Technical SEO Checklist for Foreign Trade Independent Sites

After six months running a foreign trade independent site — 80 articles, 200 backlinks — organic traffic is stuck around 2,000 per month; a competitor with only 30 articles has three times your traffic. The gap usually isn’t in content volume, it’s in technical SEO — your pages aren’t being crawled by the spider, or crawled but not indexed, or indexed but loading too slow and getting demoted.

Foreign trade technical SEO isn’t about piling up tools, it’s about打通 four chains in sequence: crawling, indexing, understanding, and experience. Prioritize fixing four types of hard issues: robots误伤, missing canonical, redirect chains, and oversized images; then run item by item according to the master table below: mobile experience, CWV, JS rendering, internal link depth — each has a pass standard. Regular sites complete the first round in a month, with an extra check during redesign or migration periods.

1. Crawling chain: verify robots.txt has no accidental blocks, sitemap only includes canonical pages returning 200, use server logs to confirm which URLs Googlebot actually crawled and whether it was rate-limited.
2. Indexing chain: each page keeps only one self-referential canonical, filter and tracking parameter pages normalized to the main version, core pages not accidentally marked noindex, parameter pages blocked by robots.
3. Understanding chain: Article, FAQ, and breadcrumb three types of Schema validation with zero errors, body internal links use real <a> tags, pagination can be crawled layer by layer to the last page.
4. Experience chain: mobile content complete, text readable, click targets no smaller than 48px, LCP under 2.5 seconds, CLS under 0.1, INP under 200 milliseconds.

What is technical SEO

Technical SEO is technical optimization that, by optimizing the website’s technical layer, enables search engine crawlers to smoothly crawl, understand, and index pages, while ensuring a smooth experience for real user visits.

It doesn’t directly produce content, but it determines content’s fate: no matter how good an article is, if it can’t be crawled or indexed, it’s as if it was never written. Before starting, first establish a baseline according to the 12 pre-launch items in the technical SEO audit checklist, then go item by item through the eight modules of this handbook.

The master table below covers eight modules, each row is executable and verifiable, with each section expanded after the table.

Check item	Recommended tool	Pass standard
robots.txt & sitemap	GSC index coverage report	Sitemap submitted, all core pages indexed, no robots rule errors
canonical & duplicate content	Screaming Frog / Ahrefs	Each page has one self-referential canonical, all duplicate URLs normalized to main version
Structured data	Google Rich Results Test	Article, FAQ, breadcrumb Schema validation zero errors
Core Web Vitals	PageSpeed Insights	LCP under 2.5s, CLS under 0.1, INP under 200ms
Mobile adaptation	GSC mobile usability report	No content occlusion, text readable, click targets no smaller than 48px
JavaScript rendering	URL Inspection	Core content visible in raw HTML, crawl and render results consistent
Internal links & navigation	Sitebulb or self-built link graph	Homepage to any core page no more than 3 clicks, no orphan pages
301 & dead links	Crawler logs / site-wide crawl	No 404 on core pages, redirect chains no more than 2 hops

Crawling and indexing: first let the crawler in

Crawling is the entry point of the whole chain. First check robots.txt: are resource directories or admin paths accidentally blocked, and does the sitemap contain URLs that don’t need indexing? Each sitemap entry should return 200 — admin, cart, and filter parameter pages all excluded. After this step, look at logs — logs can tell you which URLs Googlebot actually crawled, how often it comes, and whether it was rate-limited — information the GSC coverage report can’t give you.

After crawling comes indexing. If core pages are stuck at “crawled but not indexed” long-term, first check if the content is too thin, then check if they’re accidentally marked noindex or if parameter URLs are siphoning authority. Crawl budget is only sensitive for sites with hundreds of thousands of pages — regular foreign trade sites just need to block low-value parameter pages in robots.

The specific写法 of robots rules and sitemaps follows robots.txt & sitemap; to see the crawler’s real behavior clearly, set up log monitoring according to log file analysis.

Page technical elements: canonical, structured data, and pagination

canonical is the first move for treating duplicate content. Filter parameters, tracking parameters, and currency switching all create URLs with identical content — each page keeps only one self-referential canonical, consolidating authority to the main version, don’t let the crawler waste effort on pointless duplicate crawls.

Structured data uses JSON-LD placed in the body area — foreign trade sites prioritize three types: Article, FAQPage, and BreadcrumbList. Validate with the Rich Results Test tool every time you change something, go live only with zero errors — errors hurt more than not doing it at all.

Pagination and infinite scroll are the easiest ways to make the crawler “not crawl everything.” Pagination pages must be crawlable layer by layer through links to the last page; infinite scroll lists must keep all entries in HTML, not just exist in JS memory.

The canonical troubleshooting process is in canonicalization (canonical); the structured data field checklist can be directly copied from structured data.

Performance and Core Web Vitals (CWV)

Google uses three metrics — LCP, CLS, INP — to score page experience. If all three fail, rankings will consistently suffer for content of the same quality. First test mobile with PageSpeed Insights — 90% of foreign trade site bottlenecks concentrate in three places: uncompressed images, no caching enabled, and server too far from visitors.

Fix in order of ROI: first compress images, enable gzip and browser caching (results in a day or two), then add CDN to push static resources to the node closest to users. After changes, must retest — the acceptance line is LCP under 2.5 seconds, CLS under 0.1, INP under 200 milliseconds.

The error meanings and item-by-item fixes for the three metrics are in Core Web Vitals (CWV); if you don’t want to troubleshoot from scratch, follow the execution order of the site speed optimization three-move combo.

Mobile-first indexing

Google uses the mobile page as the indexing baseline. No matter how complete the desktop version is, if the mobile version has missing content blocks, font size too small, or buttons covered by popups, indexing and rankings will suffer连带. The verification method is straightforward: open each core page on a phone — is any information missing, can you smoothly complete one inquiry path?

Two high-frequency failure points: first, mobile uses JS to collapse the body, so the mobile version the crawler catches is empty; second, interstitial popups cover the first screen. When mobile readability fails, first fix the structure, don’t patch by piling on code.

Self-check item by item against the mobile-first indexing: 9 checkpoints for mobile experience checklist — more efficient than guessing one by one against GSC reports.

JavaScript SEO

As long as core content depends on JS rendering, there’s a risk of “the crawler sees an empty page.” Google can execute JS, but it pays two costs: new content queues into the rendering queue, slowing indexing speed; secondary crawls consume server resources. Server-side rendered sites like WordPress are usually safe — SPAs and headless architectures need focused troubleshooting.

Troubleshooting method: in URL Inspection, compare “view crawled page” with the rendered DOM, confirm title, body, and internal links are all in the raw HTML. Critical content must never exist only in JS variables or API responses.

SPA改造 paths, how to choose between SSR and pre-rendering — see JavaScript SEO.

Site architecture and navigation

Architecture determines how authority flows. The most common mistake on foreign trade sites is a two-layer structure: all product pages hang directly under the homepage, click depth is only 1, but with too many pages, homepage authority gets spread like “one spoon divided among a hundred bowls.” The correct approach is a flat architecture within three levels: homepage → category page → product page, using breadcrumbs and sidebar navigation to pass authority down level by level.

Two details most easily missed: each filter click generates a new parameter URL — unhandled, it eats up crawl budget; orphan pages have no internal links pointing to them, so crawlers can hardly find them.

The authority distribution logic and navigation改造 examples are in flat architecture & navigation; the写法 for preventing parameter page runaway is in faceted navigation SEO.

Image optimization

Images account for the bulk of foreign trade site page weight — an uncompressed product image can easily be 1MB. Do three actions at once: rename files to descriptive English (cargo-pants-waist-adjustable.jpg not img_231.jpg), write alt clearly describing product key points, compress to WebP and under 200KB.

Image hosting and CDN can automatically handle cropping and format conversion, but the前提 is the system will add alt and descriptive filenames to images — otherwise you’ve only optimized file size, and half the SEO value is still missing.

Naming conventions, alt写法, and the full compression tool workflow — see image SEO optimization.

Migration and multilingual

Changing domains, changing CMS, or modifying directory structure all count as site migration — the easiest thing to lose is link assets. There’s only one principle: all old URLs 301 to equivalent new pages, equivalent where possible, those that can’t be equivalent go to the nearest category page — never land on 404, and don’t make long chains of 301 followed by 301, max two hops.

Multilingual sites use hreflang to declare language and region relationships. Common errors: hreflang pointing back to itself, language groups not linking back to each other, only written in HTML but not synced into XML sitemap. Done wrong, multilingual pages get judged as duplicate content.

The complete pre-during-post migration operation checklist follows the site migration SEO checklist; the multilingual hreflang matrix写法 is in hreflang multilingual/multi-region deployment pitfall avoidance guide.

FAQ

How often should technical SEO be checked?

Do a full audit before and after launch; for daily operations, spot-check 5 core pages monthly, run the full checklist quarterly, and add an extra check during redesign or migration.

How to make pagination pages fully crawlable?

The crawler can follow links from the first page to the last, each page has an independent title; infinite scroll must keep all entries in HTML. Details in pagination & infinite scroll SEO.

Does crawl budget need dedicated optimization?

Foreign trade sites with under 50,000 pages don’t need to worry — just block filter parameter pages in robots; after scale grows, optimize systematically, reference crawl budget optimization.

What to do if rankings drop after site redesign?

All old URLs 301 to equivalent new pages, don’t leave 404s; use GSC to submit the redesign and observe index recovery. Details in dead links & redirect chain cleanup.

Next steps

• Today: open GSC index coverage report, export “not indexed” URLs, group by page type.
• This week: run a full audit on the homepage and 3 core category pages using the master table at the top, mark failing items in red.
• After fixing robots, canonical, and redirect three types of hard issues, use URL Inspection to request re-crawl for core pages one by one.
• Every last Friday of the month, spend 1 fixed hour reviewing CWV and index coverage, paste data into a shared sheet.