With SSR (server-side rendering), the server generates HTML containing the content before sending it to the browser. With CSR (client-side rendering), the server first provides basic HTML, and the browser generates the content after downloading and executing JavaScript. The key questions are which content needs to appear early and be readable by crawlers, and whether the server or the browser handles rendering. Consider SSR first for pages where public content and search visibility matter. CSR may suit interaction-focused sections with less need for indexing. A hybrid approach is also possible.
The difference lies in the initial HTML, not just the final page
Both approaches can ultimately display a complete web page, but they differ in how the content appears. For an article page, SSR can fetch the article data on the server, assemble the title and body, and then return the HTML. CSR may initially display an empty container or a loading indicator, adding the article only after the browser retrieves the data.
Just because users can eventually see the text does not mean the initial response contains it. Content that appears only after JavaScript runs also requires an extra processing step for crawlers.
Compare initial display and subsequent interactions separately
- SSR: Usually displays the main content sooner, but generating HTML on demand for every request increases the server’s computational load. This is especially important during traffic peaks.
- CSR: The initial load requires JavaScript to download and run. Users are more likely to encounter a blank screen on slow networks or less powerful devices. Once loading is complete, partial updates can provide smooth interactions.
CSR moves page rendering to the browser, saving server-side rendering resources, but the server must still deliver files and API data. Welly’s article argues that SSR usually performs better for initial display and LCP (Largest Contentful Paint), but this does not mean every SSR page is necessarily faster.
Easier access does not guarantee higher rankings
If SSR’s initial HTML already contains the main text, crawlers can read it without waiting for JavaScript to run. With CSR, content may be missed or indexing may be delayed, so it is important to consider how well different crawlers handle JavaScript.
Sources differ in how they describe the SEO benefits. Welly’s 2025 article places greater emphasis on SSR’s advantages, while awoo’s Q&A record published that year reports that SSR offers no direct advantage, but CSR is more prone to indexing problems. A more cautious interpretation is that making content easier for crawlers to read does not guarantee higher rankings. An event Q&A should not be treated as a statement of current platform rules, either.
Check whether your implementation includes content in the initial HTML
An iThome Next.js example shows that when article data is fetched in a Client Component’s useEffect, it has not yet been retrieved when the server sends its initial response. That response therefore does not contain the article body. The example was adjusted to fetch the data on the server before rendering the article. What matters is where data fetching and rendering happen, not which framework is used.
yesharris recommends prioritizing SSR for important pages such as home, product, service, and article pages, while using CSR for carts, checkout pages, or purely interactive modules. This is advice on choosing a rendering approach based on each page’s search needs, not a rule every website must follow.
SSR and CSR: perspectives from across the web
What else should you consider when using infinite scrolling or product filters?
Beyond the rendering approach, crawlers also need paths to discover content. Content further down an infinite scroll may go unindexed because crawlers do not scroll down like users do; consider providing traditional pagination links or other internal links. If product filters are triggered only through code and have no corresponding URLs, consider giving filter criteria their own URLs and links.
Does Next.js dynamic always improve search performance?
Not necessarily. The iThome example notes that dynamic enables code splitting, but if you delay loading important sections, you still need to check whether their content is included in the initial HTML. The example's default settings still render static content on the server. Setting ssr: false skips server-side rendering, which may prevent important content from being read immediately.
How can you check whether the initial HTML contains important content?
You can inspect the HTML returned by the server directly to see whether the title and body text are already present, then compare it with the rendered result. In the iThome author's comparison project, Flutter Web's initial body consisted mainly of loading scripts, while a plain Vite Vue SPA contained only an empty container. This describes only that comparison project's initial responses and should not be generalized to all websites using these technologies.
Should you switch to SSR if “Crawled – currently not indexed” increases?
You should not switch based on this metric alone. Drawing on SEO experience, the iThome author recommends tracking the ratio of “Crawled – currently not indexed” pages to valid pages. Possible responses include changing the rendering approach, improving the content, or using noindex, depending on the problems with the pages and the website's needs. The article also notes that these cases may not be entirely avoidable, and the ratio should not be treated as a universal official standard.
References
- 從 IT 技術面細說 Search Console 的 27 組數字 KPI (7) :網頁的排除ithelp.ithome.com.tw — Discusses: Should you switch to SSR if “Crawled – currently not indexed” increases?
- AI SEO怎麼做?11個優化重點,讓品牌被AI搜尋推薦!www.yesharris.com · 2026-05-06T13:23:17+08:00 — Discusses: What else should you consider when using infinite scrolling or product filters?
- CSR、SSR 是什麼?定義、對 SEO 影響與網站選擇建議welly.tw · 2025-02-12T10:09:27.170Z — Source used in the article above.
- 【Search Central Live Deep Dive】Day2:索引才是王道!技術 SEO × JavaScript × 多語系策略全攻略|awoo 活動實記www.awoo.ai · 2025-07-25T07:14:49+00:00 — Source used in the article above.
- 【Day 27】用 Next.js 進行 SEO:Meta 設定與常見誤區ithelp.ithome.com.tw — Discusses: Does Next.js dynamic always improve search performance?
- Day 26|Flutter Web 的 SEO 困境:爬蟲看到的是什麼ithelp.ithome.com.tw — Discusses: How can you check whether the initial HTML contains important content?



