/performance-check

name: performance-check
description: "Pull live metrics from every connected analytics MCP into one cross-channel snapshot: KPI scoreboard with RAG status vs profile targets, period-over-period trends, industry benchmarks, top wins and concerns, and 3-5 recommended actions — then persist the snapshot via performance-monitor.py for trend history. Triggers on \"/digital-marketing-pro:performance-check\", \"how are our marketing metrics\", \"pull current KPIs\", \"quick performance snapshot\", \"are we hitting our targets\". Reads the brand profile for KPI targets and industry benchmarks; reports data gaps for unconnected platforms. Pairs with /digital-marketing-pro:performance-report, which turns these snapshots into the stakeholder narrative."
user-invocable: true
triggers:
  - check marketing performance
  - pull current KPIs
  - performance snapshot
  - how are our marketing metrics
  - compare performance to targets
  - marketing performance check
  - quick KPI health check
  - check campaign performance

/digital-marketing-pro:performance-check

Purpose

Pull live metrics from all connected analytics MCPs and produce a comprehensive performance snapshot. Compares current performance to KPI targets defined in the brand profile, previous-period benchmarks, and industry averages. Designed for quick health checks — run it daily, weekly, or on-demand to stay on top of marketing performance without switching between platforms.

**Scope (vs `/digital-marketing-pro:performance-report`):** this skill is the **live-pull + snapshot-persistence** layer — it fetches current metrics from the platforms and saves a snapshot for trend history. When you need a formatted, narrative deliverable for stakeholders (executive summary, channel commentary, prioritized recommendations, branded formatting), run `/digital-marketing-pro:performance-report`, which consumes the snapshots this skill persists rather than re-pulling. Use `performance-check` to *see the numbers now*; use `performance-report` to *tell the story*.

Input Required

The user must provide (or will be prompted for):

• **Time period**: Today, this week, this month, this quarter, or a custom date range (e.g., "last 14 days", "Jan 1 - Jan 31")
• **Channel focus** (optional): Specific channels or platforms to prioritize (e.g., "paid search only", "email and social").

If omitted, all connected platforms are included

• **Comparison period** (optional): Period to compare against — previous period, same period last year, or custom range.

Defaults to the equivalent previous period

• **KPI targets** (optional): Override targets for this check.

If omitted, targets are pulled from profile.json goals and KPI settings

• **Granularity** (optional): Daily, weekly, or aggregate view. Defaults to aggregate for the selected period

Process

1. **Load brand context**: Read `~/.claude-marketing/brands/_active-brand.json` for the active slug, then load `~/.claude-marketing/brands/{slug}/profile.json`. Apply brand voice, compliance rules for target markets (`skills/context-engine/compliance-rules.md`), and industry context. Also check for guidelines at `~/.claude-marketing/brands/{slug}/guidelines/_manifest.json` — if present, load restrictions. Check for agency SOPs at `~/.claude-marketing/sops/`. If no brand exists, ask: "Set up a brand first (/digital-marketing-pro:brand-setup)?" — or proceed with defaults. 2. **Detect connected analytics MCPs**: Check `.mcp.json` and active MCP connections to identify which platforms are available (google-analytics, google-ads, meta-marketing, linkedin-marketing, tiktok-ads, mailchimp, stripe, mixpanel, amplitude, shopify, etc.). Log any expected platforms that are not connected so the user knows about gaps in coverage. 3. **Pull metrics from each connected platform**: Request key metrics for the specified time period:

• Traffic: sessions, users, pageviews, new vs returning (break out GA4's **"AI Assistant"** default channel — referrals from ChatGPT, Gemini, Copilot, Perplexity, etc. — so AI-sourced traffic isn't buried under Referral/Direct)
• Ads: impressions, clicks, spend, CPC, CPM
• Conversions: leads, purchases, sign-ups, goal completions
• Revenue: total revenue, average order value, transaction count
• Engagement: open rate, click rate, bounce rate, time on site
• Platform-specific: email deliverability, social reach, video views, app installs

4. **Aggregate into unified dashboard**: Normalize metrics across platforms into a single cross-channel view with consistent naming, currency conversion if multi-currency, and de-duplicated conversion counts where platforms overlap 5. **Calculate KPIs vs targets**: Compare actuals to targets from `profile.json` goals — flag green (on track or exceeding), yellow (within 10% of target), or red (missing by >10%). Include absolute and percentage variance for each KPI. 6. **Compare to previous period**: Calculate period-over-period change for every metric and attach trend direction (up/down/flat) with percentage change. If year-over-year data is available, include as a secondary reference point. 7. **Benchmark against industry**: Reference `skills/context-engine/industry-profiles.md` for the brand's industry to contextualize performance relative to category averages. Flag metrics significantly above or below industry norms. 8. **Identify notable findings**: Surface the top 3 wins (best-performing metrics or biggest improvements), top 3 concerns (underperforming or declining metrics), and any material changes that warrant deeper investigation. Before labelling a conversion-rate change "statistically significant," confirm it with `python "${CLAUDE_PLUGIN_ROOT}/scripts/significance-tester.py" --control-visitors {n} --control-conversions {n} --variant-visitors {n} --variant-conversions {n} --confidence 0.95` — do not call a movement significant off a raw percentage delta. 9. **Generate recommended actions**: Based on the d