About these figures
The project observations below describe selected engagements and the reporting period stated in the methodology. They are not a representative industry survey or a forecast for your project. Client-level source records are not published with this article.
TL;DR
- 32 of the 103 SaaS sites we crawled have empty body content before JS hydration. Google indexes them anyway, but with a delay measured in days and a much higher rate of partial indexing.
- CSR-only React stacks underperform SSR / SSG stacks by 40-50 percentage points on indexed-page share. The gap is the dominant driver of organic search underperformance for funded SaaS.
- The most common blocking issue is internal links rendered via onClick handlers (41% of sites). Crawlers can't follow them. Sites lose interlinking signal that compounds over months.
Funded SaaS startups invest heavily in marketing pages and product content. They also disproportionately deploy JavaScript-heavy stacks where what you see in the browser is not what crawlers see in the raw HTML. We wanted to put numbers on the gap.
We sampled 103 funded SaaS companies (recent Series A rounds via Crunchbase Discover), crawled their marketing surface (typically the homepage, ~10 product pages, and ~10 blog posts), and compared the rendered DOM against the raw HTML response. We cross-checked indexed-page share against Ahrefs index data and, where available, Search Console coverage reports. The render-vs-raw approach mirrors Google's own publicly-documented two-wave indexing model, per the Google Search JS SEO basics.
Three original metrics: Render-Dependent Indexability (RDI), Crawl Budget Efficiency (CBE), and JS-Stripped Coverage (JSC). All defined below with formulas.
Methodology
The 103-site sample was drawn from public Series A rounds in the last 18 months. Filters: B2B SaaS, English-language site, >100 indexable URLs in sitemap. We crawled the home, top product page, and a top blog post for each, captured both raw HTTP response and post-hydration DOM, and ran a structured diff. Indexed-page share comes from Ahrefs index counts compared to sitemap counts.
Finding 1: CSR-only stacks lose 40+ points of indexability
Chart 1, Render-Dependent Indexability (RDI) by stack
% of textual & link content present in raw HTML before JS hydration. 100 = body text present in the initial response; 0 = entirely JS-dependent. This does not establish indexing status.
| Stack | Sample (n) | RDI |
|---|---|---|
| Next.js (SSR/SSG) | 38 | 96% |
| Next.js (CSR only) | 14 | 42% |
| React + Vite (SSG) | 12 | 91% |
| React + Vite (CSR) | 18 | 35% |
| Astro / 11ty (SSG) | 9 | 98% |
| WordPress | 5 | 95% |
| Webflow | 4 | 92% |
Sources: Crunchbase Series A list; custom crawler comparing rendered DOM vs raw HTML; Google Search Console index coverage where accessible; Ahrefs index data. Figures rounded.
Stacks split cleanly into two groups. Server-rendered or statically-generated stacks (Next SSR/SSG, Vite SSG, Astro/11ty, WordPress, Webflow) sit at RDI 91-98%. Client-rendered stacks (Next CSR, React+Vite CSR) sit at RDI 35-42%. The gap is consistent across sample slices, vertical, site age, traffic volume.
Finding 2: RDI predicts indexed-page share with high confidence
Chart 2, RDI vs % of pages indexed by Google
X = RDI; Y = share of pages found in Google's index. Each dot = a stack.
| Stack | RDI | % indexed |
|---|---|---|
| Next.js (SSR/SSG) | 96% | 81% |
| Next.js (CSR only) | 42% | 38% |
| React + Vite (SSG) | 91% | 76% |
| React + Vite (CSR) | 35% | 31% |
| Astro / 11ty (SSG) | 98% | 88% |
| WordPress | 95% | 79% |
| Webflow | 92% | 72% |
Sources: Crunchbase Series A list; custom crawler comparing rendered DOM vs raw HTML; Google Search Console index coverage where accessible; Ahrefs index data. Figures rounded.
RDI plotted against indexed-page share gives the cleanest correlation in the dataset. The break point is around RDI 70, below it, indexed share collapses fast. Above it, sites get most of what their sitemap promises into Google's index. This is the single most useful diagnostic in the report, measure RDI on your site before you debate any other technical SEO concern.
Finding 3: The top three issues are universal and fixable
Chart 3, Top indexability issues across 103 SaaS sites
% of sample affected. Bar colour = severity (3 catastrophic / 2 blocking / 1 cosmetic).
| Issue | % affected | Severity |
|---|---|---|
| JSON-LD added post-hydration | 47% | 2 |
| Internal links rendered via onClick | 41% | 3 |
| Image src lazy-injected via JS only | 39% | 2 |
| Heading hierarchy collapsed in raw HTML | 36% | 2 |
| Body content empty without JS | 32% | 3 |
| Meta description set client-side | 28% | 2 |
| Pagination/infinite scroll JS-only | 26% | 3 |
| Canonical missing or incorrect | 24% | 3 |
| Title tag injected by JS | 22% | 3 |
| Hreflang only after hydration | 19% | 2 |
Sources: Crunchbase Series A list; custom crawler comparing rendered DOM vs raw HTML; Google Search Console index coverage where accessible; Ahrefs index data. Figures rounded.
Most-affected issues:
- JSON-LD added post-hydration (47%) depends on successful rendering. Google supports this implementation, but clients that do not execute JavaScript will not see it. Test the rendered URL before treating the markup as missing.
- Internal links via onClick (41%), clicking a div with a router push doesn't produce a crawlable href. Internal linking equity evaporates.
- Images src lazy-injected via JS (39%), image search and image alt-text contributions are missed; CLS impacts Core Web Vitals on top.
How we score render-dependence and indexability
1. Render-Dependent Indexability (RDI)
RDI = (Tokens in raw HTML body / Tokens in rendered DOM body) x 100
This measures how much text is delivered before JavaScript runs. A lower score indicates a stronger rendering dependency; it does not establish whether Google has indexed the page.
2. Crawl Budget Efficiency (CBE)
CBE = Indexed pages / Pages requested per month by Googlebot
Use the same URL population and reporting period when comparing indexing and crawl data. Repeated crawls, canonicalisation and pages intentionally excluded from indexing affect the ratio, so it is a diagnostic measure rather than a universal health threshold.
3. JS-Stripped Coverage (JSC)
JSC = (Working internal links with JS off / Working internal links with JS on) x 100
The interlinking-equity proxy. Below 60 means most internal navigation depends on JS, which depresses both crawl reach and link equity.
Four counterintuitive patterns
- Indexed-page share decays over time on CSR sites even when the sitemap is clean. We tracked 8 sites over 6 months, CSR sites lost an average of 11% indexed-page share without any change to content.
- Validate structured data in the rendered page. Server rendering can simplify delivery and testing, but Google also supports JavaScript-generated JSON-LD. An organic traffic change cannot be attributed to schema alone without suitable controls.
- Hreflang errors compound silently when set client-side. International SaaS sites with hreflang only after hydration index correctly in en-US but lose signal in regional locales.
- The pagination / infinite scroll JS-only pattern is the worst-leverage issue. 26% of sample sites; almost all blog post indexing depends on it. Switching to crawlable pagination is the highest-ROI fix in the issue table.
Recommendations
For SaaS founders
Move marketing pages off CSR. The product app can stay CSR if it must, search engines don't index app routes meaningfully, but marketing surfaces (home, pricing, blog, docs, comparison pages) should be SSR or SSG. The traffic lift is significant and the migration is bounded. Our companion structured-data guidecovers suitable JSON-LD types, current search-feature support and measurement. And once you're indexed, indexing decay is the discipline that keeps you there.
We run this work as part of our technical SEO for SaaS engagement: prerender or SSR the marketing surface, audit and fix the issue list above, monitor indexation recovery for 90 days post-deploy.
For engineers building new SaaS
Default to SSR or SSG on the marketing surface. Next App Router with PPR, Astro, or Vite SSG with a prerender step all work. We recommend delivering the main content, canonical, metadata and internal links in the initial response where practical, then verifying what search engines can actually render. We bake this into every SaaS web app development engagement we ship.
Limitations
The 103-site sample is biased toward English-speaking, VC-funded SaaS. Bootstrapped SaaS, regional SaaS, and enterprise software in Java/.NET stacks have different shapes. Index counts from third-party tools approximate Google's real index, verify with Search Console for your own site.
The single highest-impact JS SEO fix
Run an RDI check on your own marketing site this afternoon. Open it with JavaScript disabled. If the body is empty, the metadata is missing, or the internal links don't work, you're on the wrong side of a 40-point indexability gap. Fixing it is the highest-leverage technical SEO move available to most funded SaaS sites.
If you want a private RDI / CBE / JSC report on your own site, send the URL, we'll run the same crawler and send the report back free.
The two companion SEO studies, what makes Google drop pages, and which schema types actually move rankings:
Research
Indexing Decay: A 217-Page, 12-Month Panel on When Google Drops Stale Content
We tracked 217 pages across four content types for a year. Decay curves, half-lives, and the refresh cadence that recovers traffic.
Research
Schema.org for SaaS: JSON-LD, Rich Results and Measuring SEO Impact
Choose suitable structured data, check current search-feature support and measure impact against your own baseline.
Two services that map exactly to the issues in this report:
Service
Technical SEO for SaaS
Prerender or SSR the marketing surface, audit and fix the issue list, monitor indexation recovery for 90 days post-deploy.
Service
Maintenance & Support
Ongoing engineering to keep your indexability, performance, and dependencies healthy after launch.
Frequently asked questions
- Does Google reliably index JavaScript-rendered content in 2026?
- Google can render and index JavaScript. Rendering errors, blocked resources, response codes, canonicalisation and content quality can still affect the result. A raw-versus-rendered HTML comparison identifies a dependency on JavaScript; it does not by itself prove a page is unindexable.
- What is the highest-ROI JS SEO fix on a React or Next.js site?
- Start with the issue preventing useful pages from being crawled, rendered or indexed. Check response codes, robots directives, canonical URLs, internal links and rendered content before prioritising structured data. Google supports JavaScript-generated JSON-LD when it is accessible in the rendered page.
- Why does indexed-page share decay on CSR sites even without content changes?
- Investigate changes in indexing with Search Console and server logs. Rendering failures, canonical changes, duplicate content, crawling restrictions and search-system changes can all contribute. An observed correlation with client-side rendering does not identify the cause.
