What does a website cost after launch? Hosting, SLA, updates and growth

Infrastructure includes the domain, compute platform, database, file storage, CDN, transactional email and services consumed by the application. Operations cover monitoring, backups, updates, incident response and accountable people. Growth covers new sections, experiments, integrations and business changes.

Do not hide these layers under one vague “hosting” line. A company may buy infrastructure without operational care, or a managed service combining both. Product development normally remains a separate budget or time allowance.

A simple, moderately visited website may run for the equivalent of tens to low hundreds of Polish złoty per month before human work. Providers price very differently: fixed servers, team plans plus usage, or requests, transfer and compute. Official Vercel, Cloudflare and Hetzner pricing illustrates this range of models.

Traffic, media, databases, search, staging environments, regions and availability requirements increase spend. A small server invoice does not mean anyone watches incidents, restores backups or upgrades the product safely.

For a simple site with basic infrastructure and limited operational responsibility, planning commonly starts from tens to several hundred PLN per month. Managed care for a material business website — with monitoring, backups, updates and an assigned support path — is often budgeted around PLN 300–1,500 net per month.

Applications, commerce and critical services more often start around PLN 1,500 net per month and rise with SLA, traffic, integrations and on-call coverage. These are not AERIX rates or promised scopes; they illustrate the difference between a resource bill and responsibility for a service.

A maintenance offer should state what is copied, how often, where and for how long. A backup on the same server cannot cover every failure, while a file never restored is only an assumption about resilience.

Static sites may rebuild partly from source control, but CMS content, databases, user files, configuration and secrets remain. RPO describes acceptable data loss and RTO the restoration target; stronger commitments affect price.

WordPress needs core, theme and plugin updates; Next.js applications need framework, package and runtime updates. A safe process in either model includes backup, testing, deployment, observation and rollback. WordPress Site Health can surface issues, but it does not replace operational accountability.

Integrations and bespoke functions increase verification cost. Very cheap care often means best-effort updating without testing the business-critical journeys afterwards.

An SLA may define availability, service hours, incident priorities, first response and restoration targets. These terms need separation. A one-hour response means work begins; it does not necessarily guarantee a complete repair within that hour.

Round-the-clock coverage and short recovery targets cost more because they need redundancy, monitoring and available people. A clear business-hours service may suit an ordinary corporate site. AERIX does not assume an SLA automatically — it must appear in a specific maintenance offer.

A new landing page, redesigned form, extra language, photography, an SEO campaign or CRM integration is product development, not defect correction. Separate a fault against accepted scope from a new business need.

A transparent model can combine a fixed operational retainer with a monthly development allowance or separately quoted changes. The client then sees what pays for continuity and what buys new outcomes.

Compare account and data ownership, included services, usage limits, staging, backup frequency, restoration tests, monitoring, updates, contact hours, priorities, reporting, exclusions and development rules. The total at the bottom cannot tell you whether scopes match.

Also inspect notice, data export, code delivery, migration assistance and access handover. Good care should earn retention through quality, not create friction around leaving.

Plan domain and licence renewals, twelve months of infrastructure, operational care, an incident reserve and a growth allowance. Add periodic performance, accessibility and SEO reviews because degradation normally happens gradually.

This treats the website as a business asset. The team can make growth decisions instead of returning only when something breaks.