ga-lite

Smaller, cacheable subset of Google Analytics JS client

This project is a non-official implementation for Google Analytics tracker script analytics.js. It uses the official API by Google to send analytics events to Google Analytics.

Version 2 is out! 🎉

Make sure you check the migration guide if you're upgrading.

Install to your project

The easiest way to install ga-lite to your project is to include the following script to your website:

<script>
(function(e,t,n,i,s,a,c){e[n]=e[n]||function(){(e[n].q=e[n].q||[]).push(arguments)}
;a=t.createElement(i);c=t.getElementsByTagName(i)[0];a.async=true;a.src=s
;c.parentNode.insertBefore(a,c)
})(window,document,"galite","script","https://cdn.jsdelivr.net/gh/neverbehave/ga-lite@master/dist/ga-lite.min.js");

// You may define your own endpoint as ga_api
windows.ga_api = "your api" // default https://www.google-analytics.com/collect

// You may use cf-worker as backend.
// Example in scripts/cf-worker.js

galite('create', 'UA-XXXXXXXX-X', 'auto');
galite('send', 'pageview');
</script>

This includes the most recent version of ga-lite to your site from the JSDelivr CDN, initializes the script with your own UA code and sends the current page's pageview event.

Hosting on your own server

If you're hosting the script on your own server, just grab the dist/ga-lite.min.js and upload it to your server. Then change the URL from the loading script (starting with https://cdn.jsdelivr.net) to point to your own server.

Install via NPM

If you're using a module bundler like Browserify or Webpack, You can also install the package straight with npm:

npm install ga-lite --save-dev

After the installation, you can use ga-lite in your code by requiring the package:

ES5:

const galite = require('ga-lite')

galite('create', 'UA-XXXXXXXX-X', 'auto')
galite('send', 'pageview')

ES6:

import galite from 'ga-lite'

galite('create', 'UA-XXXXXXXX-X', 'auto')
galite('send', 'pageview')

This imports the package from your node_modules/ and bundles it to your app's bundle file. This way you don't need to use the loader script to load the tracker from an external file.

Public API

This project uses same public API as the official analytics.js script. You can refer to the usage from the Google's official analytics.js documentation.

Only use the galite global function instead of ga function.

Migrating from analytics.js

As ga-lite uses same public API as the official analytics.js, the migration from analytics.js is really straightforward:

1. Include the ga-lite script
2. Replace any ga() function call with galite()

If you find any features that are not implemented yet, please open an issue.

Developing

If you have nvm, execute nvm i in the project directory to use a compatible version of node.

Install the project by running:

yarn

This installs the needed dependencies for developing this project.

Running tests

This project uses Jest unit tests to cover most of the code in this repository. Also Cypress tests are used to test behaviour in actual browser.

You can run the tests by running:

yarn test

This runs both unit tests and browser tests in your local machine.

Running unit tests

You can run only unit tests by running:

yarn test:unit

This runs all Jest tests inside test/unit-tests/ folder.

Running browser tests

You can run all browser tests by running:

yarn test:browser

This runs all the Cypress specs from test/browser-tests/specs/ folder.

Building

You can compile this project with the yarn command:

yarn build

Once the compilation has ended, webpack has compiled the assets from /src folder to /dest folder.

Do not add the dest/ folder files manually to your pull request. The deploy is done automatically.

Deploying

The deploy is done automatically by Travis. Creating a new release is done with the following yarn command:

yarn release

This command will ask for new version number and it will create the build commit and tag for you automatically.

You'll need to push the tag and commit to git manually.

Features

This project was first created to be a simple, cacheable alternative to Google Analytics to gain Google Page Speed test score of 100/100.

However the project exceeded expectations with popularity, and the version 2.0 is built to support many of the official analytics.js script's methods:

• Page Tracking
• Event Tracking
• Social Interactions
• User Timings
• Exception Tracking

How does this script differ from the official analytics.js script? Here's a comparison:

Feature	ga-lite	analytics.js
Supported by Google	No *	Yes
Size	~ 8 kB	~ 30 kB
Cacheable as long as you want	Yes	No (only 2 hours)
Open Source	Yes	No
Can be hosted on your own server	Yes	No
Use Cookies	No	Yes
Gets disabled on "Do not track" browser flag	Yes	No
Number of official features	Most	All
Can be installed from NPM	Yes	No
Can be bundled to your vendor.js bundle	Yes	No
Browser support	Modern	Modern and IE10+

* But uses Google's official, supported API

Privacy

While analytics.js uses Cookies by default, ga-lite only requires a single entry in local storage with key uid to store the visitor's identification.

Contributing

If you'd like to request a feature, or if you find any official analytics.js pubic feature missing, please open an issue.

If you'd like to contribute, please fork the repository and use a feature branch. Pull requests are warmly welcome.

Note: Do not add the dist/ folder files to your pull request, as the master branch should have the most up-to-date version's build files included. The new versions will be automatically fetched by jsdelivr to the CDN.

Motivation

This project was born, because it is currently impossible to use Google Analytics' official JS library to track your site if you want to achieve 100/100 in Google PageSpeed Insights.

This happens, because Google Analytics' official JS library has a cache header with the length of 2 hours. As PageSpeed Insights forces longer cache times, a custom GA library is practically the only way to achieve 100/100 points.

This project also makes it possible to:

• Get your site to load faster (as this is smaller lib than the official)
• Be sure your site won't serve any code you can't check yourself

You can read more from the blog post that's written about the library.

Migrating from v1

Note that version 2 is a complete, not backwards compatible rewrite from the initial ga-lite library.

You must now use the galite function to initialize your tracker and send the initial pageview. So the following code in v1:

var galite = galite || {}
galite.UA = 'UA-XXXXXX'

Will need to be converted to:

galite('create', 'UA-XXXXXXXX-X', 'auto')
galite('send', 'pageview')

Notice that ga-lite does not send any events or page views anymore on your behalf. You must explicitly call galite('send', 'pageview') to send the initial page view event.

Onunload tracking

The version 1 also automatically sent an event on the page's unload event. This was created, because Google Analytics only shows you "time on page" between the first and last event that you have sent.

Now normally this is fine, as you expect your user to browser different pages and spend a lot of time on your website. But when user bounces (visits only a single page), the time is not tracked realistically.

For example: User comes from Reddit to read your blog post. You only send the pageview event to Google Analytics. User reads the article and leaves the site. GA then shows the time spent on page as 0, even if the user did read the whole article and spent time with it.

This can be overcome by sending a beacon on the unload event, which v1 did automatically. To avoid any drop in your site's metrics, please include the following script to your page (after loading ga-lite):

<script>
  window.addEventListener('unload', function() {
    galite('send', 'timing', 'JS Dependencies', 'unload')
  })
</script>

FAQ

Why does ga-lite error with Pingdom?

When you run a website speed test with Pingdom's speed test, the connection to www.google-analytics.com that ga-lite makes fails. You'll see a connection error appearing:

Plus the connection to www.google-analytics.com in the page waterfall fails:

This error happens because Google Analytics blocks requests that the Pingdom bot makes. This error does not mean that ga-lite is broken or implemented incorrectly.

If you want to perform a website speed test that loads ga-lite correctly, use Webpagetest.org.

Licensing

The code in this project is licensed under MIT license.