Site Search
Adding on-site search to a website — via a hosted crawl-index, a client-side static index, or a server/API-backed engine.
Definition
Site search lets visitors query a site’s own content. The key architectural fact (from Adding Search To Your Site In 15 Minutes): static/Jamstack sites have no server to query at request time, so search must come from either a client-side JS layer, a build-time static index, or an external hosted index that crawls the site. The article demonstrates the hosted-SaaS route (Site Search 360): sign up, it crawls and indexes from the public web, you style a no-code UI, and paste a JS snippet before </body> — live in ~15 minutes, with features like URL-pattern result grouping.
The spectrum is now two-sourced:
| Approach | How it works | In-wiki source |
|---|---|---|
| Hosted-crawl SaaS | Vendor crawls the public site; you paste a JS snippet | Adding Search To Your Site In 15 Minutes / Site Search 360 |
| Build-time static index | Post-SSG indexer writes a WASM/chunked bundle into the output; no server | Getting Started with Pagefind / Pagefind |
| Server/API-backed | Hosted or self-hosted engine (Algolia, Meilisearch, Typesense, Elasticsearch) | You May Not Need Algolia (decision rule, not a how-to) |
Pagefind fills the named registry gap for the build-time pole. You May Not Need Algolia supplies the decision rule: if the index rebuilds from the same static output users browse, start with Pagefind; use Algolia when search needs live writes, business rules, or user-specific ranking. A full Algolia (or Meilisearch) implementation guide is still a hole. The architectural principle is established.
Key Properties
- No server on static/Jamstack → search via client-side JS or an external/hosted index.
- Three approaches: hosted-crawl SaaS, build-time static index, server/API-backed.
- Hosted-crawl tradeoffs: fast setup vs. third-party JS weight, cost, and crawl-freshness.
- Relates to Web Architecture & Scalability and Rendering Strategies (CSR / SSR / SSG / ISR / RSC) (why static sites need this at all).
Examples from Sources
| Example | Source |
|---|---|
| Hosted Site Search 360 crawls a Jamstack site and adds search via one JS snippet | Adding Search To Your Site In 15 Minutes |
In the Sources
| Source | Context |
|---|---|
| Adding Search To Your Site In 15 Minutes | Hosted-crawl approach (Site Search 360). |
Related
Concepts: Web Architecture & Scalability, Rendering Strategies (CSR / SSR / SSG / ISR / RSC), Core Web Vitals Entities: Site Search 360, WordPress
Data gap
Only the hosted-crawl approach is documented so far. Backfill sources on build-time indexes (Pagefind) and API-backed search (Algolia/Meilisearch) to complete the spectrum.