# Keyword, Competitor and Backlink Explorers These three workspaces add external research to the local operator edition. They use your own encrypted DataForSEO connection. No provider account, balance, subscription, ranking guarantee or paid request is included. No private analytics or page content is sent: each request carries only the explicit research query and its displayed scope. Search results remain distinct from GSC/GA4 measurements. ## Workspaces **Keyword Explorer** has four modes: a bulk overview (up to 20 keywords, one per line), seed-keyword suggestions, related searches and topic ideas. Where supplied, rows include estimated monthly volume, CPC in USD, paid competition, organic keyword difficulty, intent, provider update time and up to 36 months of volume. Local filters cover text, minimum volume, maximum difficulty, intent and English question keywords. Missing values remain unavailable; zero stays zero. **Competitor Explorer** has domain overview, ranked organic keywords, top pages, search competitors and keyword gaps. A gap request uses the first domain as the competitor with keywords and the second as the comparison domain missing them. The provider searches its database; absence does not prove no real-world ranking. Traffic and ad-equivalent values are provider estimates, not measured visits or revenue. Overview includes ranking counts for positions 1–3, 1–10 and 1–20 when provided. Repeated identical captures can be compared, row by row, in the UI. **Backlink Explorer** has profile totals, individual backlinks, referring domains, anchor text and monthly link history (up to 366 days per request). Select exact domain/subdomains and live/all/lost status, plus dofollow/nofollow filters. Individual backlinks also support new and broken-link filters. The provider's first/last seen, broken/lost flags and index counts may lag the web. Rank is requested on the provider's 0–100 scale. Spam scores are provider heuristics, not Google judgments or automatic recommendations to disavow links. All three retain dated local history. Reopen exact saved results without another provider request, sort/filter locally, or export the filtered table to CSV and the full normalized capture/provenance to JSON. CSV cells protect spreadsheet formula prefixes. Rows can be added to a shortlist with revision-checked notes. A shortlist item can create a deduplicated editorial-review task for an audited real website. Neither saving, exporting nor task creation publishes content or sends outreach. ## Setup and billing 1. Open an Explorer → **Usage & connection**. Save your own API login/password. This reuses the same encrypted DataForSEO connection as Growth monitoring. Labs and Backlinks may require separate product activation, balance or fees with the provider; successful access to one does not establish access to all. 2. Review the shared Explorer monthly call cap and planning budget (default 100 attempts / USD 10). This is separate from each Growth monitoring budget. 3. Select a research mode, keyword/domain and scope. Common geographic codes are shown, and custom provider-supported location/language codes can be entered. 4. Approve that exact input and its per-request allowance. **Run research — paid** queues one provider task. The local worker must be running. Every refresh and next provider page needs a new approval; opening saved data is free of new API requests. Pagination returns 10, 50 or 100 rows, with offsets up to 10,000. 5. Compare reported costs with the provider's bill. The request allowance is a local planning reservation, not a price quote or invoice ceiling. Unknown charges after timeouts or malformed responses remain possible. No retry is automatic. Set billing controls with DataForSEO too. ## Persistence and failure behavior Explorer state is an optional field in the existing SQLite application state; existing installations need no migration. No customer data is replaced. Bounds: 60 research captures, 100 returned rows per paged capture, 500 shortlist entries, 500 recent request records and 10 active queued requests. Export/remove captures when full. Shortlist evidence survives deletion of its source capture. The API validates inputs and only uses a fixed allowlist of provider endpoints. Responses are bounded to 2.5 MB and requests time out after 60 seconds. No endpoint, custom filter expression or arbitrary request URL can be supplied by the browser. Only normalized bounded fields are retained; provider error bodies are not shown. The worker atomically claims a request and reserves the allowance in the UTC-month ledger before external IO. An idempotency key and active-query deduplication prevent duplicate enqueue operations. Provider credentials are fingerprinted. Cancellation, pause, provider changes or expired ownership prevent publication. Expired requests fail permanently; no ambiguous paid call is retried. Known costs are retained even if the output is discarded. An overage raises the recorded allowance and pauses new Explorer work until the operator reviews settings. **Pause & cancel active requests** prevents results from pending/running work. A request already sent to the provider may still incur a charge. Merely changing usage caps preserves prior spending and cannot undo a sent request. ## Provider contracts and validation DataForSEO reference documentation: - [Keyword overview](https://docs.dataforseo.com/v3/dataforseo_labs/google/keyword_overview/live/), [suggestions](https://docs.dataforseo.com/v3/dataforseo_labs/google/keyword_suggestions/live/), [related keywords](https://docs.dataforseo.com/v3/dataforseo_labs/google/related_keywords/live/), [ideas](https://docs.dataforseo.com/v3/dataforseo_labs/google/keyword_ideas/live/). - [Domain overview](https://docs.dataforseo.com/v3/dataforseo_labs/google/domain_rank_overview/live/), [ranked keywords](https://docs.dataforseo.com/v3/dataforseo_labs/google/ranked_keywords/live/), [relevant pages](https://docs.dataforseo.com/v3/dataforseo_labs/google/relevant_pages/live/), [competitors](https://docs.dataforseo.com/v3/dataforseo_labs/google/competitors_domain/live/), [domain intersection](https://docs.dataforseo.com/v3/dataforseo_labs/google/domain_intersection/live/). - [Backlink summary](https://docs.dataforseo.com/v3/backlinks/summary/live/), [backlinks](https://docs.dataforseo.com/v3/backlinks/backlinks/live/), [referring domains](https://docs.dataforseo.com/v3/backlinks/referring_domains/live/), [anchors](https://docs.dataforseo.com/v3/backlinks/anchors/live/), [history](https://docs.dataforseo.com/v3/backlinks/timeseries_summary/live/). `tests/explorers.test.ts` covers contracts, missing/zero metrics, scope validation, request approval, idempotency, costs/limits, cancellation, credential changes, expired leases, shortlist and task behavior, CSV safety and request authentication. `npm run test:explorers-ui` uses isolated temporary storage and a synthetic provider with the actual API routes and browser. It exercises all modes, history/cache, filters, notes, export, task handoff and responsive layouts. Development checks make no paid calls and do not establish live provider entitlement or availability. Local verification for this change: 180 tests and application/connector typechecks pass, as does the production build. Official documented response examples were parsed for eleven endpoints. The isolated Explorer browser test covers all 14 research modes; existing growth/content/intelligence and monitoring browser suites also pass. Tests use no customer credentials or paid DataForSEO calls. The running local app responds and its worker heartbeat is current.