Core Web Vitals: why speed directly affects sales

Core Web Vitals is a set of three key metrics Google defined to measure real-world user experience, and — contrary to what most e-commerce managers assume — it isn't just "the tech team's problem." These metrics affect your search ranking on one hand, and on the other, they determine whether a visitor stays on your site and completes checkout. In this article we break down LCP, INP and CLS along with their threshold values, show speed's impact on sales in hard numbers, and lay out improvement steps tailored to e-commerce sites.

What is Core Web Vitals, and why does it matter?

Since 2021, Google has used page experience as a ranking signal, defining that experience through three measurable metrics: loading speed (LCP), responsiveness (INP), and visual stability (CLS). Together these three are called Core Web Vitals. What they have in common is that they're measured on real users' devices, not in lab conditions — reflecting the experience your customer actually has on a bus, with a mid-range phone, over a spotty connection.

For e-commerce, the importance is twofold. First, visibility: between two pages of comparable content quality, Google tends to favor the one with a better experience, and this directly affects your category pages' organic traffic. Second, and bigger, is conversion: a slow, unresponsive, jumpy page loses the sale before the visitor even sees the product.

The three metrics and their thresholds

Google defines three bands for each metric — "good," "needs improvement," and "poor." Evaluation is based on the 75th percentile of user measurements: if at least three-quarters of your users fall in the "good" band, the page passes for that metric.

LCP: Largest Contentful Paint

LCP is the time it takes for the largest visible element on the page — in e-commerce, typically the product image or hero banner — to render. It's the metric that shapes the "the page has loaded" feeling. Under 2.5 seconds is the target; the usual culprits are unoptimized large images, slow server response (TTFB), and render-blocking CSS/JS.

INP: Interaction to Next Paint

INP is the metric that replaced FID in 2024, measuring how quickly the page responds visually to every interaction — tapping "add to cart," selecting a variant, opening a menu. Delays above 200 milliseconds create a "stuck" feeling. The biggest enemy is heavy JavaScript that locks up the main thread; on e-commerce sites this usually means third-party marketing scripts.

CLS: Cumulative Layout Shift

CLS measures unexpected jumps in content while the page loads. You're about to tap "Add to Cart" when a late-loading banner slides in and you tap the wrong thing — that's the scenario that tanks both your CLS score and your customer's patience. The main causes are images without specified dimensions, late-loading ad/campaign slots, and custom fonts that load in afterward.

Speed's impact on conversion: what do the numbers say?

The relationship between speed and sales has been measured consistently for years, and the results confirm intuition:

• When load time goes from 1 second to 3 seconds, the probability of an immediate bounce rises by 32%; at 5 seconds that increase reaches 90%.
• About half of mobile users abandon a page that doesn't load within 3 seconds.
• According to Deloitte's cross-industry research, even a 0.1-second speed improvement boosted conversion on retail sites by up to 8%.
• Industry case studies report double-digit conversion increases for e-commerce sites that moved LCP from the "poor" band into the "good" band.

The logic behind these numbers is simple: slowness levies a small abandonment tax at every step of the checkout funnel. If the category page is slow to open, users never reach the product; if variant selection lags, nothing gets added to cart; if the checkout page is heavy, the order never completes. Speed alone doesn't guarantee sales — but when product, price and trust are already in place, it creates a multiplier effect. For the other links in the funnel, see our article on 7 proven ways to increase conversion rate.

Speed's impact on SEO: indirect but cumulative

Core Web Vitals is a ranking signal, but its role needs to be framed correctly. Google has explicitly said that experience signals don't override content quality: a great-experience empty page won't outrank a slow but highly relevant one. The signal is what tips the balance among competitors of similar quality — and in e-commerce, category pages are exactly in that kind of competition, with dozens of sites listing nearly identical products under similar titles.

The indirect effects are often bigger than the direct signal. A slow page increases bounce rate as users go back to the search results, and this behavior pattern gradually lowers the page's perceived quality over time. Crawl budget is also affected by speed: Googlebot crawls slow-responding sites with fewer pages, which for large catalogs means new products get indexed later. We covered growing organic traffic through category pages in detail in our e-commerce SEO guide.

Lab data or field data?

The most common confusion in speed measurement is: "My PageSpeed score is 90, but Search Console still says poor." That's because these are two different types of data.

Lab data (Lighthouse) tests your page once in a controlled environment, using a standard device and connection profile. It's excellent for diagnosis: it shows, line by line, which image is oversized and which script is blocking. Field data (CrUX — Chrome User Experience Report) is measurements from real Chrome users who actually visited your site over the past 28 days. This is the data Google uses for ranking — not the lab score.

The gap between the two carries information: if lab is good but field is poor, your real users are on weaker devices and connections than your test profile — very common in Turkey's mobile-heavy traffic. The right workflow is: set your targets based on field data, and use lab tools for diagnosis and verification. Since field data is a rolling 28-day window, it takes about a month for an improvement to fully show up in the report — so be patient.

Improvements specific to e-commerce

Instead of the generic "make the site faster" advice, focus on the four areas that pay off the most on e-commerce templates. The order of work matters too: improve the templates that drive the most traffic and revenue first (usually product and category pages); the homepage may look like the most prestigious page, but it's rarely the real bottleneck in the purchase journey. Since more than three-quarters of traffic in Turkey comes from mobile, run all your tests on a mid-range Android profile, not desktop.

Image optimization

Most of an e-commerce page's weight is images. Switch to WebP/AVIF, serve correctly sized images for each breakpoint (responsive images), and lazy-load off-screen images — but never lazy-load the main product image that's your LCP candidate; load it with priority instead. Specifying width/height on every image also closes off the biggest source of CLS.

Third-party script audit

Between analytics, marketing pixels, live chat and heatmaps, the average e-commerce site carries dozens of third-party scripts, and they're public enemy number one for INP. Run an inventory every quarter: remove what's unused, defer the rest until after interaction, and move to server-side tagging where possible. Removing a single bloated marketing script can win you more than weeks of code optimization.

CDN and caching

Serving content from the point closest to the user lowers both TTFB and LCP. A CDN for static assets is standard now; the real difference comes from smart caching of category and product page HTML output. Architectures that separate out dynamic parts like stock and price and serve the page skeleton from cache keep field LCP in the "good" band even during campaign traffic spikes.

Critical CSS and the render path

Inlining the CSS needed for the first viewport and deferring the rest breaks the render-blocking chain. Loading custom fonts with font-display: swap and holding the place with a system font also gives both LCP and CLS room to breathe.

"Speed isn't a campaign, it's a hygiene standard: not a score you hit once and walk away from, but a budget you have to defend again with every new feature."

Measurement tools: what should you track with what?

To avoid getting lost in a crowd of tools, be clear on each one's role:

1. Search Console – Core Web Vitals report: Shows the overall state of your site's field data and which groups of pages have issues; this is your main dashboard for tracking.
2. PageSpeed Insights: Shows field and lab data side by side for a single URL; start your diagnosis here.
3. Lighthouse / DevTools: Your tool for deep diagnosis and verifying changes during development.
4. Real user monitoring (RUM): Unlike CrUX's 28-day lag, web-vitals measurement you add to your own site shows problems the same day and provides segment breakdowns (device, page type, country).

A cadence to follow: check your RUM dashboard weekly, review Search Console monthly, and run Lighthouse on critical templates before every major release. That way regressions get caught before customers ever notice. Also avoid reducing measurement to a single overall score: a "get PageSpeed above 90" target can push teams toward fine-tuning that raises the score but doesn't touch the user. The right targets are always metric- and template-specific — for example, "bring field LCP on the mobile product page down from 3.1 seconds to 2.4 seconds." A target like that both makes progress measurable and lets you put the improvement's revenue impact side by side with conversion data.

Conclusion

Core Web Vitals isn't a technical exam — it's your customer's patience, quantified. LCP measures "how fast do I see it," INP measures "how fast do I get a response," and CLS measures "is it jumping around in front of my eyes" — and the answers to these three questions affect both your Google ranking and your revenue. The good news: most speed problems in e-commerce come from a known handful of patterns, and with the right infrastructure, most of them never happen in the first place. You can review which performance layers come ready-made with your infrastructure choice on our solutions page, and see what each package covers on our pricing page.

Completing this checklist on your current infrastructure can take weeks; on Şimşek Software's infrastructure, layers like image optimization, CDN, smart caching and critical CSS come standard — our customers' stores go live already targeting the "good" band in field data. To map out your own site's Core Web Vitals report card together and see the revenue you're losing to speed, request a demo; let's start with measurement.