# Competitor Research

> SEO competitor research — find the real SERP competitors, compare authority, traffic and backlinks, mine keyword and content gaps, spot weak pages you can outrank, and turn them into a page plan. Use when the user asks who their competitors are, competitor analysis, keyword gap, content gap, backlink gap, or "what are competitors ranking for". For general keyword ideas use keyword-research.

- Skill: `khadinakbarlabs/competitor-research` (Agent Skill)
- Install (CLI): `npx get-skill add khadinakbarlabs/competitor-research`
- Raw SKILL.md: https://api.skillmd.com/api/skills/khadinakbarlabs/competitor-research/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: khadinakbarlabs (https://skillmd.com/u/khadinakbarlabs)
- Updated: 2026-11-16
- Page: https://skillmd.com/skills/khadinakbarlabs/competitor-research

---


# Competitor research

Tools: Apify Actors per `../../references/tools-and-apis.md` (run protocol). Log paid calls in `ledger.json`.

**Paid checks:** before any paid run, say it is a paid check billed to the user's own Apify account and wait for a clear yes. Do not state an amount; give it only if the user asks what that task costs (then call `quote_cost` and tell them). Always call `quote_cost` anyway to get the spending cap for each run, and set it without comment.

## Web connector mode
Cover the whole step with one call: say it is a paid check, wait for a yes, then `quote_cost` with `pack: "competitors"` for the caps and `call-actor` on `khadinakbar/*` only. Show five findings, not a data dump (`../../references/plain-language.md`). Keep the snapshot in chat; offer `save_results` under a project name if they want history.

## 1. Identify real competitors (by SERP, not by opinion)
- Take 5–15 money keywords (from `context.md`, the homepage/money pages, or the user's GSC export).
- Pull SERPs for each with `khadinakbar/scrape-google-serp`; count which domains recur in the top 10 → SERP competitors. Separate **business competitors** (sell the same thing) from **content competitors** (publishers, directories, marketplaces, Reddit, listicles) — the latter are link/mention targets, not rivals.
- Keep 3–5 business competitors (merge with the user's list).

## 2. Snapshot table
For you + each competitor: referring domains and backlink summary (`khadinakbar/website-backlink-checker`, `mode: summary`), how many of the tracked SERPs each ranks in and at what positions, top pages by traffic, page types that win (BoFu landing pages, comparison pages, tools/templates, blog), AI-visibility hints (are they named in AI answers? see `ai-search-visibility`). Label all third-party traffic numbers as estimates.

## 3. Keyword and content gaps
- Expand the topic set with `khadinakbar/dataforseo-keyword-research` (`mode: keyword_ideas`, 1–10 `seedKeywords` per run, ≤50 rows per run; each run carries a per-request fee, so batch seeds into one run); add `mode: search_volume` for specific phrases and `khadinakbar/google-trends-scraper` for seasonality.
- Gap = keywords where competitors appear in the SERP top 10 and you don't (from the SERP scrapes) or where you have no matching page.
- Classify intent from the SERP (page types ranking) and the phrasing — prioritise "best X for Y", "X vs Y", "X alternatives", "X pricing/cost", "service + city", templates/tools.
- **Weak-page mining** (high-yield): for each candidate keyword, inspect the ranking URLs. A ranking page is *weak* if it has 0–3 referring domains, the keyword missing from its title/H1/slug, a broad/accidental page type, or stale/thin content. Weak top-10 results = keywords you can win with one dedicated, intent-matched page.
- Cluster keywords sharing a SERP (>50–70% top-10 URL overlap = one page; otherwise separate pages).

## 4. Backlink gap
- Diff referring domains (`website-backlink-checker`, `mode: referring_domains`, ≤100 rows per competitor): domains linking to ≥2 competitors but not you. Classify: directory/listing (claim now), review site, listicle/roundup (pitch inclusion), publisher/editorial (digital PR), partner/integration, community. Output feeds `link-building`.

## 5. Cited for rivals, not for you
If AI results exist for this brand (from `ai-search-visibility` or `citation-loop`), read the sources in those answers and list the sites that AI cites for two or more competitors but never for the brand. For each: the site, the page, the competitors it is cited for, and the kind of page (list, review site, forum thread, news). Hand the list to `link-building` / `outreach` as first targets. If there are no AI results yet, say so and offer the AI check (paid; ask first). Do not guess the list.

## 6. Outputs
- `seo-workspace/<domain>/competitors.md`: competitor table, what each does well, the 3 strategic gaps to exploit.
- `seo-workspace/<domain>/keyword-opportunities.csv`: `keyword,volume,kd,intent,ranking_url,weak_page_reason,proposed_page`.
- A page plan: top 10 pages to create or upgrade, each with target keyword cluster, page type (per SERP), title/H1 draft using `Keyword | benefit | Brand`, and internal links to add.

## Rules
- Pages compete, not domains: always reason at page level.
- Don't propose mass templated pages (doorways / scaled content abuse). Each proposed page needs a real reason to exist.
- Competitor data is research only; never contact competitors or copy their content.
- Finish with the three-line footer (Done, Cost so far, Next).

