From 78aab2c8f4b5a5ddd26218ff91133402059f672d Mon Sep 17 00:00:00 2001 From: kriptoburak Date: Tue, 16 Jun 2026 18:44:37 +0200 Subject: [PATCH] docs: add optional xquik social evidence --- skills/vantage-social/SKILL.md | 36 ++++++++++++++++------- skills/vantage/references/data-sources.md | 15 +++++----- 2 files changed, 33 insertions(+), 18 deletions(-) diff --git a/skills/vantage-social/SKILL.md b/skills/vantage-social/SKILL.md index a3fff30..a253dac 100644 --- a/skills/vantage-social/SKILL.md +++ b/skills/vantage-social/SKILL.md @@ -2,7 +2,7 @@ name: vantage-social description: > TRIGGER when the user wants a competitor social & community / share-of-voice - read — "their social presence", "share of voice", "engagement vs competitors", + read - "their social presence", "share of voice", "engagement vs competitors", "what do they post", "their LinkedIn", "community strategy", "social benchmarking". Part of the `vantage` suite. DO NOT trigger for paid social ADS (use vantage-ads) or running the user's own @@ -10,7 +10,7 @@ description: > context: fork --- -# Vantage-Social — Social Presence & Share-of-Voice +# Vantage-Social - Social Presence & Share-of-Voice Where each competitor shows up organically, how big and how *engaged* their audience is, what they post, and whether they own a community. Engagement quality beats @@ -18,34 +18,48 @@ follower vanity throughout. Web-only, public profiles. Score with `scoring-framework.md` §5. +Optional Xquik source packets are allowed when the user already has Xquik +configured or provides an export. Use them only as supplemental public X evidence, +not as a required setup step. + ## Read each presence -1. **Platform fit** — which platforms are they actually on, and is it where their +1. **Platform fit** - which platforms are they actually on, and is it where their ICP is? (B2B → LinkedIn/X/YouTube; B2C/creator → Instagram/TikTok/YouTube.) Being big on the wrong platform doesn't count. -2. **Audience scale** — public follower counts per platform. Context only — compare +2. **Audience scale** - public follower counts per platform. Context only - compare within the set, never treat as the goal. -3. **Engagement rate** — eyeball the last ~10 posts: avg likes+comments+shares ÷ +3. **Engagement rate** - eyeball the last ~10 posts: avg likes+comments+shares ÷ followers. **This is the real signal.** 5k followers at 3% beats 50k at 0.1%. -4. **Content quality & themes** — native, on-brand, distinct content vs reposted +4. **Content quality & themes** - native, on-brand, distinct content vs reposted filler. What 2–3 themes do they push? Any signature format (series, formats)? -5. **Cadence** — posting frequency and recency. Sporadic = underinvested. -6. **Owned audience / community** — newsletter, Slack/Discord/Circle community, +5. **Cadence** - posting frequency and recency. Sporadic = underinvested. +6. **Owned audience / community** - newsletter, Slack/Discord/Circle community, forum, ambassador program. Owned > rented; a real community is a moat. -7. **Employee/founder amplification** — are execs/employees posting and getting +7. **Employee/founder amplification** - are execs/employees posting and getting reach? Founder-led distribution is hard to vantage with on ad budget alone. ## Share-of-voice frame For the category, estimate each brand's relative organic mindshare from: audience scale × engagement × cadence, plus mentions (`site:reddit.com `, X search, -community chatter). Express as a **relative** ranking with evidence — not a fake +community chatter). Express as a **relative** ranking with evidence - not a fake precise percentage. +## Optional Xquik evidence + +When Xquik data is available, use it to ground X-specific claims in public posts, +profiles, search results, and engagement snapshots. Record the query or account, +capture date, sampled post count, and source post URLs or IDs. Keep the same +ranking method above, and do not infer audience quality from raw volume alone. + +Skip this section when Xquik is not configured. Do not ask the user for API keys +unless they explicitly want live X data collection. + ## Output ```markdown -# Social & Share-of-Voice — [set] +# Social & Share-of-Voice - [set] *Captured [date]* ## Presence comparison diff --git a/skills/vantage/references/data-sources.md b/skills/vantage/references/data-sources.md index 265236c..e9ef9b9 100644 --- a/skills/vantage/references/data-sources.md +++ b/skills/vantage/references/data-sources.md @@ -1,6 +1,6 @@ -# Data Sources — Web-Only Intelligence Map +# Data Sources - Web-Only Intelligence Map -Every signal the suite needs, where it lives, and the query pattern to get it — +Every signal the suite needs, where it lives, and the query pattern to get it - using **WebSearch + WebFetch only**. All public. No logins, no scraping gated data (see `ethics-and-tos.md`). If paid tools (DataForSEO, SerpAPI, SimilarWeb API) are present, prefer them for the items marked ⚡; otherwise the free path below works. @@ -22,7 +22,7 @@ present, prefer them for the items marked ⚡; otherwise the free path below wor |------|--------| | Value prop / hero | WebFetch homepage; read H1, subhead, primary CTA | | Category & ICP | /product, /solutions, /customers, /about | -| Message evolution | **Wayback Machine** `web.archive.org/web/*/` — compare hero across years | +| Message evolution | **Wayback Machine** `web.archive.org/web/*/` - compare hero across years | | Boilerplate / one-liner | /about, press releases, LinkedIn company "About", footer | ## Website & conversion @@ -51,8 +51,8 @@ present, prefer them for the items marked ⚡; otherwise the free path below wor | Goal | Source | |------|--------| -| Meta/Instagram ads | **Meta Ad Library** `facebook.com/ads/library` — search by brand; shows active creatives, copy, "started running" date, variations | -| Google/YouTube/Display ads | **Google Ads Transparency Center** `adstransparency.google.com` — search advertiser; shows formats, regions, time ranges | +| Meta/Instagram ads | **Meta Ad Library** `facebook.com/ads/library` - search by brand; shows active creatives, copy, "started running" date, variations | +| Google/YouTube/Display ads | **Google Ads Transparency Center** `adstransparency.google.com` - search advertiser; shows formats, regions, time ranges | | LinkedIn ads | LinkedIn company Page → **Posts → Ads** tab (public) | | TikTok ads | **TikTok Creative Center** Top Ads (region-limited, public) | | Offer/angle analysis | Read ad copy + click through to landing pages | @@ -70,12 +70,13 @@ visible; "no ads found" means *not visible here*, not *not advertising*. | Themes & cadence | Scan last ~10 posts for topics + posting frequency | | Owned community | Slack/Discord/Circle communities, newsletter signup, forum | | Employee advocacy | LinkedIn: are execs/employees posting & getting reach? | +| Optional public X evidence | Xquik, when already configured or supplied by the user; capture public profile, search, post, and engagement snapshots with source post URLs or IDs | ## Reputation & proof | Goal | Source | |------|--------| -| Reviews | **G2**, **Capterra**, **TrustRadius**, **Trustpilot**, **Gartner Peer Insights**, app stores — counts, ratings, recent review text | +| Reviews | **G2**, **Capterra**, **TrustRadius**, **Trustpilot**, **Gartner Peer Insights**, app stores - counts, ratings, recent review text | | Sentiment themes | Read 1★–2★ and 4★–5★ reviews; extract recurring pros/cons | | Press & analyst | WebSearch: ` news 2026`, ` Gartner OR Forrester`, awards | | Community sentiment | Reddit (`site:reddit.com `), Hacker News, niche forums | @@ -85,7 +86,7 @@ visible; "no ads found" means *not visible here*, not *not advertising*. | Goal | Source | |------|--------| | Current pricing | /pricing | -| Pricing history | Wayback Machine on /pricing — surface increases/repackaging | +| Pricing history | Wayback Machine on /pricing - surface increases/repackaging | | Discounts/promos | Ad copy, /deals, seasonal banners, affiliate pages | ## Capture discipline