Image SEO Checklist

Filenames: describe before upload

Filenames are the first signal crawlers read, and the default camera name — IMG_4821 — says nothing. The fix is mechanical: rename to a short descriptive phrase with hyphens, like rooftop-garden-planter-box.jpg — three to six words naming what the image shows, matching how someone would search for it. Do this at upload time, not after publication; renaming later changes the URL and orients existing links to a redirect or a dead end.

The discipline extends to honesty. Filenames stuffed with unrelated keywords — best-cheap-fast-seo.jpg on a photo of a sandwich — are the spam pattern they have always been, and modern systems discount them. Accuracy is the optimization: the filename describes the image, the surrounding text provides context, and together they tell crawlers what the image is without tricks. For product catalogs, a consistent naming scheme — product-color-angle — doubles as internal organization, which is how a checklist item becomes a workflow asset.

Alt text: function, content, and common failures

Alt text serves three masters simultaneously: screen readers announce it, crawlers index it, and browsers display it when the image fails to load. Good alt text describes the image's content and function in one sentence — 'bar chart of quarterly revenue rising from 2.1 to 3.4 million' beats 'chart' and beats 'beautiful amazing revenue chart of success'. Write what you would say to someone on the phone who cannot see the page.

The recurring failures, in order of frequency: empty alt on meaningful images — invisible content for screen-reader users and crawlers alike; decorative images with alt text that clutters screen-reader output — decorative images deserve empty alt deliberately, not by accident; keyword-stuffed alt that reads like a tag cloud; and alt text describing the file instead of the content — 'photo, jpeg, 800 pixels'. The decision rule: if removing the image would lose information, the alt text carries that information. If it is purely decorative, mark it so and move on.

File weight: the ranking factor visitors actually feel

Page-speed metrics are ranking signals, and images dominate page weight, so image optimization is SEO work, not just performance work. The sequence that matters: resize to display dimensions first — serving 4,000-pixel originals into 800-pixel slots is the most common waste on the web; choose the efficient format — WebP for pages you control, at quality settings verified against visible artifacts; and set budgets — hero images under roughly 200 KB, content images under 100, thumbnails under 50.

The measurement loop closes in the browser: load the page, inspect image transfer sizes in the network tab, and run the speed assessment tool of your choice. Largest Contentful Paint — the time until the main visible content renders — is usually image-bound, which means the biggest image on the page deserves the most optimization attention specifically. One hour of systematic image dieting routinely moves speed scores by tens of points, which is more than most content changes can claim.

Dimensions, responsiveness, and layout stability

Images without declared dimensions load as zero-height boxes that snap open when the file arrives, shoving content downward — the layout-shift problem that speed metrics punish and users hate. Declaring width and height in markup reserves the space before the image loads; the browser calculates the aspect ratio and holds the slot. This is a one-line fix per image with measurable score consequences.

Responsive serving goes further: different devices need different sizes, and a phone should not download the desktop hero. Modern markup — srcset and picture elements — serves appropriately sized files per screen, and the performance gain on mobile is often larger than any compression win. The practical floor: explicit dimensions on every image, responsive variants for anything above roughly 100 KB, and lazy-loading for images below the fold — which defers their download until the visitor scrolls near. The combination reserves space, shrinks transfers, and stabilizes the page in one pass.

Crawlability: sitemaps, lazy-loading caveats, and context

Crawlers must find and understand images, which fails in characteristic ways. Lazy-loading implemented carelessly can hide images from crawlers that do not scroll — verify that deferred images still appear in the page source or use the standard loading attribute, which crawlers handle correctly. Image sitemaps help large catalogs — sites with thousands of product images — get discovered completely. And robots rules that accidentally block image directories remove the entire collection from image search silently.

Context is the signal crawlers weight most: surrounding headings, captions, and paragraph text tell them what an image means more than any attribute does. An image embedded under a relevant heading with a real caption outranks the identical file floating in an unexplained gallery. Captions also earn their keep with readers — they are among the most-read text on any page. The crawlability checklist: images reachable, lazy-loading crawler-safe, context surrounding, and for large catalogs, a sitemap declaring the set.

Licensing and the risk items nobody checklists

The last checklist items are legal rather than technical, and they are where SEO work intersects liability. Every published image needs a provenance: created by you, licensed explicitly, or from a source whose license you actually read — 'free' collections vary wildly in whether attribution is required, commercial use is permitted, and model releases exist for recognizable people. Stock licenses are contractual; a usage that exceeds them — a license limited to web use appearing in print — is a breach even if the image is otherwise perfect.

The protective habits: keep a record of where each image came from and under what terms; prefer originals you control for anything brand-critical; and treat AI-generated imagery under your platform's current disclosure policies. Search-engine image traffic is worth having, but an image lawsuit costs more than the entire SEO program it supports. The checklist ends where it began — deliberately: filenames, alt, weight, stability, crawlability, license. Six items, in publication order, each one cheap.

Running the image SEO audit on an existing site

The checklist becomes powerful when applied as an audit rather than a future promise. The sweep starts with inventory: list the site's images with their sizes, formats, filenames, alt text status, and the pages they appear on. Every metric has a cheap bulk signal — file sizes sort numerically, missing alt text searches textually, filenames reveal themselves at a glance. The sorted lists produce the work order automatically: the heaviest files first, the unmarked images next, the descriptive-naming opportunities last.

The heavy-file triage deserves particular attention because it compounds through every page metric that matters: load time, engagement, and the quality signals built from them. An oversized hero image on a high-traffic page is worth more optimization than a hundred minor fixes in obscure corners, and the audit's sort order surfaces exactly those cases. The fix pattern repeats per file — resize to display dimensions, choose the right format, set quality by viewing contract, re-measure — and each fixed page takes its weight off permanently.

Alt text remediation follows a different rhythm because it is judgment work, not arithmetic. The pattern to install: every content image describes its informational contribution in a sentence; every decorative image gets an empty marker that tells assistive tools to skip it; keyword usage stays natural because the text serves humans first. Finishing the audit means recording the standards where future images get made — upload checklists, CMS field requirements — so the backlog shrinks forever rather than regenerating. The audit is a weekend of work on a mid-size site; the habits it installs are what make it the last such weekend.

Frequently asked questions