Foreign trade keyword research has a counterintuitive rule: the keywords with the highest search volume usually aren’t the ones bringing the most inquiries. In the industrial valve category, “industrial valves” gets tens of thousands of monthly searches, but one search simultaneously corresponds to hundreds of suppliers; “stainless steel butterfly valve API609” only gets a few hundred monthly searches, but the people clicking through are a batch of buyers who’ve already written down specs and part numbers. Pick the wrong direction in keyword selection, and the higher you rank, the more you lose. This guide walks from the first step of keyword discovery to final layout, with directly actionable moves at each step.
Quick conclusion: foreign trade keyword research isn’t a one-time keyword find, it’s a weekly iterative cycle: judge intent, dig long-tail, test difficulty, cluster keyword groups, rank priorities, map to URLs, and finally land on the content calendar. The core standard for keyword selection isn’t search volume, it’s how close to a deal. Following this guide through, you’ll get a complete map from candidate keywords to page URLs, plus the first step you can start on this week.
Six stages each handle a segment: judge intent — label each candidate keyword with four-category search intent, intent written into the first column of the keyword library as a filter condition for subsequent screening and sorting; dig long-tail — use Google autocomplete suggestions, People Also Ask, customer inquiry original sentences, and industry forum questions to spread the keyword library to the thousand level, each channel stored separately and source-labeled; test difficulty — combine tool KD with manual SERP verification, look for break-in pages in the top 10 with thin content, young domains, and few backlinks; cluster keyword groups — group by search intent, semantically similar keywords with high result overlap share one page, main keyword goes in title and first paragraph, the rest become H2 sections and FAQ; rank priorities — score and sort by “traffic × difficulty × commercial value” three dimensions, low-difficulty high-value keyword groups done first, big keywords left until authority builds up; map and land — one URL only handles one main keyword and one keyword group, correspondence written into the mapping table, finally scheduled into the content calendar for weekly iteration.
What is foreign trade keyword research
Foreign trade keyword research is a systematic method of analyzing the real words and phrases overseas buyers enter into search engines, judging intent, difficulty, and commercial value, and据此 planning page topics and URL structure.
The biggest difference from domestic search operations is the data source: you’re not serving Chinese expression, but the real question phrasing of overseas markets in English, Spanish, German, etc. — for the same product, buyers in different markets may call it completely differently. The research output isn’t a “keyword list” either, it’s a four-column mapping table of “keyword — intent — difficulty — URL” — it directly determines what the content team writes, how they write it, and where they publish it for the next three months. Three to six months after launch, reconcile this table with Search Console data once, re-route keywords that didn’t rank up — do this step well, and the keyword library gets more accurate with use.
| Process step | Core action | Common tools | Output |
|---|---|---|---|
| Step 1 · Discovery | Collect candidate keywords from GSC, competitor pages, customer inquiries | Search Console, Ahrefs | Thousand-level candidate keyword library |
| Step 2 · Intent | Label each keyword with search intent by four categories | Manual labeling | Keyword table with intent tags |
| Step 3 · Difficulty | Combine tool KD with manual SERP verification | Ahrefs, Semrush | Difficulty tiered list |
| Step 4 · Cluster & map | Group synonyms and near-synonyms, map to URLs | Sheets, clustering tools | Keyword group mapping table |
| Step 5 · Schedule & land | Sort by commercial value, schedule into content calendar | Content calendar, content templates | Publishing schedule |
Search intent: first answer why the buyer searched this keyword
Search intent determines what the page should look like. Under the same product topic, questions starting with “what is” are informational intent, suitable for blogs and explainer pages; those starting with “best X for” are commercial investigation intent, suitable for comparison articles; “buy X” and “X price” are transactional intent, directly give product pages and quote pages. Get intent judgment wrong, and no matter how well the content is written, it won’t retain people — users click in and bounce in two seconds. Labeling each candidate keyword with intent is the first column of data in the keyword library. The four-category judgment standards and landing page correspondence are detailed in search intent four categories; to further map keywords to different stages of the sales funnel, reference keyword intent layering.
In practice, informational intent keywords are the content team’s “granary,” while commercial investigation and transactional intent keywords are the real “grain tickets.” Write intent labels into the keyword library, and every subsequent screening and sorting step can directly filter by label. In reality, it’s normal for 30% of intent labels in a keyword table to be wrong — spending half a day relabeling old keywords is more cost-effective than digging 500 new ones.
Long-tail keyword discovery: first spread the library to a thousand then reel in
The problem with most foreign trade sites isn’t too few keywords, it’s too narrow: digging one page under product keywords and stopping, missing question words, model numbers, regional words, and application scenario words entirely. The correct order is first open up volume then reel in — Google autocomplete and dropdown suggestions, People Also Ask, original sentences repeatedly appearing in customer inquiries, and question titles from industry forums — these channels can lay out hundreds of candidate keywords in a day. Once volume is sufficient, enter clustering and screening, don’t agonize over individual keywords in the early stage. Store keywords from each channel separately and label the source, making it easy to review which channel actually nurtured inquiry keywords. The specific tactics and pitfalls of discovery channels are in how to dig long-tail keywords. Don’t rush to delete keywords showing zero search volume in tools — they might be high-conversion keywords in sampling blind spots, the judgment method is in zero search volume keywords.
Keyword difficulty: manually look at SERP, don’t blindly trust tool scores
The KD given by tools is a probability model, only representing a “difficulty index,” not “whether you can win.” For the same KD 40 keyword, if the homepage is full of old articles from industry media vs. full of independent site product pages, the opportunities are completely different. Manual verification only looks at three things: are there pages in the top 10 with thin content, young domains, and few backlinks — if yes, there’s break-in space; are the result pages content pages or product pages — directly reflecting the search engine’s judgment of intent; how often does the #1 ranking page update — determines whether you need to do time-sensitive content. The pitfalls and correct reading of KD scores are in how to read keyword difficulty; to see a complete year-long research routine for a category, 2026 keyword research in practice has month-by-month cases.
Keyword group clustering: let one page eat dozens of related keywords
Clustering solves the problem of “how many keywords can one page eat.” There’s only one principle: group by search intent. Semantically similar keywords with high search result overlap are grouped into one keyword group sharing one page; keywords with different intent must be split apart. After the keyword group is determined, the main keyword goes into the title and first paragraph, the rest become H2 sections, FAQ, or tables for natural coverage — one page eating dozens of related keywords is more labor-saving than splitting into dozens of thin pages, and authority is more concentrated. A good clustering table should let the content team know which page a certain sentence should be published on without thinking. The specific operation steps of clustering are in using clustering to build keyword groups; if you find multiple pages competing for the same keyword when grouping, re-demarcate according to the思路 of fixing keyword cannibalization.
Priority ranking: use three-dimensional scoring to decide which article to write first
With the keyword library and keyword groups ready, which article to write? Score on three dimensions: “traffic potential × difficulty × commercial value.” Traffic potential looks at the keyword group’s search volume and seasonal trends, difficulty looks at the step 3 evaluation results, commercial value looks at how close this keyword is to a deal — keywords that can attach products, compare prices, and trigger inquiries naturally have higher value. After sorting by the product of the three, quickly win “low-difficulty, high-value” keyword groups first, leave high-difficulty big keywords until the site’s authority builds up. Scoring isn’t pulling numbers out of thin air — weights can be set first then fine-tuned: traffic 30%, difficulty 40%, commercial value 30%, after running two rounds backfill according to results. The weight setting and replicable calculation table for three-dimensional scoring are in keyword priority ranking.
Competitor keyword gaps: find keywords competitors rank for that you don’t
Competitor gaps are the highest-ROI keyword source: the keyword has already been verified that people search for it and people do it, you only need to judge whether you can do it better. The method is to list 3-5 direct competitors, grab their core keywords ranking in the top 20, and do a difference set with your keyword library. From the difference keywords, pick a batch with controllable difficulty and high commercial value to add to the content schedule — much more stable than blindly creating new keywords. Another layer of meaning of gaps is defense: if competitors have seized your brand keywords, you need to lay out brand defense content in advance. The specific method of gap analysis is in competitor keyword gap analysis; the dual-line strategy for brand and non-brand keywords is in brand vs non-brand keywords.
Keyword to URL mapping: each page only guards one main keyword
Mapping is the last checkpoint from keyword library to content, and also the most easily skipped. The rule is simple: one URL only handles one main keyword and one keyword group, and write the correspondence into the mapping table: URL, main keyword, covered keywords, intent, owner, status. When new keywords come in, first check the table to avoid building duplicate pages for existing pages; when two keyword groups collide on intent, merge first then decide the URL. The mapping table can also tell you “where exactly is the page for this keyword” during content updates, saving time searching the whole site. After the mapping table is built, any new keyword entering the library first passes through the table, blocking keyword competition from the source. Mapping rules and field design are in keyword to URL mapping.
Content landing: arrange the keyword library into an executable calendar
Selected keyword groups ultimately become pages. Foreign trade sites commonly have three types of content forms: guide-type long articles eat informational keywords, Q&A pages eat question keywords and顺便 grab featured snippets, product comparison pages eat commercial investigation keywords. Low-difficulty long-tail keywords are suitable for template batch production — fill in the blanks after putting on a unified skeleton, single-article cost kept very low; for categories with strong seasonality, remember to schedule 90 days in advance, don’t wait until search volume rises to start writing. The skeleton structure of long-tail keyword templates is in long-tail content templates; the method of using Q&A structure to grab featured snippets is in Q&A content SEO; the advance layout rhythm for seasonal keywords is in seasonal keyword trends.
FAQ
The following five questions come from the highest-frequency questions from foreign trade operators — conclusion first, then expand.
How often should foreign trade keyword research be done?
Recommend a systematic review once per quarter, with supplementary discovery anytime new products launch or competitors change. The keyword library is a continuously maintained asset, not a one-time deliverable.
Are free tools enough?
Yes. GSC, Google autocomplete, and competitor pages can cover most topics — paid tools just help you save time, they’re not a necessary condition.
Which keywords should a new site do first?
Prioritize low-difficulty, high-commercial-value long-tail keywords, start with KD under 20. Leave big keywords for after the site builds up authority, avoid going head-to-head right out the gate.
Can one keyword correspond to multiple pages?
Not recommended. The same main keyword is only handled by one URL, semantically similar keywords merged into one keyword group — otherwise multiple pages will compete with each other for rankings.
How to judge whether a keyword has commercial value?
Look at whether product pages, price pages, or comparison pages appear on the search result homepage. If they appear, there’s commercialization space — more valuable than pure informational keywords.
Implementation is five steps: open Search Console and export the last 16 months of queries, label each with intent and commercial value, build the first version of the candidate keyword library; grab keywords ranking in the top 20 from 3 direct competitors, do a difference set with your keyword library, list a gap keyword list; do intent four-category and semantic clustering for candidate keywords, map each keyword group to one URL, fill all into the mapping table; sort keyword groups with three-dimensional scoring, schedule the top 5 keyword groups into this month’s content calendar and assign owners; weekly supplement new keywords from on-site search logs and customer inquiries, keeping the keyword library synchronized with real demand.
Figure: Key points of foreign trade keyword research: six-step closed loop from keyword library to pages (compiled by 运营GO)
