
“How much does a website cost?” is the wrong question, and everyone quoting you knows it. The honest question is what changes at each price — because a 1K website, a 5K website and a 25K website are not the same product at different markups. They're different products: different scope, different system depth, different aftercare. Here's the whole ladder laid out — what each tier actually contains, what it costs to deliver in the AI era, and how to tell which one a given project genuinely needs.
One disclosure that makes this guide different: we build the tooling agencies use to deliver these projects, so the delivery-hour numbers below reflect AI-assisted pipelines — the honest modern baseline. The scopes, though, are just the market: this is what the tiers have always meant, minus the mystery.
The agency staple: eight to ten pages, a real design system, native builder output, forms, SEO foundations and a proper launch. The full anatomy — scope, day-by-day plan, prompts, margin — is in how to create a 5K website using NibWP. What the extra 4K buys over the 1K tier is one word: persuasion. Custom sections that sell, a system that keeps every page coherent, and the polish cold traffic notices without knowing why.
Between the staple and the flagship sits the project band where an existing thing becomes a better thing: the 8K redesign (new design system, every page rebuilt, URLs preserved), the 10K template conversion (a premium design executed as a reusable library), and the 15K migration (full platform move with rankings protected). The common thread: the risk is continuity, so the deliverable is proof — checklists, redirect maps, before/after packs.
Every tier above 1K should end in a plan, because sites rot: links die, updates lag, content stales. The 500-a-month care plan is the recurring layer that keeps any of these builds worth what was paid for it — and for the agency, it's where project revenue becomes business revenue.
The tiers get concrete when real briefs hit them. A physiotherapy practice wanting 'a professional site patients trust': persuasion tier — 5K, ten pages, booking form wired to their calendar; the 1K version would cost them exactly the trust they asked for. A retailer with 900 SKUs 'moving off the old system': the storefront is a 3–5K build, but the catalog is a data operation priced per hundred — the honest quote separates them, and the dishonest one discovers the difference at week six. A founder with 'an idea like Airbnb for rehearsal spaces': that's the 25K system tier wearing a 5K sentence — model, search, booking loop, corpus; the kindest thing a professional does is price it truthfully and offer the phase-one version instead of selling the 5K mirage.
Whatever the tier, four costs live on the client's side of the table and belong in the budget conversation even though no agency invoices them. Content time: someone must answer briefs, supply copy or approve drafts — the calendar cost that stalls more projects than money does. Photography: stock reads as stock at every tier; real photos are the cheapest upgrade per euro on the whole ladder. Decision latency: every review round the client adds costs a working day; committees cost weeks — worth saying out loud when timelines matter. And domain-adjacent subscriptions that outlive any build: hosting, email, the booking tool, the analytics. A quote that surfaces these isn't padding — it's the difference between a project priced honestly and one that ends with a surprised client and an agency holding the blame for arithmetic nobody did.
Payment terms are scope's shadow, and they shift up the ladder. At 1K, take it in one — invoicing a fixed-scope product in halves just doubles the admin on a thin-margin item. At 5K, the classic half up front, half at launch keeps both sides honest through the review cycles. The 8–15K transformation band earns three stages tied to gates, not dates — signature, staging approval, launch — because the client's biggest fear is paying for an unfinished migration, and gate-based payments convert that fear into structure. At 25K, monthly against milestones with the corpus and monetization phases named separately: system builds have real dependency on client-side data delivery, and the payment schedule should make the client's deadlines as binding as yours. One rule spans every tier: the aftercare plan bills separately from day one — folding the first plan months “free” into the project price teaches the client the plan is worth nothing, which is the most expensive discount an agency gives.
Usually they're quoting different tiers at the same URL — one heard “presence,” another heard “persuasion.” The scope questions above surface the disagreement before the invoice does.
It's made them cheaper to deliver. Market prices move slower than delivery costs, which is why this is the best moment in a decade to run an agency — and why fixed-scope productized tiers beat hourly billing.
Yes, if the 1K site is built on real foundations — a builder, clean structure, sane URLs. A 1K site built on a page-builder island or a closed platform is a rebuild, not an upgrade.
Alongside, at every tier: the ladder prices structure and execution. Great content on a 1K structure beats thin content on a 25K one — budget both, weight content higher than most clients do.
Whatever the tier, an honest quote decomposes. Ask for the split across these six lines, and learn more from the shape than the total:
Budgets and needs misalign constantly; the honest moves are scope moves, not discount moves. Down-tier by cutting whole capabilities, not quality: a 5K need on a 3K budget becomes six pages instead of ten and one landing page instead of three — same system, same polish, smaller surface. Up-tier by phase, not by lump: the 25K directory on a 12K budget launches as model + templates + 100 listings, with monetization as phase two funded by the audience phase one builds. What never works: the same scope at 60% price — that money comes out of verification and aftercare, the two lines whose absence you discover at the worst possible moment. An agency that discounts scope-silently is showing you where they cut when pressed; prefer the one that says which capability leaves the quote.
Because unscoped quotes aren't comparable — the middle of three numbers describing three different products is noise. Fix the scope with the four questions first; then quotes describe the same product, and the spread tells you about the agencies instead of the ambiguity.
They own the tier below 1K, and honestly: a founder validating an idea should DIY. The ladder starts where a site must persuade strangers or carry operations — the point where structure, verification and aftercare begin producing returns a template can't.
The ladder is scope, not markup: presence, persuasion, transformation, operation. Pick the tier by what the site must do and what breaks if it's wrong — then insist on the tier's proof: a checklist at 1K, a system at 5K, a redirect map at 15K, a data model at 25K. The delivery machinery behind all of it is the agency AI stack — pricing here.