@digitaloptgroup/analytics NPM

Digital Optimization Group - Javascript Analytics

This library is made to work with Digital Optimization Group's Headless A/B testing CMS. It provides a range of event tracking automatically, plus it provides a public api to allow custom event tracking.

At 15kb gzipped it's lighter than Google Analytics.

What's tracked by this library?

a/b test variations actually appearing in a user's viewport (not just assignment)
time a variation is in a user's viewport
time on site and time on page
visibility of the page (is tab visible?)
mouse distance
total scrolling distance by page
scroll depth per page
errors
arbitrary outcomes set up by the user of this library
page views
orientation changes (mobile devices)
mousedown activity on a variation

All events are placed on a timeline so that outcomes can be properly attributed to variations actually appearing in the viewport. For example, we shouldn't attribute an order that occured before a variation was in the viewport to that variation when running an a/b test.

Usage

npm

npm install --save @digitaloptgroup/analytics

yarn

yarn add @digitaloptgroup/analytics

Initialze

import { initTracker } from "@digitaloptgroup/js-analytics";

const config = {
  apiKey: "your-api-key",
  projectId: "your-project-id",

  // optional advanced configuration
  // see advanced further down this README
  vid: "visitor_id",
  rid: "request_id",
  startTimestamp: "proxy_startTimestamp"
};

const {
  pathChange,
  outcome,
  caughtError,
  getIntersectionObserver
} = initTracker(config);

Track Pageviews

const pathname = "/about-us";
pathChange(pathname);

Track Outcomes

Tracking events with outcome is intended as a means to track any custom user actions that you would like to associate with your a/b test variations.

outcome takes two arguments outcomeName and metadata. The name must be a string. The metadata must be an array containing an arbitrary number of objects with the type signature { key: string, value: string }.

const outcomeName = "addToCart";
const metadata = [
  { key: "sku", value: "product_123" },
  { key: "price", value: "125.99" }
];
outcome(outcomeName, metadata);

Caught Errors

If you would like to report a caught error you may do so with caughtError.

const metadata = [
  { key: "type", value: "auth_failed" },
  { key: "url", value: "/login" }
];
caughtError(metadata);

Observe / Unobserve Variation

initIntersectionObserver provides viewport tracking for a/b test variations. It is intended to be used with Digital Optimization Group's Headless CMS.

Automatically Tracked Events

In addition to the events listed above, this library automatically tracks the following events:

Euclidian Mouse Distance

The library will sample mouse position every 250ms and calculate the euclidean distance from prior x and y coordinates. It will report the total every 3 seconds.