competitive-intel

name	competitive-intel
version	1.0.0
description	Competitive intelligence on one or more companies — paid ads, organic post activity, tech stack, tracking pixels, social footprint, and hiring signals. Compares multiple competitors side-by-side when asked. Uses ad_search, website_intelligence, web_tech_stack, post_keyword_search, and profile_activities on key executives. Proactively invoke on "what ads is X running", "compare X vs Y", "what tech does X use", "competitive scan".
benefits-from	["richapi-gtm","account-research"]
allowed-tools	["Bash","Read","Write","Edit","AskUserQuestion"]
triggers	["what ads is X running","compare X vs Y","what tech does X use","competitive scan","competitive intel","look at their landing page"]

competitive-intel

Quick, comparable, evidence-based snapshots of competitors.

Preamble

~/.claude/skills/richapi-gtm-skills/bin/richapi-skills-preflight

Phase 0 — define the question

Ask if unclear — the cost varies 10× by angle:

The user really wants to know…	Use
"What ads are they running?"	ad_search + ad_details
"What's their tech stack?"	web_tech_stack (cheap) or website_intelligence (comprehensive)
"What are they posting about?"	post_keyword_search with fromCompany
"Who are their thought leaders?"	enrich_profile + profile_activities on 2-3 named execs
"How do we stack up vs them?"	Full compare mode — run a narrow version of all the above on each side

Phase 1 — ad intelligence

ad_search (0.2/result) scoped to the target company. You can:

• Pass advertiser: "<Company>" or LinkedIn Ad Library URL
• Add dateRange to scope to last 30/90 days
• Add country if geo-specific

After getting the list, run ad_details (2 credits each) on the top 2-3 most interesting ads only — don't burn credits on every ad.

What to extract

• Creative angles: what pains / promises / CTAs do the ads emphasize?
• Audience signals: targeting breakdown (if available) — industries, titles, seniority
• Cadence: how often they launch new creatives (churn rate signal)
• Channels: image vs video vs carousel share

Phase 2 — tech stack + GTM signals

For a fast read: web_tech_stack (2 credits) on their marketing site.

For the full picture: website_intelligence (5 credits) — one call, all of:

• Tech stack (100+ detectors)
• Pixels / analytics (Meta, LinkedIn, TikTok, etc. — reveals ad spend channels)
• Social profile links
• Emails on the site
• SSL / domain metadata

Rule: If the user wants >2 companies, website_intelligence is almost always the right call. Running tech_stack + pixels + social_links separately costs 2+1+2 = 5 anyway, and loses the aggregated view.

Interpreting pixels → GTM playbook

Pixel seen	Implies
LinkedIn Insight Tag	Running LinkedIn Ads — high-intent B2B
Meta Pixel + Google Ads + TikTok	Multi-channel paid acquisition (DTC or SaaS-wide)
HubSpot / Marketo / Pardot	Running marketing automation → organized demand gen
Segment	Data-mature — tracks events to multiple downstream tools
Intercom / Drift / Zendesk	Support ops maturity
Cal.com / Calendly / Chili Piper	Sales booking flow (meeting-driven)

Phase 3 — content activity

Last 30d posts from the company page: post_keyword_search with { fromCompany: "<companyUrn>", datePosted: "past-month" } (6 credits).

Ask before running — 6 credits is the biggest single line item here.

Extract themes:

• Product-led (feature launches, tutorials)
• Thought-leadership (opinions, research)
• Employer-brand (hiring, culture)
• Customer-led (case studies, testimonials)

Phase 4 — exec voice (optional)

For 1-3 named executives: enrich_profile + profile_activities (POST only).

Cost: 1 + 2 = 3 credits each. Cap at 3 execs = 9 credits.

What this tells you:

• Where they're consciously positioning themselves
• What they openly dislike about competitors (pay attention — they often name names)
• What customers they're engaging with publicly

Phase 5 — compare-mode synthesis

When comparing 2+ companies, output a side-by-side grid:

                      Acme Corp            Beta Industries
Size (LinkedIn)       501-1000             201-500
Ad cadence (30d)      12 new creatives     3 new creatives
Ad angle              "Save time"           "Save money"
Pixels                LI + Meta + HubSpot   LI only
Tech maturity         Enterprise            Scrappy
Company posts (30d)   21 (product-led)      8 (thought-leadership)
Top exec voice        CEO — thought lead    CMO — product demos
Inferred strategy     Demand gen at scale   Founder-led, narrower ICP

Then a 2-3 sentence what-this-means interpretation for the user.

Cost ceiling

Sensible defaults per company:

• Light scan (ads + tech only): ~10 credits
• Standard compare: website_intelligence (5) + 5 ad_details (10) + company posts (6) = ~21 credits per company
• Deep dive: add exec voice (9) + fuller ad sweep = ~35 credits per company

State the bill before running. For compare mode with 3+ companies, always confirm.

Anti-patterns

• Don't run ad_details on every ad — top 3-5 only.
• Don't interpret tech stack without seeing the site — pixels can be stale.
• Don't compare companies on metrics that vary by company-size (total posts, total employees) — normalize by headcount.
• Don't make claims about strategy you can't point at (ad, post, hire) as evidence. Always cite.

Follow-ups

• "Want to find decision makers at [winner of the comparison]? → account-research"
• "Want to build a prospect list that looks like their customers? → build-prospect-list"

Offer to schedule this

"Want this as a weekly intel report? Every Monday I can re-pull ads + posts + tech signals for these competitors and email you the diff vs last week. → scheduled-workflow."

Weekly competitor diffs catch campaign launches and positioning shifts within a week — any slower and the intel is a history lesson.