ShipworkSite inspection
Image weight

Crawlability

PaginationAre paginated pages set up right?Internal searchIs site search eating crawl budget?URL parametersAre parameters creating duplicates?Log analyserWhat does Googlebot actually crawl?Robots builderNeed a robots.txt file?Soft 404sAre pages “not found” but returning 200?JavaScriptCan crawlers see it without running JS?FreshnessIs my site quietly going stale?Robots.txtDoes robots.txt say what I think?

Indexing

Sitemap lastmodAre my sitemap dates valid?SitemapIs my sitemap actually fetchable?IndexingIs Google allowed to index my pages?CanonicalsIs Google indexing the wrong URL?Index signalsDo sitemap and index signals agree?Sitemap generatorNeed a sitemap.xml for my site?Bulk URL statusWhere does each of these URLs really land?Redirect builderNeed the redirect rules?RedirectsAre my URLs answering directly, over HTTPS?

On-page

Keyword ideasWhat are people searching for?SERP previewHow does my page look in Google?Content qualityAre your pages too thin or too alike?Image weightAre images slowing the page?CannibalizationAre pages competing with each other?On-page checkDoes the page use its target phrase?ImagesAre my images accessible and loading?Meta tag builderWhat should my title and share tags say?DuplicatesDo my pages compete for one query?

Links

Orphan pagesWhich pages can no link reach?Link graphHow deep do your pages sit?Anchor textDo links say what they point at?Outbound linksAre external links still alive?Broken linksAre internal links sending visitors nowhere?

Structured data

Rich resultsIs my markup eligible for a rich result?Structured dataIs my product schema valid?Schema coverageDo my key pages carry structured data?Social previewHow does my page look when shared?Schema vs pageDoes markup match the page price?Schema builderNeed valid JSON-LD?

International

Hreflang sitemapDo page and sitemap hreflang agree?HreflangDo my language versions link back?Hreflang builderNeed the hreflang tags?Page DoctorWhy is this page not doing well?Redirect planMy site has dead links — what redirects do I write?Sitemap diffIs anything missing from my sitemap?Robots simulatorWhat does my robots.txt actually block?Can Googlebot?Can Googlebot fetch this URL?
Every error, explainedGuides
Pricing Learn Error guides Run free audit

+1 free check

Welcome to Shipwork

Sign in free and your next check this month is on us.

Continue with GoogleContinue with email

We email you a 6-digit code. No password.

No card needed. Reports and watches follow you to any device.

Are your images slowing the page down?

Images are usually the largest thing on a page and the easiest to fix. Legacy formats, missing srcset and oversized files are the three reasons a page is slow on a phone while it looks fine on your laptop.

Reads the images on a set of pages, checks format and srcset, and measures a few for weight. Saved to your account if you are signed in to a paid plan.

How this check works

Shipwork parses the images on a set of your pages and reports three things: whether any use a modern format (WebP or AVIF), whether they carry a srcset so phones do not download desktop-sized files, and whether the heaviest ones exceed a sensible size. It reads raw HTML and a few image headers, so it sees what a browser actually has to fetch.

What the image weight check measures

The image weight check reads the image tags on the pages it samples, up to 20 in a free run, and looks at their source addresses and attributes. It counts images served from legacy formats such as JPEG, PNG or GIF when none use WebP or AVIF, images with no srcset, and images with no lazy loading.

It then requests a small sample of up to eight image addresses to read the real content type and size from the response, because a server can deliver a modern format from an older-looking address. Images whose response is larger than 250 KB are listed, so the sample reads the served type rather than trusting the file extension.

How to read an image weight result

A legacy-format warning means the images checked used older formats and none served a modern one by type or address. A missing srcset warning means the page offers the browser only one image size, so a phone may download a desktop-sized file. Heavy images over 250 KB are the clearest lead: they are the ones most likely to slow the largest contentful paint.

These are leads, not proof. The tool probes a limited sample, an image request can fail for reasons unrelated to weight, and an image service may transform an image before it reaches a visitor. The check cannot measure real-user speed; it identifies files worth compressing or converting, which you should confirm in a browser.

Common causes and why they matter

Heavy images usually arrive through bulk upload: full-resolution photographs uploaded straight from a camera, theme assets that were never optimised, and banners exported as PNG. Missing srcset is a template issue, because the same single source is printed on every screen size. Missing lazy loading is common on older themes and page builders.

Image weight is often the largest part of a page's total bytes, so it is a direct lever on when the main image appears and how much data a visitor spends. That matters most on mobile connections. Lighter images, responsive sources and deferred below-the-fold loading all improve the first screen without changing the design.

How to fix heavy and slow images

Compress images to a sensible quality and dimensions rather than uploading full-resolution files, then serve WebP or AVIF where the platform supports it. Add srcset so the browser can choose a smaller file on smaller screens, and lazy-load images that sit below the first screen while leaving the main above-the-fold image eager.

On a store, fix this at the source: set your upload and resize rules so new images are compressed on the way in, and check whether an add-on can convert and resize existing ones. Replacing an image means re-uploading it, so correct the process rather than editing a single product. Re-run the check afterwards to confirm the size falls.

When to use related checks

Use image check for alt text and width and height attributes, which is a different question from weight. Use Core Web Vitals when you want what real visitors actually experienced rather than a file-level estimate.

The Core Web Vitals guide and the crawl budget guide cover how image delivery and page weight interact with speed and crawling. Fix the bytes first, then confirm the effect in real-user data.

Questions

Why WebP or AVIF?
They are 25-50% smaller than JPEG or PNG at the same quality, and every modern browser supports them. Smaller images mean a faster largest contentful paint.
What is srcset?
A list of image sizes a browser can choose from. Without it, a phone downloads the full desktop image, which is slower and wastes data.
How big should an image be?
There is no single number, but product and hero images over a few hundred KB are usually compressible without visible loss.
How do I check image weight on my website?
Run the check on your domain. It reads the images on the pages it samples, flags legacy formats, missing srcset and missing lazy loading, and requests a small sample to read the real type and size. It reports images over 250 KB, but the sample is limited, so inspect your heaviest pages directly too.
Why WebP or AVIF instead of JPEG or PNG?
Modern formats usually produce a smaller file than JPEG or PNG at comparable quality, and current browsers support them. Smaller files mean the main image appears sooner and visitors download less data. The check flags pages where only older formats appear.
What is srcset and why does it matter?
It is a list of image sizes the browser can choose from. Without it, a phone can download a desktop-sized image, wasting data and time. With it, the browser picks a file suited to the screen, which is part of a faster, lighter page.
Does the image weight check measure my Core Web Vitals?
No. It reads image attributes and a small sample of served files, which is a file-level view. Real-user speed comes from field data, covered by the Core Web Vitals check. Use this check to find heavy files, then confirm the effect in real-user data.

Keep reading

What this reads

  • Reads public pages from outside your network, the way a crawler does.
  • No signup, and nothing is stored on the free checks.
  • One of 83 checks across discovery, content, markup, trust and catalogue data.
83checks
0signups needed
0stored on free checks
Cool, just ran an audit and got some insights! Let’s see if I can fix them :)
Tenta Dev, @DevTenta
Whoah! this site is crazy, nothing to roast at all
Ainreek, @AinreekAI
cool as heck, tried it on my blog and the insights it gave are helpful and not bluff.
Refaat 🍉 (Mouse), @1Mouse__