Competitive research / supplied evidence
Competitor Keyword Gap Tool
Compare organic ranking exports for your website and relevant competitors. Find terms absent from your supplied records, inspect keywords where another site has a better position, and review the pages behind those differences.
Bring your own CSV or TSV exports, or paste records into the tool. This workflow compares supplied evidence. It does not retrieve competitor keywords, contact a keyword database, fetch current search results or crawl the ranking pages. You choose the source, market and snapshot that the comparison represents.
A useful gap report starts with a clear question: what should your audience be able to find on your website? Competitor coverage can help you investigate that question. It cannot decide which topics belong in your business, prove that a page deserves to rank, or predict the traffic an edit will earn.
01 / Set up a useful comparison
Same market. Same snapshot.
A question worth answering.
Prepare exports that answer the same question
Use one provider for your target and every competitor. Match the country, device, search engine and ranking snapshot date. A desktop report for one country should not be treated as an equivalent observation to a mobile report for another country.
Choose organic ranking records with integer positions. Google Search Console average position describes an aggregate and is unsuitable for this comparison, even when a displayed average happens to be a whole number. Keep Search Console performance analysis in its own workflow.
Check the export settings before uploading. A complete provider export, a top-ten filter and a small sample have different coverage. Record which you have. If you cannot establish the coverage, leave it unknown and make your conclusions narrower. An absence from a short file is weak evidence about a website’s wider visibility.
The SEO tools hub provides related utilities for investigating questions that an export comparison leaves open. Start with the data you have, then choose the next check according to the decision you need to make.
Choose competitors and a sensible scope
Compare your target with up to four competitors serving a similar audience. Include sites whose ranking pages you can learn from, rather than selecting only the largest domains in your industry. A broad publisher may cover thousands of subjects unrelated to your offer.
Choose domain scope when subdomains belong in the review. Choose exact hostname when a blog, shop or regional host needs separate treatment. Choose exact page URL when comparing specific pages. The ranking URL column is required for an exact-page comparison; domain comparisons can use keyword and position columns without it.
Use equivalent scopes across the comparison. Comparing one focused service page with an entire competitor domain can reveal many unrelated terms and make the opportunity list harder to interpret. Confirm the page type and audience before calling that difference a content gap.
When format is uncertain, use the Search Intent Checker to review a supplied search snapshot. A query may call for a product category, comparison, guide or interactive tool. Its wording alone does not settle the choice.
02 / From records to review
Build the comparison
- Enter your target. Select the scope that matches the website, hostname or page you intend to review.
- Add competitors. Use one to four relevant competitors and collect comparable organic exports for each.
- Import the records. Upload CSV or TSV, or open the paste area for the corresponding site. Keep each site’s evidence separate.
- Review the mapping. Map keyword and position. Add ranking URL and volume when your export provides them. An exact-page comparison needs the URL mapping.
- Record the context. Enter provider, country, snapshot date, device and search engine. Identify full, filtered, sample or unknown coverage as accurately as you can.
- Check the record count. The form displays a maximum of 10,000 combined records. This is a shared comparison limit, not 10,000 records for every competitor.
- Complete human verification. The displayed code expires after ten minutes and works once. Request a new code when needed.
- Compare and inspect. Read the categories with their definitions, examine the supplied URL evidence, then export the rows needed for your review.
The target form includes a declaration for an export containing no keyword records. Use that only when it accurately describes the evidence you supplied. A deliberately empty target and a file that failed to import should never be treated as equivalent situations in your analysis.
03 / Presence is not performance
Understand the categories before counting opportunities
| Category | What it means here | What to check next |
|---|---|---|
| Missing | A competitor has a known supplied position and the target has no keyword record. | Confirm export coverage and whether an existing target page already answers the topic. |
| Behind | At least one competitor has a better known supplied position than the target. | Compare the actual pages and the usefulness of their answers. |
| Shared | The term appears in the target’s records and at least one competitor’s records. | Review the ranking evidence; shared can also be behind or ahead. |
| Target only | The term appears only in the target’s records. | Assess relevance and whether competitors’ exports cover that topic. |
| Unknown position | A record exists, but a usable position is unavailable. | Preserve that uncertainty instead of assigning an invented rank. |
These categories do not form a set of separate totals that can always be added together. A behind keyword can also be shared. A missing keyword means missing from the target export, not that the site has never ranked for the query.
Repeated keywords use the best supplied position for each site. Review the best-ranking URL and other supplied pages together. Multiple URLs are a reason to investigate; they do not by themselves establish that your pages compete harmfully. The Keyword Cannibalization Tool offers a separate query-and-page review when that becomes the next question.
04 / Put the definitions to work
Read a small worked example
The table below is manually authored demonstration evidence, not the tool’s Example payload, an observed search result or a generated report. Dashes mean no keyword record; “unknown” means a record exists without a position. All sites are assumed to use the same provider, country, device, engine and snapshot date.
| Keyword | Target position | Competitor A | Competitor B | Review interpretation |
|---|---|---|---|---|
| study planner | 18 | 7 | 9 | Shared and behind; inspect the target page before creating another. |
| revision timetable | — | 5 | 8 | Missing from this target export; confirm coverage and relevance. |
| exam checklist | 4 | 11 | — | Shared; target is better placed than the one supplied competitor. |
| course notes | 9 | — | — | Target only within these supplied records. |
| learning template | unknown | 6 | — | Shared with an unknown target position; not missing. |
| study calendar | 9 | 9 | — | Shared with equal known positions; no rank advantage in this pair. |
In this illustration, “study planner” provides a useful starting point for page review. The competitor positions do not tell you why those pages rank better. Read the pages, identify differences that matter to a reader and decide whether an edit belongs on your existing page.
If the supplied volumes for that term are 100 and 120, with another value blank, the observed supplied range is 100–120. The blank is not zero, and the range is not a new search-volume estimate. Disagreement should remain visible so a reviewer can inspect the source values.
05 / Make a useful next decision
Choose the page.
Keep the evidence attached.
Choose work your readers will benefit from
Begin with relevant shared terms where your supplied position is 11–20 and a competitor is in the top ten. This is a review heuristic, not a promise of an easy ranking improvement. The right change might be a clearer explanation, an original example, a useful comparison or a more direct path to completing a task.
Next, inspect missing terms appearing across several suitable competitors. Remove unrelated brands, incompatible products and topics outside your audience. Several competitors covering a subject can justify investigation; it does not oblige you to publish a page for every wording variation.
Group related questions around the job a page should do. One strong guide may answer several closely related queries. A category page may serve a different need from a tutorial using similar words. Record the intended destination and reason for the decision in the Keyword Mapping Tool before assigning content to a writer.
Avoid copying a competitor’s outline without evaluating it. Your page needs information, evidence or functionality that helps its own visitor. More words, more repeated phrases and more pages are not substitutes for a useful answer.
Keep ranking evidence separate from assumptions
The tool normalizes keyword case and spacing while keeping accents and different word forms distinct. Treat that as record matching, not semantic clustering. Two phrases can be related in meaning without being the same comparison key.
Mapped ranking URLs are checked against the selected domain, hostname or page, and excluded rows are reported. Those checks concern the supplied addresses and chosen scope. They do not establish that a page currently exists, returns a particular status or is indexed by Google.
If a target page appears absent because of a possible technical restriction, investigate it with the Indexability Checker. Keep that technical evidence alongside the gap review. A missing export row does not diagnose a robots rule, canonical choice or indexing problem.
Reported volume helps sort an already relevant list. The tool does not estimate keyword difficulty or supply a traffic forecast. It also does not identify search intent automatically. Keep business relevance, source coverage and actual page usefulness ahead of a large number in a spreadsheet.
06 / Carry the context forward
A focused work queue.
A complete evidence record.
Export a focused work queue
Use CSV for the currently filtered rows you want someone to review. Download the full JSON report when you need to retain the wider source context and ranking evidence. These downloads serve different purposes: a selected work queue should not be mistaken for the entire comparison.
Record the export date and the ranking snapshot date separately in your project notes. An old ranking observation does not become current because a report was downloaded today. If you repeat the exercise, align the new exports before interpreting differences as movement.
Assign a concrete next step to each chosen topic: inspect an existing page, verify coverage, review the search results, update a section or propose new content. A short queue with clear evidence and ownership is easier to use than a long list labeled “opportunities.”
Data handling and practical limits
The published tool policy says the plugin does not store imports or reports. Shared abuse protection retains a salted IP hash, tool identifier and usage date. Hosting and security-provider policies apply separately. This statement does not establish that all processing happens in your browser or that the surrounding website records no information.
Download any report you need before clearing the comparison. Keep source files and permission to use them within your own project process. The tool compares what you supply; it cannot recover an export you no longer have or fill gaps in a provider’s keyword coverage.
Before you compare
Questions about the comparison
Sources, scope and what a supplied ranking difference can tell you.
Does entering a domain retrieve its keywords?
No. You provide the ranking exports. The tool does not connect to Ahrefs, Semrush or another provider, and it does not fetch a live results page. A domain identifies the comparison scope; it is not authorization or a request for automated collection.
Can I combine exports from several providers?
Use the same provider and comparable settings for a useful comparison. Different databases, refresh schedules and position conventions can produce apparent gaps that reflect the sources rather than a real difference between websites.
Can I use Search Console if I round its positions?
No. Rounding an average does not turn it into a single organic ranking observation. Use compatible ranking exports here and evaluate your own Search Console performance separately.
Should every missing term become a new page?
No. Check the export’s coverage, topic relevance, search intent and existing pages first. Several phrases may belong on the same page, while another phrase may call for a different type of resource or no work at all.
Does a better competitor position explain the cause?
No. A position is an observation in the supplied dataset. It does not isolate the effect of content, links, technical changes or any other factor. Inspect the underlying pages and use additional evidence before choosing an action.
Is a keyword gap the same as a backlink gap?
They answer different questions. This comparison concerns supplied ranking terms and pages. Use the Backlink Gap Tool when your investigation concerns referring-source overlap. A keyword difference alone does not prove that a page needs additional links.