Core Web Vitals for B2B: Treat Speed as a Sales Tool

Core Web Vitals measure what users experience: how fast content appears, how responsive the site feels, how stable the layout is. For B2B, they affect rankings and whether a buyer stays long enough to request a quote.

A slow service page is a sales problem. The champion opened you on a phone. The hero was blank for four seconds. They left. That never shows up as “lost deal” in the CRM. It shows up as a bounce on /services/web-design.

The Three Metrics

LCP — Largest Contentful Paint

How long until the main content is visible. Target: 2.5 seconds or less (good). 2.5–4.0s needs work. Over 4.0s is poor.

On B2B pages the LCP element is usually the hero image, the H1, or a background video. Video as LCP is an own-goal. A compressed image plus a real heading is easier to get under 2.5s.

Slow LCP means they never reach problem, solution, or proof. You paid for the click. They saw a white screen.

INP — Interaction to Next Paint

How quickly the site responds to clicks and taps. Target: 200ms or less. 200–500ms needs work. Over 500ms is poor.

This replaced FID. It covers the whole visit, not just the first click. Laggy nav, a form that freezes on submit, a calendar embed on the main thread — buyers read that as sloppy.

Forms are the INP kill zone. Heavy client JS and chat widgets sit on the main thread. The tap on Send should not wait on a chatbot.

CLS — Cumulative Layout Shift

How much content jumps while loading. Target: 0.1 or less. 0.1–0.25 needs work. Over 0.25 is poor.

Classic B2B CLS: a cookie banner, a late webfont, an injected HubSpot form, or a chatbot popping over the CTA. The buyer goes to tap Request a quote. The layout shifts. They tap the wrong thing or give up.

Reserve space for images (width/height or aspect-ratio). Load fonts so fallback metrics don’t explode the heading. Don’t inject embeds above existing content.

Why B2B Sites Struggle

• Heavy CMS themes — WordPress page builders ship CSS and JS for every widget. Most of it isn’t on the page.
• Unoptimized media — 2MB hero PNGs, autoplay background video, icon fonts for five icons.
• Third-party scripts — Analytics (fine, if tagged well), chat, ad pixels, heatmap tools, three tag managers. Each one can wreck INP.
• Complex marketing pages — Carousels, counters, scroll libraries. Decorative motion on a quote page.
• Legacy hosting — PHP origin, no CDN, TTFB already 1.5s before the image starts.

Ecommerce has the same physics. B2B adds long pages and forms that must work. You can’t “fix SEO” with title tags if LCP is 5s.

The Fix (In Order)

1. Measure the real site — PageSpeed Insights and CrUX for field data. Lighthouse for lab. Lab-only 99s that fail in the field mean you optimized a cable connection, not your buyers.
2. Fix LCP first — Compress and size the hero. Preload it if it’s the bottleneck. Cut render-blocking CSS. Get TTFB down (CDN, static or edge HTML). Don’t start with a footer icon.
3. Then INP — Remove unused JS. Delay chat until idle or first tap. Keep the contact form boring. Hydrate only the islands that need it (Astro helps).
4. Then CLS — Image dimensions, font strategy, no late banners above the fold without reserved space.
5. Re-measure in production — Search Console → Core Web Vitals. Staging Lighthouse is not the score Google uses.

Prioritize money templates:

• Homepage
• Top service pages
• Industry pages
• Contact

A perfect blog post and a 6s quote page is a bad trade.

What Speed Buys You

Faster pages rank more easily when quality is similar. More important: more people reach the form. Track that with form_submit / generate_lead, not with a vanity Lighthouse screenshot.

Don’t invent “we raised conversion 37% by improving LCP.” Improve the metric, then look at leads on the same URLs. That’s the sales argument.

---

We build B2B sites that hit 90+ Lighthouse scores. Plan your rebuild.