Open Analytics

Self-hosted, accurate analytics tool for your website in a clean minimalist interface.

Demo (w/ fake data)

This project is a derivation of open-analytics-core. So like its predecessor it logs the following to a MongoDB database: Device Ip, User-agent, IsMobile, Date, Active Time (Seconds), Visited url, Location, Referrer...

And uses those data to visualize:

Average time on each page
Current active users
Most used device (Mobile/Desktop)
Visits by each country
Websites most users are comming from
Top visited links
New vs returning users
Sum of all users activity per day

Setup

First Step:

Clone this repository.

Second Step:

Add this code inside the body tag of whatever page you want to include in the analytics:

// If you need to include in multiple pages, add the latter script tag to a seperate .js file to avoid unnecessary lines of code.
<script src="https://cdn.socket.io/4.4.0/socket.io.js"></script>
<script>
    let isFirstTime = true;
    if(document.cookie) {
    	isFirstTime = false
    } else {
        document.cookie = "name=hasVisited;path=/";
    }
    const socket = io("ADDRESS", {secure: true}); // replace ADDRESS with the ip of the api address & port.
    socket.on('socketClientID', function (socketClientID) {
        socket.emit('clientMessage', {url: window.location.href, referrer: document.referrer, isFirstVisit: isFirstTime});
    });
</script>

Third Step:

Edit the ./src/config.json file accordingly:

{
    "api": { // Edit here accordingly
        "port": 8070,
        "uri": "mongodb://localhost:27017",
        "databaseName": "DatabaseName",
        "collectionName": "CollectionName",
        "address": "http://192.168.1.7" // DONT include the "/" at the end
    },
    "tls_support": { // Leave this as is if you dont need HTTPS support
        "enabled": false,
        "key": "", // key.pem file path (relative to "./api/server.js")
        "cert": "" // cert.pem file path (relative to "./api/server.js")
    }
}

Fourth Step:

Run (cd api && node server.js) & npm start in the main folder, the dashboard will be available in http://localhost:3000.