Research brief: Built to Last — why most SMB sites rebuild every 3-4 years (piece 5 of 15)

Status: Research material — not a finished article. Compiled May 2026.

Thesis

Most SMB websites get rebuilt every 2-4 years (Orbit Media: average across the Inc 5000 is 2 years 4 months). Orbit's own client base — sites under continuous care — averages 6 years 4 months between rebuilds. Lifespan is a function of architecture and maintenance, not a fixed property of websites.

The argument is not "never rebuild." It is "build so the rebuild is a choice driven by business strategy, not a forced move driven by accumulated debt."

What kills sites in 3-4 years

1. Plugin abandonment + vulnerabilities. 96% of WP vulns in 2024 were in plugins; 1,614 plugins removed from the .org repo for unpatched issues; 35% of 2024 vulns still unpatched in 2025.
2. Performance decay. Median desktop page = 2.56 MB and rising YoY; only 40-59% of mobile pages pass LCP.
3. Accessibility lawsuits. 5,000+ ADA filings in 2025; EAA enforceable since June 28, 2025; 95% of sites fail basic WCAG; 40% of new ADA filings are pro se (AI-assisted).
4. Page-builder lock-in. Divi 4 → 5 is a one-way migration; Elementor pricing restructures force decisions; Bricks RCE shows builders are attack surface.
5. Platform decay (Doctorow). Squarespace acquires Google Domains; Shopify sunsets checkout.liquid; WP Engine vs Automattic. Platforms are M&A and policy inventory.
6. Content decay + link rot. Ahrefs 14B-page study: 96.55% of pages get zero Google traffic. Zittrain: 50% of Supreme Court citation URLs suffer reference rot.

What the survivors do differently

Daring Fireball (since 2002), Berkshire Hathaway (since 1997, 198 KB pages), Craigslist, Pinboard, GOV.UK Design System, Stack Overflow — see Long-lived sites worth studying: Daring Fireball, Berkshire Hathaway, Craigslist, Pinboard, GOV.UK, Stack Overflow, Wikipedia for the catalogue. Common pattern: content separated from presentation, minimal dependencies, URL stability as a design principle, boring architecture as a feature.

Honest caveats

• The recycled "average website lifespan = 2y7mo" stat is unverified at primary source (HubSpot citation could not be located).
• No primary surveyed dataset of Canadian SMB rebuild cycles exists.
• The 10-year cost model (Scenario A: rebuild every 3 years ~$115k-$185k vs Scenario B: foundation-first ~$235k-$260k) is directionally correct but ignores SEO opportunity cost of 3× 523-day recovery curves in Scenario A.
• Some rebuilds are legitimately driven by business pivots, capability ceiling growth, or compliance forcing functions — not all rebuilding is decay-driven.