convez.ai
Skill
SameerKhan

saas-growth-dashboard

SameerKhan/saas-growth-dashboard

Ver no GitHub
BomPluginScore · de 0 a 100

Open-source BigQuery-first growth intelligence dashboard for SaaS founders. Connect GA4, ChartMogul, Amplitude, Ahrefs, and 15+ more sources. Includes Claude Code skill for automated CEO reports.

Instalação

$npx convez-ai SameerKhan/saas-growth-dashboard

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
2
Forks
0
Acompanhando
0
Saúde
57%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
1

README

A tradução para o português deste README ainda está na fila.

📊 SaaS Growth Dashboard

The open-source, BigQuery-first growth intelligence dashboard for SaaS founders

Connect your existing tools via MCP servers, pipe everything into BigQuery, and get a single-page executive dashboard with automated daily reports.

Built and battle-tested at Social Champ, now open-sourced for any SaaS to use.

Live Demo    License: MIT

Dashboard Overview

👉 View the fully interactive live demo — all 20 sources connected with synthetic data. Click through every tab.


📖 The Story Behind This

I spent 2 months building something I wish existed 5 years ago.

It started simple. I just wanted to see our Google Analytics and Google Ads data in one place.

Then I connected Amplitude. Then ChartMogul. Then Ahrefs. Then MongoDB, Crisp, ClickUp, GitLab, Meta Ads, Bing Ads, YouTube, Threads, Reddit, Clarity, Lighthouse, GrowthBook, Serpstat...

And before I knew it, I had 20+ data sources feeding into one dashboard.

Here's why I couldn't stop.

Every morning at Social Champ, we opened 8 tabs. GA4 for traffic. ChartMogul for MRR. Amplitude for activation. Ahrefs for SEO. ClickUp for sprint progress. Crisp for support tickets.

Each tool told us one piece. None of them told us the full story.

Which blog post brought the signup that converted to a paying customer? Which paid campaign is burning money on keywords nobody converts on? Which 70% of our blog gets literally zero traffic? How many customers are at churn risk right now?

These questions are impossible to answer when your data lives in 15 different silos.

So I built a single page that answers all of them.


✨ What It Does

🟠 Pulls from 20+ sources via MCP servers and APIs (GA4, ChartMogul, Amplitude, Ahrefs, Clarity, MongoDB, and more)

🟠 Stores everything in BigQuery in a layered warehouse (raw → modeled → executive). Costs $0-5/month for most SaaS

🟠 Generates an interactive HTML dashboard with color-coded KPIs, department health grades (A through F), and prioritized actions

🟠 Works with Claude Code, Codex (ChatGPT), Cursor, or any MCP-compatible tool. You can literally ask "run the daily CEO report" and get a 3-post executive brief

🟠 Connect what you have, skip what you don't. Start with GA4 and one revenue source. Add more as you grow

The best part? Our daily standups went from 20 minutes of "let me pull up the numbers" to opening one page.

💡 Every SaaS founder deserves to know what's actually happening in their business. Not from 8 tabs. Not from a spreadsheet someone fills by hand every Monday. From one place.


🚀 Quick Start (Under 1 Hour)

Prerequisites

  • ☁️ Google Cloud project with BigQuery enabled (free tier works)
  • 🐍 Python 3.11+
  • 🤖 Claude Code, Codex (ChatGPT), Cursor, or any MCP-compatible AI tool (for the AI skill, optional)
  • 📊 At least one data source (GA4 recommended as starting point)

Step 1: Clone and configure

git clone https://github.com/SameerKhan/saas-growth-dashboard.git
cd saas-growth-dashboard

# Copy config templates
cp config/config.example.yaml config/config.yaml
cp .env.example .env

# Install dependencies
pip install -r requirements.txt

Step 2: Edit your config

Open config/config.yaml and fill in:

company:
  name: "Your Company"
  domain: "yourcompany.com"

bigquery:
  project_id: "your-gcp-project"

sources:
  ga4:
    enabled: true
    property_id: "123456789"

Add credentials to .env:

BIGQUERY_PROJECT_ID=your-gcp-project
GA4_PROPERTY_ID=123456789

Step 3: Run your first refresh

python run_daily_growth_refresh.py --lookback-days 7

Step 4: Generate the dashboard

python run_growth_dashboard.py --open

Your dashboard opens at reports/growth_dashboard.html. 🎉

Step 5: Install the AI skill (optional)

Works with any MCP-compatible tool. For Claude Code:

cp -r skill/ ~/.claude/skills/saas-growth-intelligence/

For Codex (ChatGPT), Cursor, or other MCP tools, point your MCP config to the same skill directory and connect the MCP servers listed in config/config.yaml.

Now you can say:

  • "Run the daily CEO report"
  • "Investigate the biggest growth anomaly"
  • "Refresh the warehouse"

🏗️ Architecture

                    MCP Servers / APIs
                          │
                    Source Connectors
                    (growth_intelligence/sources/)
                          │
                    BigQuery Warehouse
                    (raw → modeled → executive)
                          │
              ┌───────────┴───────────┐
              │                       │
        HTML Dashboard          AI Skill
        (reports/)      (Claude Code / Codex / Cursor)
              │                       │
         Browser/CDN          Mattermost/Slack

💾 The BigQuery-first principle

Every metric is fetched once, structured once, and reused many times. The warehouse has three layers:

LayerPurposeExample
RawDirect API extractsraw_ga4_channel_daily
ModeledJoined, deduplicated, business-logic viewsdaily_growth_kpis
ExecutivePre-computed summaries for reportingceo_daily_brief

🔌 Supported Sources (20+)

Connect what you have, skip what you don't. Every source is optional except BigQuery.

🌐 Traffic & Acquisition

SourceTypeWhat You Get
Google Analytics 4MCP / APISessions, users, channels, landing pages, conversions
Google Search ConsoleAPIOrganic clicks, impressions, avg position, queries by page
Microsoft ClarityMCP / APIDead clicks, rage clicks, scroll depth, UX friction scores
Lighthouse / PSIAPICore Web Vitals (LCP, CLS, TBT), performance scores, accessibility

💰 Revenue & Billing

SourceTypeWhat You Get
ChartMogulMCP / APIMRR, ARR, churn rate, LTV, ARPA, customer movements (new, expansion, contraction, churn)
StripeMCPSubscription validation, billing reconciliation

🧪 Product & Activation

SourceTypeWhat You Get
AmplitudeMCP / SnapshotSignup-to-activation funnel, user journeys, feature adoption
MongoDBMCPNPS feedback, trial conversions, feature depth, job delivery health, payment signals
GrowthBookMCP / SnapshotA/B test results, experiment lift, confidence levels

📣 Paid Media

SourceTypeWhat You Get
Google AdsAPICampaign spend, conversions, CPA, search term intent (brand/competitor/generic)
Meta AdsAPIFacebook + Instagram ad performance, reach, CPM, ROAS
Bing AdsAPIMicrosoft Advertising campaigns, CPC comparison

🔍 SEO & Content

SourceTypeWhat You Get
AhrefsMCPDomain rating, backlink profile, organic competitors, keyword rankings
SerpstatAPISERP top-100 snapshots, keyword volume, difficulty, CPC

📱 Social Media & Community

SourceTypeWhat You Get
Meta (Social)APIFacebook Page reach, Instagram engagement, post performance
YouTubeAPIViews, watch hours, subscribers, video scorecard, traffic sources
ThreadsAPIViews, likes, replies, reposts, follower growth
RedditAPI (PRAW)Brand mentions, competitor tracking, opportunity threads

⚙️ Engineering & Support

SourceTypeWhat You Get
ClickUpMCPSprint health, task completion, bugs, tech debt, roadmap
GitLabAPIMerge request pipeline, code velocity, team workload
CrispAPISupport tickets, SLA tracking, resolution time, CSAT

💡 You don't need all of these. Start with GA4 + one revenue source (ChartMogul or Stripe) and add more as you grow. The dashboard shows "Not Connected" placeholders for missing sources.


📸 Dashboard Tabs

🏠 Overview

Executive narrative with AI-generated daily operating report, department health scores (A–F), and prioritized action items.

Overview Hero

Team briefings give each department a one-glance status card with the metrics that matter most to them.

Team Briefings


🔍 Acquisition & SEO

Section brief with scored health grade, organic traffic by channel, top landing pages, and keyword rankings from Ahrefs + GSC.

SEO Overview

Organic traffic → signup funnel split by stage (BoFU / MoFU / ToFU / Product / Other) with daily conversion tracking.

SEO Funnel Conversions

Blog posts by age × traffic matrix. Instantly see which content is driving traffic and which 70% is sitting at zero.

Blog Age x Traffic

Blog pages dropping out of the top 10, impression cliff alerts with MoFU/BoFU stage tags, and pages that lost >80% search visibility.

SEO Blog Decay


💰 Revenue & Activation

MRR and customer snapshot with health grade, revenue brief (What Changed / Why It Matters / What To Do Next), and ranked action items.

Revenue MRR

Daily MRR movement strip showing new, expansion, contraction, and churn by day with running totals.

Revenue Daily Strip

Churn risk scorecard ranking customers by health score, support intelligence with SLA tracking, and MRR at risk breakdown.

Churn & Retention


📱 Social Media

Cross-platform daily metrics, posting cadence tracker, YouTube video scorecard, and Reddit community intelligence.

Social Media

Daily posting cadence heatmap across Instagram, Facebook, YouTube, and Threads with posted/missed tracking.

Posting Cadence


📣 Paid Media Command Center

Cross-platform spend allocation bar (Google Ads + Meta Ads + Bing Ads), weekly performance at a glance, and search term intent split.

Cross-Platform Paid

Section brief with performance grade, ranked campaign insights, and winners vs waste detection.

Paid Media Overview

Search term intent analysis splitting spend into brand vs competitor vs generic, with campaign drill-down and waste term flagging.

Paid Search Intent


🏥 Site Health & UX

Core Web Vitals from Lighthouse, Microsoft Clarity UX friction data (rage clicks, dead clicks, scroll depth), and performance scores by page.

Site Health


⚙️ Engineering & Development

Sprint ship forecast with confidence tags (WILL SHIP / ON TRACK / AT RISK / WON'T SHIP), sprint command center, and velocity trend.

Engineering Sprint

Bug radar by severity with report bugs and internal bugs, plus GitLab MR pipeline with contributor leaderboard.

Engineering Bugs + GitLab


🧬 Product Health

Per-account health scorecard scoring top 200 customers on posting activity, team size, feature usage, and error rate.

Product Health Scorecard

Feature adoption bars across 10 features, trial-to-paid conversion funnel, and multi-feature depth analysis.

Feature Adoption


🤖 AI Skill Commands

Works with Claude Code, Codex (ChatGPT), Cursor, or any AI tool that supports MCP (Model Context Protocol). After installing the skill, just talk to your AI:

CommandWhat It Does
daily-ceo-report📋 Generate 3-post executive report with KPIs, breakdowns, and actions
refresh🔄 Refresh all BigQuery tables from connected sources
anomaly🔎 Investigate the biggest anomaly using alert thresholds
dashboard📊 Generate and open the HTML dashboard
warehouse-brief📨 Send warehouse health status to Mattermost/Slack

⚙️ Configuration Reference

config/config.yaml

The main configuration file. See config/config.example.yaml for the full schema with comments.

SectionPurpose
companyName and domain (appears in dashboard header)
bigqueryProject ID and dataset
sourcesToggle and configure each data source
competitorsDomains and names for competitive analysis
bofu_keywordsBottom-of-funnel keywords to track
contentSitemap URLs for content inventory
deliveryMattermost/Slack webhooks

config/alert_thresholds.yaml

Customize when alerts fire:

AlertTriggerSeverity
🔴 Traffic DropSessions < 80% of previous dayHigh
🔴 Paid DropAd spend < 70% of previous dayHigh
🔴 Revenue DropMRR < 98% of previous dayHigh
🔴 Churn SpikeChurn > 120% of previous dayHigh
🟡 Reach CrashSocial reach < 50% of 7d avgHigh
🟡 Engagement DropEngagement < 60% of 7d avgMedium

.env

Credentials and API tokens. Never committed to git. See .env.example for all supported variables.


📁 Project Structure

saas-growth-dashboard/
├── run_daily_growth_refresh.py      # CLI: refresh the BigQuery warehouse
├── run_growth_dashboard.py          # CLI: generate the HTML dashboard
├── config/
│   ├── config.example.yaml          # Configuration template
│   └── alert_thresholds.yaml        # Alert threshold defaults
├── growth_intelligence/
│   ├── __init__.py
│   ├── config.py                    # YAML + env config loader
│   ├── branding.py                  # Lazy brand/company context (name, domain, keywords)
│   ├── bigquery.py                  # BigQuery client wrapper
│   ├── schemas.py                   # Raw table specs + modeled SQL (raw → modeled → executive)
│   ├── orchestrator.py              # run_daily_growth_refresh: fetch → load → model → brief
│   ├── dashboard.py                 # HTML dashboard generator (all tabs)
│   ├── google_ads_client.py         # Google Ads API client helper
│   ├── http.py                      # Shared Google API session
│   ├── mcp_exec.py / claude_exec.py / agent_exec.py  # Optional MCP bridge helpers
│   └── sources/                     # One file per data source (30 connectors)
├── skill/
│   ├── SKILL.md                     # Claude Code skill definition
│   └── prompts/
│       ├── daily-ceo-report.md
│       ├── refresh-growth-warehouse.md
│       ├── investigate-anomaly.md
│       ├── generate-dashboard.md
│       └── send-warehouse-brief.md
├── demo/
│   ├── growth_dashboard_demo.html   # Simplified demo
│   ├── growth_dashboard_full_demo.html  # Full production demo
│   └── sample_data/                 # Synthetic data for all sources
├── .env.example
├── requirements.txt
└── LICENSE                          # MIT

Note on the MongoDB connectors: the six mongo_* / churn_risk / product_anomalies connectors are reference implementations built against a social-media-scheduling product schema. Remap collection names via sources.mongodb.collections in config.yaml, and expect to adapt their queries to your own product's schema. Every other connector is fully generic.


🔧 Adapting to Your SaaS

Step 1: Copy the source connectors

The source connectors in growth_intelligence/sources/ are already generic (they use GA4, GSC, ChartMogul APIs directly). Copy them and update only the config references.

Step 2: Customize your modeled tables

The modeled SQL in growth_intelligence/schemas.py defines how raw data is joined. Key customizations:

  • 🔄 Activation funnel stages — your product's signup → activate → convert flow
  • 📄 Content page type classification — blog, pricing, feature, etc.
  • ⚠️ Churn risk scoring signals — specific to your product's usage patterns

Step 3: Customize the dashboard

The dashboard generator in growth_intelligence/dashboard.py builds the HTML. Key customizations:

  • 🎨 Company name and branding
  • 📊 Department scorecards — which sections matter for your team
  • 🎯 Color thresholds — what CPA is "good" for your business?

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Priority areas:

  • 🔌 New source connectors — Add support for tools like Mixpanel, Heap, HubSpot, Intercom
  • 📊 Dashboard improvements — Better visualizations, new panels, mobile responsiveness
  • 🤖 Skill enhancements — New prompts, better anomaly detection, forecasting

❓ FAQ


📜 License

MIT — see LICENSE.

👨‍💻 Credits

Built by Sameer Ahmed Khan, founder & CEO of Social Champ.

Techstars Toronto '22 alum, 4x P@SHA ICT Award winner, and the guy whose social media tool Guy Kawasaki uses every day. Three failed startups before Social Champ worked. Software engineer by degree, marketer at heart.

LinkedIn X Instagram Facebook YouTube


Open-sourced because every SaaS founder deserves one place to see everything that matters and actually grow. 🚀

⭐ Star this repo if it helped you. It keeps us shipping.

Ver o original no GitHub

Bom para

Relacionados