Skip to main content

Make a GHL Funnel Page Indexable on WordPress

Purpose​

When a GHL funnel page is served on a WordPress site (apex domain) through the LeadConnector plugin → Native HTML Embed, Google reports the page as "Excluded by 'noindex' tag" even though crawling is allowed. This SOP documents the verified fix: the funnel's JavaScript injects a noindex meta at render time that no native GHL setting can disable, so it must be neutralized on the WordPress side. Applies to the SC website and any GHL-on-WordPress client.

Scope​

GHL funnel pages embedded on WordPress via LeadConnector Native HTML Embed. Not for funnels served on their own go. subdomain (those are controlled by the domain-level robots.txt — see Set robots.txt for GHL Domains).

Background — why this happens​

  • GHL has no per-funnel index/noindex toggle; robots is managed only at the domain level (Domains & URL Redirects → robots.txt code).
  • The funnel runtime writes <meta name="robots" content="noindex"> into the page head at render time (client-side JS). So the raw served HTML is clean, but Googlebot — which renders JavaScript — sees the noindex. A server-side fix alone is necessary but not sufficient.

Prerequisites​

  • GHL admin access; WordPress admin access to the site running the LeadConnector plugin.
  • The WPCode Lite plugin, the stack standard's one approved code-snippet exception for this snippet (stack standard), or the child theme's functions.php. Whatever hosts it must genuinely execute PHP: a header/footer "code" field that outputs raw HTML will not work.
  • curl for the static check and Google Search Console access for the authoritative render check.

Procedure​

1. Diagnose (confirm it's the JS-injected noindex)​

curl -s https://<domain>/<funnel-path>/ | grep -i 'name="robots"'
  • If the served HTML shows no noindex but GSC → URL Inspection → Live Test reports Indexing allowed? No: 'noindex' detected in 'robots' meta tag, the funnel JS is injecting it. (GSC renders JS; curl does not — this gap is the whole signal.)

2. Fix the layout first (complex pages)​

In LeadConnector → Funnels → the funnel → Edit Page: for pages with pricing tables / grids, uncheck "Keep my WordPress header and footer intact." The WP theme's CSS collapses the funnel's grid when this is on. The page then renders with the funnel's own header. Text-only pages can keep it on. Save → purge CDN.

3. Strip the noindex (server + client) via a WPCode PHP snippet​

Add a PHP snippet (Run Everywhere), listing one path per funnel slug. It removes any server-rendered noindex and injects a guard that removes the JS-injected noindex from the rendered DOM Googlebot evaluates:

add_action('template_redirect', function () {
$path = rtrim(parse_url($_SERVER['REQUEST_URI'], PHP_URL_PATH), '/');
if (!in_array($path, ['/pricing', '/platform'], true)) { // one entry per funnel slug
return;
}
ob_start(function ($html) {
$html = preg_replace(
'#<meta\s+name=(["\'])robots\1\s+content=(["\'])[^"\']*noindex[^"\']*\2\s*/?>#i',
'', $html
);
$guard = '<script>(function(){function s(){document.querySelectorAll(\'meta[name="robots"]\')'
. '.forEach(function(m){if(/noindex/i.test(m.content))m.remove();});}s();'
. 'new MutationObserver(s).observe(document.documentElement,'
. '{childList:true,subtree:true,attributes:true,attributeFilter:["content","name"]});})();</script>';
return preg_replace('#<head([^>]*)>#i', '<head$1>' . $guard, $html, 1);
});
}, 0);

Activate → purge CDN.

4. Duplicate-content hygiene​

The same funnel is reachable on its go. subdomain. Keep the go. subdomain crawlable (do not robots-disallow or redirect it — the embed needs it) and let Google consolidate to the apex.

No tool emits a canonical on these routes. Verified on symphonycore.com 2026-10-31: /pricing/ and /platform/ serve no <link rel="canonical"> and no robots meta. LeadConnector serves the funnel HTML itself, so Rank Math never runs on the route. The self-canonical that SearchAtlas OTTO used to inject went with it. Until a canonical is added, consolidation rests on internal links pointing at the apex URL and on the apex being the URL submitted in GSC.

To add one, put <link rel="canonical" href="https://<domain>/<path>/"> in the funnel step's own head tracking code in the funnel builder. ⚠ Not yet verified on an SC site: after adding it, confirm with curl that it appears exactly once on the apex route, and with GSC URL Inspection that the Google-selected canonical is the apex.

5. Sitemap / discovery​

GHL funnel routes are not native WordPress pages, so Rank Math's sitemap does not include them, and Rank Math has no way to add external URLs (the Additional Pages feature belonged to the retired Auctollo plugin; decision D6). Use GSC → URL Inspection → Request Indexing for each apex URL; a sitemap is not required for indexing. Link each funnel route from a WordPress page that is in the sitemap so crawlers discover it. If many funnel pages must be indexed, raise it against D6, whose named mechanism is GHL's own sitemap.

Verification​

  1. Static: curl -s https://<domain>/<path>/ | grep -i 'name="robots"' → no noindex; confirm the MutationObserver guard is present.
  2. Authoritative: GSC → URL Inspection → Live Test → Page availability → Indexing allowed? Yes (this renders JS; curl cannot judge the fix). Then Request Indexing.

When to rebuild natively instead​

Rebuild the page natively in WordPress/Elementor — no GHL funnel embed — when any of these is true:

  1. The GSC render still shows noindex with the snippet and guard in place.

  2. The page needs the global WordPress site nav, which the header toggle can't deliver without breaking layout.

  3. Native HTML Embed is corrupting the served markup. Check before trusting the embed:

    curl -sL https://<domain>/<path> | grep -c '<!--&gt;'

    Zero is required. Non-zero means the delivered HTML is damaged, and no amount of noindex-stripping will make the page rank.

The native rebuild removes the funnel JS entirely and is the guaranteed fix.

Worked example — symphonycore.com, 2026-09-17​

Native HTML Embed re-serialises the funnel HTML and prepends an escaped > to the data of every HTML comment, so Vue's server-rendered anchors <!----> are emitted as <!--&gt;-->. Most close harmlessly. Where the escaping breaks comment pairing, the opener swallows real markup until the next -->.

PageEscaped anchorsSwallowingMarkup buried inside comments
/platform5765137.5 KB, 1,143 elements — including the page's only <h1>
/pricing13,1151,05163.9 KB, 1,157 elements

/platform also serves no DOCTYPE, so it renders in quirks mode.

Hydration appears to repair the DOM for human visitors, which is why this is invisible in a browser and was found only by reading the server response. The same pass hoists <script> elements out of their containers to the end of <body>, which silently breaks any embed script that mounts relative to its own tag. That is how the prospecting audit widget vanished from /audit.

Ruled out — the snippet in §3 is not the cause. /audit was corrupted with that snippet's MutationObserver guard absent, while /pricing and /platform carry it. Corruption tracks the display method, not the snippet. Do not delete the snippet on this theory; it remains load-bearing.

Not established. Whether the escaping originates in the LeadConnector plugin or in a WordPress-side output filter on this site. A filter escaping comments in the plugin's output would produce the same signature. The grep check is valid either way, and the remedy is the same.

Notes​

  • JS-managed robots directives are a documented Google gray area. This guard is verified working (confirmed via GSC Live Test) but is a workaround; the native rebuild is the durable option.
  • Keep one SEO/sitemap source of truth on the site (avoid multiple SEO plugins fighting over robots/ meta/sitemap).

Revision History​

DateVersionChange DescriptionAuthor
2026-06-151.0Initial SOP — verified fix for LeadConnector native-embed noindexSymphony Core Systems Team
2026-09-171.1Added a third rebuild trigger: Native HTML Embed measured corrupting the served markup on symphonycore.com. Worked example with the /platform and /pricing numbers, the one-line check, and the snippet explicitly ruled out as the causeSymphony Core Systems Team
2026-10-311.2SearchAtlas references removed. Step 4: verified that no canonical is emitted on funnel routes now OTTO is gone, with a head-code canonical as the unverified remedy. Step 5: the Auctollo Additional Pages route removed (D6). WPCode recorded as the stack standard's approved exceptionSymphony Core Systems Team