Across the whole site, real Chrome visitors pass Core Web Vitals (LCP 2.17 s, INP 184 ms, CLS 0.09), and LCP got 17% faster. The homepage is the exception: real visitors see LCP 3.0 s, INP 328 ms and 2.0 s before the first byte arrives, while Google's lab test got the server response in 10 ms. That gap points at what happens before the page arrives (redirect hops, worth about 0.75 s in the lab, and HTML that is not served from cache for real visitors) and at 962 KB of JavaScript, about 460 KB of it unused, of which the testing tool alone is 318 KB. The largest element is a line of text, so this is not an image problem. For an agency that sells websites, SEO and CRO to mid-size businesses, the homepage is the page prospects judge first and where proposal requests start.
First action
Trace the redirect hops and cache status of the homepage from a cold mobile connection, with and without campaign parameters.
How to measure it
Homepage field LCP and INP (p75)
- Now
- –
- Target
- LCP under 2.5 s and INP under 200 ms
- By
- 60 days
Steps
Trace the hops
Map every redirect from webfx.com, http and tagged campaign links to the final URL, and check the cache status of the homepage HTML.
Web Development Lead, Interactive 1 day
Make it one hop
Point every variant straight at https://www.webfx.com/ and update ad, email and sister-site links.
Web Development Lead, Interactive 2 days
Cache the HTML
Serve the homepage from the edge cache for all visitors; load visitor-specific content after the first paint.
Web Development Lead, Interactive 3 days
Paint the text first
Preload the hero font, inline critical CSS and remove render-blocking files.
Front-End Developer 2 days
Gate the scripts
Load the testing tool only during live tests, defer other tags until interaction and drop unused JavaScript.
Front-End Developer 1 week
Watch the field data
Track the homepage in PageSpeed and Search Console weekly until it passes.
Jr. Web Marketing Specialist - Analytics Track weekly
Publish the proof
Before-and-after speed and proposal-request rates as a WebFX case study.
Director of Marketing 1 week
Why it works
Collapse the redirect chain so webfx.com, http and tagged campaign links land on https://www.webfx.com/ in one hop. Serve the homepage HTML from the CDN cache for every visitor and move any visitor-specific content to load after the first paint. Preload the font used by the hero paragraph, inline critical CSS and remove render-blocking files. Load the testing tool only while a homepage test is live, defer other tags until interaction, and split out the unused JavaScript. Watch the homepage's 28-day field data until LCP is under 2.5 s and INP under 200 ms, then publish the before and after as a WebFX case study.
Prospects reach the homepage from brand search, ads, review sites and the sister brands, often on mobile; two seconds before the first byte is time they spend deciding to leave. The rest of the site already passes, so this is one page of work on the highest-value traffic, and the result doubles as proof for WebFX's own web design and CRO pitch.
Ready to use
ChecklistHomepage speed checklist
[ ] One redirect hop from webfx.com, http and tagged links [ ] Homepage HTML served from edge cache [ ] Visitor-specific content loads after first paint [ ] Hero font preloaded, critical CSS inlined [ ] Testing tool loads only during live homepage tests [ ] Other tags deferred until interaction [ ] Unused JavaScript split out [ ] Field LCP under 2.5 s and INP under 200 ms (28-day)
Risks and how to handle them
Caching the HTML breaks personalisation or a live test
List live tests first and move visitor-specific content to load after the first paint.
Changing redirects drops campaign tracking
Test that UTM parameters and ad click IDs survive the single hop before switching.
Evidence
- Google PageSpeedHomepage, real Chrome visitors (tested November 19, 2026): fails Core Web Vitals on LCP and INP, with LCP 3.0 s, INP 328 ms, CLS 0.07 and time to first byte 2.0 s, while the lab test measured the server response at 10 ms. Lab mobile score 81 (desktop 82). The largest element is a text paragraph ("For marketing teams who need to prove revenue impact, not just report performan..."). Top fix: "Avoid multiple page redirects" (about 0.75 s), then render-blocking requests (0.25 s). 1,808 KB over 163 requests, 962 KB of JavaScript with about 460 KB unused; total blocking time 519 ms.Source
- Google PageSpeedWhole site, real Chrome visitors: passes Core Web Vitals, with LCP 2.17 s, INP 184 ms and CLS 0.09, and LCP 17% faster than the previous period. The homepage is the exception.Source
- Google PageSpeedHomepage third parties: 10 services (Google Tag Manager, Google Analytics, Google Ads, VWO, G2, Sentry, Nutshell and others) add 360 KB, of which the VWO testing tool is 318 KB over 9 requests. Accessibility score 87 with five issues: no main landmark, insufficient colour contrast, headings out of order and two ARIA attribute errors; one SEO issue, links without descriptive text.Source