Define KPI - Number of Visitors

[Aidajafarbigloo]

Establish a KPI to track the total number of visitors.

[Sundraiz-Shah]

@Aidajafarbigloo

- Viable Option to Solve KPI and Cookies Consent

1. Session-based tracking (no IP)
   We use Django's built-in session framework to identify unique visitors. A session is basically a temporary ID assigned to each browser — it tells us "this is a new visitor" without storing any personal data. We chose this because it's privacy-friendly, built into Django already, and requires zero extra tools.
2. Custom Middleware for counting
   Middleware sits between the user's request and your app — it intercepts every single request automatically. We chose this because it works silently in the background, touches no existing code, and is easy to turn on/off.
3. Django Admin as the KPI Dashboard
   Instead of building a separate dashboard, we use Django's built-in admin panel to display the daily stats table. We chose this because it's already there, free, and easy to setup.
4. Cookie Consent Banner (custom HTML)
   A simple Accept/Decline banner in the base template. We chose this because it's lightweight, requires no paid service, and covers basic GDPR compliance.

-What Other Options are Available

1. Visitor Tracking
   Google Analytics: popular but requires complex consent setup
   Plausible: privacy-friendly, no cookies, but a third-party script
   Umami: self-hostable, cookieless, but needs separate server setup

2. Request Counting
   Server Logs (Nginx/Apache): no Django code needed, but requires extra tooling to read and analyze
   Django Silk: great for debugging but not designed for KPI monitoring

3. KPI Dashboard
   Grafana: powerful and visual, but overkill for simple counts
   Metabase: user-friendly dashboards, but needs its own infrastructure
   Power BI: great for business reporting, but heavy and costly for this use case

4. Cookie Consent
   Cookiebot: handles GDPR automatically, but has a paid tier
   Osano: solid compliance tool, but again paid for full features
   django-cookie-consent: a ready-made Django package, but more complex to configure than we need right now

-Why We Didn't Choose These Options
Our goals are simple count visitors, count requests, no personal data, no extra infrastructure, fast to build. Here's why each alternative falls short of that:

1. Visitor Tracking

• Google Analytics
  GA is a powerful tool but it's designed for marketing teams, not developer KPIs. It tracks way more than we need — demographics, behavior flows, ad conversions. More importantly, it stores data on Google's servers, requires explicit cookie consent setup, and is increasingly blocked by ad blockers — meaning our visitor count would be inaccurate anyway. It's overkill for a simple visitor counter.

• Plausible
  Plausible is actually a great privacy-friendly tool and cookieless by design. However, the hosted version is a paid subscription. The self-hosted version requires setting up and maintaining a separate server. Since we already have Django, spinning up extra infrastructure just to count visitors doesn't make sense for our current goal.

• Umami
  Same story as Plausible — self-hostable and privacy-friendly, but it needs its own dedicated server, database, and maintenance. That's a disproportionate amount of effort when our goal is simply to see daily visitor and request counts inside our existing app.

2. Request Counting

• Server Logs (Nginx/Apache)
  Web server logs do record every request, so the raw data is there. The problem is that logs are just plain text files — you'd need a separate tool like GoAccess or ELK Stack to parse, aggregate, and visualize them. That's an entirely separate pipeline to build and maintain. Our goal is to see counts quickly inside our existing Django app, not manage a logging infrastructure.

• Django Silk
  Silk is a profiling tool built for development environments — it helps developers debug slow queries and inspect requests during testing. It stores detailed data on every request which is great for debugging but actually a privacy concern in production. It's also not designed to surface KPIs like daily totals. Using it in production for a manager-facing dashboard isn't what it was built for.

3. KPI Dashboard

• Grafana
  Grafana is the industry standard for monitoring dashboards and it looks impressive. But to use it, you need to set up Prometheus to collect metrics, a time-series database to store them, and then Grafana to visualize them. That's three separate systems to configure. Our goal is a simple daily table — not a real-time monitoring suite. The setup cost far outweighs the benefit at our scale.

• Metabase
  Metabase is a business intelligence tool that connects to your database and lets non-technical users build reports. It's useful, but it requires its own server and installation. We'd essentially be running two applications just to see two numbers. Django Admin already shows our SiteStats table cleanly — Metabase adds nothing extra for our use case.

• Power BI
  Power BI is a Microsoft enterprise tool built for complex business reporting across large datasets. It has licensing costs, requires data connectors, and is designed for business analysts, not developers tracking web requests. It's solving a completely different problem at a completely different scale than ours.

4. Cookie Consent

• Cookiebot

Cookiebot automatically scans your site, categorizes cookies, and generates a legally compliant banner. Sounds ideal — but the free tier is very limited and the paid plans are subscription-based. Since our consent needs are minimal (one banner, one consent cookie), paying a monthly fee for a tool that does far more than we need isn't justifiable.

• Osano

Similar to Cookiebot — an enterprise-grade compliance platform. It's built for companies managing dozens of third-party cookies across multiple domains. We have one consent cookie to manage. Using Osano would be like hiring a full legal team to write a single paragraph.

• django-cookie-consent package

This is actually a solid Django-specific package and was the closest runner-up. It supports consent groups, logging, and template tags out of the box. However, it adds model migrations, URL configurations, and template tag setup that we don't need for a simple Accept/Decline banner. Our custom banner does the same job in some lines of HTML and JavaScript.