Many paddlesport retailers are small specialists, where the same few people run the shop, the website and sometimes a paddling centre. Their sites tend to grow by addition. A booking plugin arrives for courses, then a delivery quote tool for boats, a colour picker, a finance widget and a hero video of someone running a rapid. Each piece is useful. Over a few seasons they add up, and the slowest part can be the server rather than the page.

This guide explains how Google measures the result and where paddlesport sites tend to lose time, starting with hosting.

How the three scores work

Google’s Core Web Vitals are taken from real visits and judged at the 75th percentile, separately for phones and desktops. In paddlesport terms:

ScoreWhat it covers hereGoodPoor
LCPthe hero photo or video poster of a boat or board appearingwithin 2.5 sbeyond 4 s
INPthe reply to a colour tap, an outfitting choice or typing a postcodewithin 200 msbeyond 500 ms
CLSdelivery quotes and booking calendars moving the page0.1 or underabove 0.25

INP replaced First Input Delay in 2024. The real-visit figures come from the Chrome UX Report, shown in PageSpeed Insights and Search Console. Lab tests such as Lighthouse help find causes, though Lighthouse cannot measure INP and uses Total Blocking Time as a stand-in.

What I measured on UK kayak and paddlesport product pages (October 2026)

On 23 October 2026 I ran five UK paddlesport product pages through PageSpeed Insights, once each, on its mobile setting: an emulated mid-range phone over slow 4G.

PagePerformance scoreLab LCPTotal Blocking TimeLab CLS
Retailer A419.1 s40 ms0.46
Retailer B784.9 s0 ms0.08
Retailer C495.4 s170 ms0.48
Retailer D3415.5 s870 ms0.08
Retailer E1215.6 s980 ms0.85

The middle page scored 41, with a lab LCP of 9.1 seconds. Blocking time was the bright spot, with a median of 170 ms, though two boat pages went past 850 ms. Layout shift was the real weakness. Three pages scored above 0.1 for lab CLS, and one reached 0.85. Lighthouse named layout shift culprits and flagged late discovery of the LCP image on all five, which fits the delivery checkers and late hero images covered below.

A few caveats. These are lab figures from a deliberately slowed phone, so they run far behind most real paddlers’ visits, and Google does not use lab LCP. None of these exact URLs had their own field data in the Chrome UX Report. Four showed only whole-site (origin) data, and two of those sites passed Core Web Vitals; the fifth had no field data at all. Five pages from a varied trade are a small sample, and the table is not a ranking.

Start with the server

Google’s guide to optimising LCP splits loading time into four parts. As a guideline, server response takes about 40% of the budget and the image download about 40%. The two delays in between should each stay under 10%. On a small shop’s site, the server’s share is worth checking first.

The Web Almanac 2024 performance chapter found that only 42% of sites had a good time to first byte on mobile. That is a whole-web figure from 2024, not one for paddlesport shops. Its ecommerce chapter found WooCommerce on 35.8% and Shopify on 19.6% of the ecommerce sites it detected worldwide. Those are home pages crawled in 2024, across all ecommerce, so they say nothing about UK paddlesport shops in particular.

On WooCommerce, an uncached product page is built from scratch on every visit. Plugins for course booking, hire calendars and product add-ons all add work, and some load their scripts on every page.

  • Cache pages at the server or CDN, excluding only the basket, checkout and account pages.
  • Minimise redirects, one of the ways Google’s LCP guide gives for cutting server response time.
  • Keep booking and calendar scripts to the pages that use them. A kayak product page does not need the course calendar’s code.
  • Ignore tracking parameters in the cache, since the LCP guide warns that unique URL parameters can defeat CDN caching.

I cover the WooCommerce side in speed, caching and scripts.

On-water imagery and hero video (LCP)

The same Web Almanac chapter found that 59% of sites had a good LCP on mobile in 2024. Paddlesport home pages often open with on-water video or a wide action photo, both heavy on a phone.

For video, LCP is timed from when the poster image loads or the first frame appears, whichever is earlier. A compressed poster keeps that figure low while the video loads behind it.

Product pages have a quirk of their own. A kayak photographed side-on is a long, thin strip, and on a phone it may cover less screen area than a block of text. LCP counts the largest image or text block, so the LCP element on a boat page can be the description or the title. When that happens, the fixes are different: render-blocking stylesheets, web fonts and scripts that delay the first draw.

When the LCP element is a photo, the standard fixes apply:

  • Put it in the HTML with src and srcset, not injected by a slider.
  • Do not lazy-load it, and give that one image fetchpriority="high".
  • Serve WebP or AVIF, sized to the layout, from the same origin as the page.

Colour pickers and outfitting options (INP)

Plastic boats often come in many colours, including multi-colour blends, and each swatch tap swaps the main image. Outfitting choices such as rudder or skeg, seat upgrade or paddle choice in a package often recalculate the price as well. INP counts each of those taps, and each key press in a postcode box. Hovering and scrolling do not count.

Product add-on plugins can be slow here, because some rebuild the whole options form after every change.

  • Update the selected swatch and the price label first, then do anything slower.
  • Fetch the new colour’s image when the buyer hovers over or focuses the swatch, and keep the old image until the new one is ready.
  • Keep swatch thumbnails small: a 40-pixel swatch does not need a full boat photograph.
  • Load chat, reviews and finance scripts after the page is usable.

Delivery checkers, booking calendars and layout shift

A postcode checker that returns a delivery quote for a boat can cause layout shift. So are embedded course calendars, weather or tide widgets on centre pages, and finance messages under the price. Google’s CLS guidance names resizing third-party widgets, inserted content, asynchronously loaded content, unsized images and web fonts as causes.

Shifts within 500 ms of a tap or key press are excluded, but a courier quote service may take longer than that. If the quote arrives a second after the buyer presses “check”, the shift is counted.

  • Reserve space for the delivery quote, with a “checking” message at the same height.
  • Give embedded calendars and widgets a fixed minimum height.
  • Set width and height on every photo, including long boat profiles.
  • Animate panels with transform rather than by changing their height.

3D boats and outfitting builders

A 3D view helps a buyer judge a boat’s length, cockpit and deck layout. Load it only when the buyer asks, with a still photo in its place, so the photo is what counts for LCP. Long hulls need sensible texture sizes and compressed models. See how a 3D product configurator is built and the kayak configurator.

A test plan for a paddlesport site

  1. Run the home page, a kayak page, a paddleboard package and a course page through PageSpeed Insights, and check server response time in the diagnostics.
  2. Request a product page that is not in the cache and time it. Compare with a second request.
  3. Find the LCP element on a kayak page in Chrome DevTools. Note whether it is the photo or a block of text.
  4. On a mid-range phone, tap through five colours and three outfitting options and note any lag.
  5. Enter a postcode in the delivery checker and watch whether the basket button moves.
  6. List the plugins and scripts loading on a product page, and remove those that belong to bookings or hire.

If you would like the work done, see Core Web Vitals or, for Shopify stores, Shopify speed optimisation. The rest of the series covers selling kayaks and paddlesports online and kayak and paddlesport SEO.

Sources