Services

How to Build a Content/Media Website: Architecture, CMS Choice, SEO/GEO, and Subscription + Ad Monetization (Real Costs & 90-Day Roadmap)

2026.07.30 · 109 views
How to Build a Content/Media Website: Architecture, CMS Choice, SEO/GEO, and Subscription + Ad Monetization (Real Costs & 90-Day Roadmap)

A content site earns from architecture, not article count. Custom vs WordPress vs Ghost, real costs, hidden fees, pitfalls, and a 90-day plan.

Share:

"We publish 20 articles a month, but traffic is stuck at 30,000, ad revenue is under NT$10,000, and AI summaries are eating half our clicks." That's what a client running a knowledge-content site for three years said the first time they came to us. They didn't lack content — they lacked the architecture to turn content into a monetizable asset. A content/media site makes money not from article count, but from getting four things right at once: content architecture, CMS choice, SEO/GEO, and monetization. This piece breaks it down, with real costs and a 90-day roadmap.

When It Fits vs. When It Doesn't

A content/media site fits when:

  • You have steady output (at least 2–3 in-depth pieces a week) with long-tail search value.
  • You want content to build authority that drives consulting, subscriptions or e-commerce.
  • Your content is citable by AI (how-to, comparison, data-driven) and you want the GEO upside.
  • You plan to monetize with "subscriptions + ads + sponsorships," not ads alone.

It doesn't fit when (honestly):

  • You lack steady content capacity — a beautiful site withers in three months without new content.
  • You only want 10 "about the company" articles — that's a brand site, not a content site.
  • You expect "traffic on launch day" — organic traffic usually takes 6–12 months to build.
  • You rely purely on ads with under 100k traffic — AdSense income will be too low to bother.

Alternatives Matrix

OptionProsConsCost tier
Custom content site (Laravel + custom admin)Full control of members, payments, SEO structure, AI readabilityHigher upfront, needs developmentFrom NT$120,000
WordPress + pluginsFast start, rich plugin ecosystemPlugin/security burden, needs performance tuningNT$40,000–100,000
Ghost (subscription content platform)Built-in memberships and newsletter, good performanceLimited customization, weaker Traditional-Chinese ecosystem~US$9/mo + build
Medium / Substack (SaaS)Zero build, built-in trafficURL and SEO not fully yours, takes a cutRev-share / free start

Full Process Breakdown (with Tools and Deliverables)

  • Phase 1 · Content architecture and strategy (1–2 weeks): map topic clusters, define categories/tags, keyword and GEO topic selection. Deliverables: content map (Notion), IA diagram (Figma / FigJam).
  • Phase 2 · Design and CMS choice (2–3 weeks): decide custom vs WordPress vs Ghost; design article, listing and member templates. Deliverables: Figma designs, CMS decision doc.
  • Phase 3 · Development and monetization (3–5 weeks): build the backend, integrate members and payments (ECPay / Stripe), paywall, ad slots, structured data. Deliverables: launch-ready site, payment test report.
  • Phase 4 · SEO/GEO and performance (1–2 weeks): structured data, sitemap, Core Web Vitals (LCP) tuning, Cloudflare CDN, GA4 and Search Console. Deliverables: performance report, tracking setup.

Full Real-Cost Breakdown

  • Development: custom content site from NT$120,000 (members, payments, paywall).
  • Payment fees (hidden): ECPay ~2.8%, Stripe ~3.4% + fixed fee, charged on every subscription.
  • Hosting and CDN: VPS ~NT$600–2,000/mo; Cloudflare free to Pro ~US$20/mo.
  • Domain and SSL: domain ~NT$400–800/yr; SSL free via Cloudflare.
  • Content production (biggest hidden cost): an in-depth piece runs ~NT$1,500–4,000 outsourced, or internal hours — the true long-term cost of a content site.
  • Maintenance: security updates, backups, feature iteration; retainer from NT$8,000/mo.

Reality vs. Client Expectation

  • Clients think: build it and traffic comes. Reality: organic traffic usually takes 6–12 months; early on you rely on content accumulation and external referral.
  • Clients think: ads = money. Reality: under 100k traffic, AdSense often pays under NT$1,000/mo; subscriptions and sponsorships carry the load.
  • Clients think: more articles is better. Reality: 50 deep, cited pieces beat 500 shallow ones — the latter drag down site-wide authority.
  • Clients think: AI eating clicks is bad. Reality: being cited builds authority and brand; the key is shifting monetization from "clicks" to "subscriptions and conversions."

Common Pitfalls and How to Avoid Them

  • Pitfall 1: messy categories/tags. Fix: define topic clusters and tag rules before launch; don't let editors add ad hoc.
  • Pitfall 2: paywall blocks SEO. Fix: partial content public + structured data (isAccessibleForFree) so engines can read while you gate payment.
  • Pitfall 3: unoptimized images, painfully slow. Fix: auto WebP, lazy-load, CDN; protect Core Web Vitals.
  • Pitfall 4: optimize only Google, ignore AI citations. Fix: answer-first intros, tables/lists, a source on each claim — works across all four AI engines.
  • Pitfall 5: payments with no cancellation/reconciliation. Fix: build auto-cancellation, invoicing and reconciliation reports in from day one.
  • Pitfall 6: no newsletter, so traffic doesn't stick. Fix: use a built-in newsletter to turn one-off visitors into a re-reachable list.

Success Metrics + 90-Day Roadmap

  • Day 30: technical baseline — sitemap submitted, Core Web Vitals green, GA4/Search Console collecting, newsletter live. Watch: indexed pages, LCP.
  • Day 60: steady cadence — weekly output, 3–5 topic clusters built. Watch: organic impressions, newsletter subscribers, avg time on page.
  • Day 90: monetization on — paywall/membership open, first sponsorship live. Watch: subscription conversion, value per visitor, AI citation appearances.

Decision Checklist

  • ☐ I can steadily produce at least 2 in-depth pieces a week
  • ☐ My content has long-tail search or AI-citation value
  • ☐ I plan to monetize via subscription/sponsorship, not just ads
  • ☐ I need full control of member and payment data
  • ☐ I accept organic traffic taking 6–12 months to build
  • ☐ I have budget for content production (the biggest long-term cost)
  • ☐ I need paywall/content-locking
  • ☐ I care about Core Web Vitals and AI readability
  • ☐ I need a newsletter to retain visitors
  • ☐ I'm willing to sign a maintenance retainer for long-term iteration

Six or more checks: a custom content site pays off. Three or fewer: start with WordPress or Ghost.

Should I choose custom, WordPress or Ghost for a content site?

It depends on monetization complexity. Pure content + simple subscription: Ghost is fastest. Many plugin needs, limited budget: WordPress. Member tiers, complex payments, full control of data and AI structure: go custom.

With AI summaries eating clicks, is a content site still worth it?

Yes, but change the monetization logic. Shift from ad clicks to subscriptions, lists and conversions; being cited actually builds your authority. The key is guiding content toward your own paid or list entry.

Will a paywall hurt SEO?

Not if done right. Partial-public content + structured data (isAccessibleForFree: false) lets engines index while users pay — the standard approach for news and knowledge sites.

Call to Action

Want to know which architecture will monetize your content? We offer a "Content/Media Website Build" service (from NT$120,000), starting with a free content-monetization audit that calculates a sensible mix of subscription, ads and sponsorship, plus a 90-day roadmap.

Share: