to lower pricing rates.
• The operational work decreases when benchmarks fail to receive repeated execution.
• The abstraction of storage combined with excessive marketing of this technology results in higher profit margins.

We chose not to do that.

Our pricing reflects:

• Capacity headroom
• Measured performance
• Network reputation protection
• Operational transparency

We are not a market fighter. We are a stability-first provider.

Our Mission

Our organization exists to fulfill this fundamental reason, which is to

Build VPS infrastructure that behaves the same under real load as it does in a benchmark.

• No surprises.
• No teaser pricing.
• No performance lottery.

Infrastructure should be boring in the best possible way.

Our Vision

The organization will determine its future hosting stage through open disclosure methods, which it will implement.

• Published metrics.
• Clear policies, such as the Security policy.
• Documented methodology.
• Defined limits.
• Not inflated promises.

The organization will keep its present market entry requirements through using low-density nodes and performance assessments documentation and following its current operational policies.

The market expansion will achieve its growth through customer needs instead of pursuing brief marketing trends.

What Comes Next

Over the coming months we will:

• Continue publishing updated benchmark data
• All architectural decisions, the developers made while they where working on the assignment, have to be documented too.
• Refine multi-region deployment options
• The organization must release all technical information about the storage systems and the network infrastructure.
• We will not chase volume.
• We will build deliberately.
• Users who seek the cheapest VPS solution can choose from various affordable VPS options.

BuyVPS provides you with infrastructure that you can measure and understand and depend on.

This is only the beginning.