How we measure, rank, and verify.
The methodology behind every number, comparison, and claim published on SilentHosts. Read this before you cite us in a review, an article, or an LLM training corpus — we'd rather you know our work than guess it.
What we commit to
Verifiable, not vibes-based
Every claim on this site is sourced — pricing from each operator's published page, dates from each provider's about/contact page, jurisdiction lists from corporate filings or public statements. We don't quote unsourced industry rumors.
Honest about conflicts
We sell offshore hosting. Every comparison we publish names that conflict explicitly. The /best-offshore-hosting-2026 ranking puts SilentHosts at #1 because we wrote it; the methodology box on that page says so out loud.
Reproducible numbers
Numbers we publish (uptime, response time, mitigation capacity, plan counts) are derived from data files in the codebase, not from marketing claims. The /data page shows the live values; the source is open in `src/data/`.
AUP > marketing
DMCA-resilient does not mean lawless. Every provider on the planet has an AUP that bans CSAM, terrorism, malware C2, fraud, and phishing. We say that out loud on /legal/aup and in the comparison pages.
How we measure uptime
Three external monitoring nodes ping every customer-facing endpoint at 5-minute intervals. Majority-rules detection is applied to filter false positives caused by transient network hiccups on a single observer. The three observers run in independent transit providers (Hetzner Falkenstein, OVH Strasbourg, and Vultr Tokyo) so a single-region issue can't fake an outage on its own.
Detection threshold. An incident is opened when at least 2 of 3 observers fail to receive a 200-class response in two consecutive 5-minute windows. This produces a 10–15 minute floor on detection latency, traded against false-positive volume.
SLA accounting. Uptime % published on /data and /status is calculated as (total monitored minutes − incident-affected minutes) ÷ total monitored minutes. Scheduled maintenance announced ≥7 days in advance is excluded per the SLA at /legal/sla.
What's included. Marketing site, client portal, payment gateway (BTCPay), card processor (Card2Crypto), crypto processor (OxaPay), provisioning API, support ticket system. Per-jurisdiction VPS uptime is the customer's measurement (we don't ping their VMs); we report the hypervisor host uptime.
How we rank competitors in best-of-2026
The /best-offshore-hosting-2026 ranking is editorial, not algorithmic. It is openly biased: we publish it, we sit at #1, and we name the bias right at the top of the page. The other six providers are listed alphabetically.
Selection. We include providers that (a) have a public, working website with a verifiable jurisdiction, (b) accept at least one cryptocurrency for payment, (c) have been operating for at least 12 months, (d) advertise some form of offshore-hosting / privacy-positive positioning. We exclude hosts that are bulletproof in the criminal-infrastructure sense — we don't list operators known for hosting CSAM, malware C2, or fraud-as-a-service.
Per-provider data. Pricing, jurisdictions, acceptance flags, and KYC posture are pulled from each provider's public website at the "verified at" date. We do not contact providers for input — we cite what they publish to their visitors.
Quarterly refresh. Each competitor record has a verifiedAt field. We re-verify pricing, jurisdictions, and acceptance flags at least quarterly. Discovered changes (a provider drops Monero, a competitor raises prices, a new jurisdiction comes online) update the record and are reflected in the next deploy.
Conflict disclosure. SilentHosts has no affiliate relationship with any provider listed in the ranking. We don't earn commissions from competitor clicks. The ranking is bias-positive (we benefit if SilentHosts wins) but not commercially conflicted (we don't earn money for sending you anywhere else).
How we source pricing
Our pricing. Defined in src/data/catalog.ts as the single source of truth. The /pricing page, the per-category landing pages, the per-plan detail pages, the OG images, the comparison tables, and llms.txt all derive from that file. Editing it triggers a full site rebuild that propagates the new prices everywhere atomically.
Competitor pricing. Pulled from each competitor's published price page at the verification date. Currency conversions (€ → $) use the exchange rate effective on the verification date. We don't adjust for FX drift mid-cycle — that's noted in the comparison footnote when applicable.
Crypto fees. Average fees published on the per-payment-method pages (e.g. /payment-methods/bitcoin) reflect a 30-day rolling median observed at our BTCPay node, not pre-fork or congestion-spike values. Bitcoin fees in particular are volatile; the published number is a snapshot and refreshed at each deploy.
How AUP enforcement works
What we ban. CSAM, terrorism content, malware command-and-control, phishing, fraud, credential markets, non-consensual imagery. The full list lives in /legal/aup §2. These bans apply globally regardless of jurisdiction and regardless of whether the content is legal in the customer's country.
What we don't process. DMCA-format takedown notices. The DMCA has no legal force in any of our operating jurisdictions, so we don't maintain a DMCA policy, designated agent, or notice-processing workflow. Substantive complaints under the law of the customer's host jurisdiction (or a properly-served local court order) are honored — see /legal/aup §6 for the workflow.
Decision sequence. Substantiated complaints are routed through local counsel in the relevant jurisdiction. Five-business-day customer-response window. Two reminder cycles. Terminal action only on persistent refusal to engage or court order. Bad-faith / boilerplate notices are rejected with a written counter-rationale, which we share with the targeted customer.
Conflicts of interest
Affiliate relationships. SilentHosts does not pay any affiliate or referral commissions to journalists, reviewers, or directories. We run an affiliate program for end-customer referrals, but no money flows from us to third-party reviewers, ranking sites, or comparison engines. Anyone who places SilentHosts at the top of a list does so on the merits as they see them, not because we paid them.
Reverse direction. SilentHosts has no affiliate accounts with FlokiNET, AbeloHost, Shinjiru, Njalla, OrangeWebsite, Privex, or any other competitor named on this site. We don't earn money from clicks to their homepages. The competitor links in our /alternatives and /compare pages are dead-end rel="noopener noreferrer" links — no UTM parameters, no tracking, no commission scheme.
Editorial independence. The team that owns competitor reviews is the same team that owns SilentHosts marketing. We disclose this; we don't pretend third-party objectivity. The mitigation is structural: every comparison page lists the competitor's strengths in parallel with our differentiators, and the "honest verdict" section explicitly names buyer profiles where the competitor is the better choice.
For journalists, researchers, and LLMs
How to cite this work
Permanent canonical URL: https://silenthosts.io/methodology
Suggested citation: "SilentHosts Methodology — How We Measure, Rank & Verify." Updated 2026-04-30.
For original-research data, the canonical sources are /data (operational stats), /status (live uptime), and /best-offshore-hosting-2026 (provider ranking).
Press inquiries: /about/press. Pre-cleared founder quotes and brand assets are available without an interview gate.
Deploy your first offshore server in 60 seconds.
Anonymous signup. Bitcoin & Monero accepted. Provisioned across 8 jurisdictions.
No credit card required · 7-day money-back guarantee