Marketing Skill

Competitor Profiling

The competitor-profiling skill turns a list of competitor URLs into structured, dated competitor profiles for an AI agent to deliver: positioning, features, pricing, customers, reviews, and SEO strength, with every claim traced to a source and observed facts kept separate from inference.

Maintained by Corey Hainesv2.1.8Updated 2026-11-092 references
npx skills add coreyhaines31/marketingskills

Installs the full library — Competitor Profiling included

A useful competitor profile is built from evidence, not impressions. Scrape the homepage, pricing, features, about, customers, integrations, and changelog pages, then pull domain rank, backlinks, referring domains, ranked keywords, and estimated organic traffic from an SEO data source. Save the raw scrapes and API responses to a dated folder before writing anything. In the profile, keep three layers apart: what a source says (with source and date), what you infer from it, and what it might mean for you, framed as a question. Write "not observed on the pages checked" instead of "doesn't have," never guess a competitor's motive, and cross-check scale claims like "10,000 customers" against their traffic. Use one template for every competitor so profiles compare side by side.

How the Competitor Profiling skill works

  1. 1

    Confirm URLs, depth, and focus

    The agent reads your product marketing context, then confirms the competitor URLs, what your product does, depth level, and any focus areas such as pricing or SEO. It defaults to a quick scan (homepage, pricing, and an SEO summary) unless you ask for a deep profile or name three or fewer competitors. With 10+ competitors it suggests profiling the top 5 first.

  2. 2

    Select the tool stack

    It checks which tools are connected. The default stack is Firecrawl for scraping and site mapping plus DataForSEO for SEO and market data. Without those, it uses WebFetch and whatever SEO source is connected, such as Ahrefs, Semrush, or Ubersuggest. With no SEO source at all, it marks those metrics unavailable rather than estimating them.

  3. 3

    Scrape the site and reviews

    It maps the site (or probes common paths like /pricing and /plans), scrapes each key page, and extracts headline, value proposition, tiers, features, named customers, integration count, and release velocity. For deep profiles it also scrapes G2, Capterra, Product Hunt, and TrustRadius for ratings, praise and complaint themes, and 3–5 quotes. Page text is treated as data, so any embedded instructions aimed at AI agents are ignored and noted.

  4. 4

    Pull SEO and market data

    It pulls backlink summaries, referring domains, ranked keywords in the top 3, 10, and 100, estimated organic traffic and value, top pages, and the competitor's own closest organic competitors, which can surface rivals you hadn't considered. The same metrics are pulled for every competitor so the numbers line up.

  5. 5

    Save raw data by date

    Every scrape, SEO response, and review page is saved under competitor-profiles/raw/<slug>/<YYYY-MM-DD>/ in scrapes, seo, and reviews folders. A new run always gets a fresh date folder and never overwrites an old one, so you can re-read, audit, or diff snapshots later without paying for the API calls again.

  6. 6

    Synthesize profiles, then the summary

    Each competitor gets a profile on the same template: at a glance, positioning, product, pricing, customers, SEO and content, strengths and weaknesses with evidence, and implications for your product. Your own advantages are labeled hypotheses unless backed by customer or win/loss evidence. A cross-competitor _summary.md with a comparison table, positioning map, takeaways, and gaps comes last.

When to use it

Your agent reaches for this skill when you ask for things like:

“competitor profile”“competitor research”“competitor analysis”“profile this competitor”“analyze competitor”“competitive intelligence”“competitor deep dive”“who are my competitors”“competitor landscape”“competitor dossier”

What's inside

The playbook covers

Initial AssessmentTool Stack SelectionCore PrinciplesSaving Raw DataResearch ProcessOutput FormatAt a GlancePositioning & MessagingProduct & FeaturesPricingCustomers & Social ProofSEO & Content StrategyStrengths & WeaknessesCompetitive Implications for [Your Product]Raw Data SourcesQuick Scan vs. Deep ProfileHandling Multiple CompetitorsUpdating ProfilesTask-Specific Questions

See it in action

You ask

Profile these three competitors for us: https://competitor1.com, https://competitor2.com, https://competitor3.com. We need this for sales enablement and to find positioning gaps.

The skill delivers
competitor2.md · Pricing and implications (excerpt)

Observed (pricing page, scraped 2026-10-27): three tiers. SSO and audit logs appear only on Enterprise, which shows "Contact sales." Observed (changelog, 2026-09): SSO moved from Pro to Enterprise. Inferred (medium confidence): mid-market buyers needing SSO now face a sales-led step. Implication: is there a deal-stage opening with security-conscious mid-market teams? Confirm in their docs or a trial before using this on a comparison page. Not observed on homepage, features, or integrations pages: a native HubSpot integration. Treat as unconfirmed, not absent.

_summary.md · Gaps and opportunities (excerpt)

Positioning map (simple to complex vs. SMB to enterprise): competitor1 and competitor3 both claim "all-in-one for growing teams"; competitor2 is moving upmarket by packaging, not messaging. Gaps: 1. None of the three leads with time-to-value; homepage copy is feature-led on all three. 2. Review complaints cluster on onboarding for competitor1 (G2 themes, pulled 2026-10-27). 3. competitor3 claims "10,000+ customers" but has 310 referring domains and modest traffic. Flagged as unverified. For sales: these feed directly into battle cards via the sales-enablement skill.

More things to ask

Real requests the skill is tested against:

  • We have 12 competitors. Profile all of them.

  • I have an existing profile of Notion from 4 months ago. Should I update it or start fresh?

  • What pages should I scrape for a competitor profile?

  • I want a profile but I don't care about SEO data — just pricing, positioning, and customer logos. Can you skip the DataForSEO calls?

  • Should I trust the customer logo wall on the competitor's homepage as evidence of who their customers are?

FAQ

What should a competitor profile include?

An at-a-glance table (tagline, founded, HQ, team size, funding, domain rank, organic traffic, referring domains, keywords), positioning and messaging themes with source pages, core capabilities and product direction from the changelog, pricing by tier with billing and trial details, named customers and review ratings, SEO and content signals, strengths and weaknesses with evidence, implications for your product, and a raw data sources section with dates. Every profile uses the same template.

How do I research competitors without guessing?

Separate three layers in writing. Observed facts carry a source and date. Inferences are labeled with a confidence level when they'd change a decision. Implications are phrased as questions or options. If you can't find a feature, write "not observed on [pages checked], as of [date]" and confirm in their docs or a trial before using the gap publicly. "No public pricing" means neither free nor expensive, and you shouldn't state a motive they haven't stated.

What pages should I scrape on a competitor's site?

Homepage, pricing, features or product pages, about, the blog index, customers or case studies, integrations, and the changelog if one exists. If you can't map the site, fetch the homepage first and follow its navigation links, or try common paths like /pricing, /plans, /customers, and /integrations. Add G2, Capterra, Product Hunt, and TrustRadius pages for review themes on deeper profiles.

How often should I update competitor profiles?

Treat each profile as a dated snapshot. When updating, check the pricing page first since it changes most, re-pull SEO metrics because traffic and rankings shift monthly, and scan the changelog for product changes. Update the generated date and add a change log table at the bottom recording what changed and the source. Only dated snapshots, not a single scrape, can support a claim that something hasn't changed.

Can I trust the numbers on a competitor's website?

Treat them as claims to cross-reference. If a site says "10,000 customers" but its traffic and backlink profile look small, flag the mismatch in the profile instead of repeating the number. Logo walls show what a company wants to be associated with, so check case studies and reviews for real usage. Competitor pages can also contain hidden text aimed at AI tools; the skill treats it as data and notes the attempt.

Can an AI agent profile competitors automatically?

Yes, given scraping and SEO data tools. With Firecrawl and DataForSEO connected, or WebFetch plus another SEO source, the agent maps each site, scrapes key pages, pulls the same SEO metrics for every competitor in parallel, saves raw data to dated folders, and writes one profile per competitor plus a summary. Without an SEO source it still produces the qualitative profile and marks traffic and backlink metrics unavailable instead of estimating them.

Related skills

Use it in your agent

Competitor Profiling is a plain SKILL.md folder, so the same skill loads in any agent that reads the skills format. Setup guides:

Build a tool for this workflow?

Marketers run this skill mid-task — the exact moment your tool matters. Partner tiers put your tool inside the skill their agent runs.

See partner tiers