FunnelQL: Funnel Query Language

Solution

Online customers can abandon their shopping cart for many reasons. As a store owner with access to Google Analytics there is information available about some of the reasons.

FunnelQL provides a solution to help customers complete their order. Pro-active communication, smart tricks with the design etc. Everything is possible with FunnelQL.

There is much to do about 'cart abandonment'. Experts estimate that 2 to 4 trillion USD per year is lost by abandoned shopping carts.

www.annexcloud.com/blog/31-shopping-cart-abandonment-statistics-for-2018/

• It’s costing a lot of money: Abandonment costs e-marketers about 2 to 4 trillion USD per year. Given the large economic impact, shopping cart abandonment solutions are becoming a priority for e-commerce marketers.
• It will get worse: Experts agree that shopping cart abandonment is on the rise.

FunnelQL can be an ideal solution and can help webshops recover more than a trillion USD in lost revenue.

FunnelQL Javascript Library

A javascript library for working with Funnel Query Language. The library enables to perform queries on the chronological context of activity of a website visitor for the purpose of Conversion Funnel Optimization.

The language and library are intended to be easy to use to unlock Conversion Funnel Optimization for users with in-depth insights into visitors/customers such as webshop owners and internet marketing specialists.

Documentation | Prototype (online demo)

Install

The package is available on NPM: npm install funnelql.

Include the library in the HTML document. The library requires a Web Worker that can be included as a inline blob or that will require a separate web worker file.

Option 1: library with separate web worker file

<script src="/funnelql.js" async></script>

The script will automatically load /funnelql-worker.js on the basis of the path of the funnelql.js file. You can define a custom location via window['funnelql-worker'] = 'path/to/funnelql-worker.js';.

Option 2: library with inline web worker

<script src="/funnelql+inline-worker.js" async></script>

To save additional bytes you can use the .nodebug.js version of each script that does not include debug and performance timing functionality.

How to use

Working with Funnel Query Language consists of two parts:

1. writing the funnel by registering tags during the visit of a website visitor.
2. querying the funnel to perform actions based on a callback or promise (e.g. a Push Notification).

Setup

Funnel Query Language is based on a IndexedDB database that can persist or expire with the browser session. Persistence enables to target returning visitors or visitors that visited on a specific day or time.

$FQL.init({
  expire: 'session'
});

Init Options

Option	Description	Default
expire	Database expiration.	session (could be never or a time in seconds)

Writing A Funnel

The Funnel data-set is based on tags that can be set with the following method:

$FQL.tag('tag-name');

During navigation, URLs are automatically registered. The following method can be used to manually register a navigation:

$FQL.nav('URL or path/');

Querying A Funnel

The following method can be used to perform a query based on a Funnel Query string or JSON:

$FQL.query( "funnel query" ).then(function(result) {
  // result = true/false (boolean)
});

The following method can be used to parse a Funnel Query string into JSON:

$FQL.parse( "funnel query" ).then(function(json) {
  // json
});

The following method can be used to perform an action as soon as a Funnel Query string or JSON matches the path of the visitor:

$FQL.on( "funnel query" ).then(function(result) {
  // result = true 
});

Alternatively, the on method accepts a callback instead of a promise and multiple queries as a Array.

$FQL.on( [ "funnel query 1", "funnel query 2" ], function(result, fql) {
  // result = true 
  // fql = funnel query string
});

The library can provide detailed debug information including Performance API timings in the browser console. The following method can be used to enable debug mode or to retrieve it's status.

$FQL.debug(true);

Funnel Query Language

How to construct a Funnel Query?

Funnel Query Language provides several methods for querying the registered funnel data-set.

• Tag (JSON string)
• URL (URL)
• Path (PATH)
• Tag Count (COUNT)
• Date or time (SINCE)
• Custom function (FN)

Tag Query

A tag can be queried using a JSON string "tag-name" or using a Regular Expression method REGEX(/tag-name/i).

"tag-x" + REGEX(/tag-name-(\d+)?/i) + "tag-y"

URL Query

An URL can be queried using the method URL("url") or using a Regular Expression URL( REGEX(/tag-name/i) ).

URL("url/to/match") + URL( REGEX(/url\.(html|js|css)/i) )

Path Query

The PATH query enables to add chronological context to a query. The path method starts a sub query that can contain a tag, url, count or function query.

The path query provides operators to define the chronological context. The path sequence can be based on URL navigation position or time.

Path Query Operators

Option	Description
>	Matches in consecutive query should follow previous query's URL position.
>=	Matches in consecutive query should match or follow previous query's URL position.
|>	Matches in consecutive query should be on same URL position as previous query's URL position.
>10	Matches in consecutive query should be at least 10 URL positions from previous query's URL position.
>10m	Matches in consecutive query should be at least 10 minutes from previous query's event position.
<10m	Matches in consecutive query should be within 10 minutes from previous query's event position.

PATH( "tag-x" > "tag-y" )

The following example matches when tag-y follows within 1 minute after tag-x.

PATH( "tag-x" <1m "tag-y" )

The following example matches when the visitor visits an URL containing cart or other-page within 5 navigations after having visited product-1.html.

PATH( URL("product-1.html") <5 URL( REGEX(/(cart|other-page)/) ) )

Since Query

The SINCE query enables to add chronological context to a query. The since method starts a sub query that can contain a tag, url, count or function query.

The first parameter of the SINCE method can contain a time reference, e.g. 1s (1 second) or 10m (10 minutes) or a date in format yyyy-MM-ddThh:mm.

SINCE( 1m, "tag-x" )

SINCE( 2018-09-02T12:10, ( "tag-x" + "tag-y" ) )

Count Query

A COUNT query enables to query based on number of occurrences of a tag.

The first parameter of the COUNT method contains the tag as JSON string or as REGEX query. The query supports the operators >, <, =, <=, >= followed by a number.

COUNT( "tag-x" > 1 )

Function Query

A FN query enables to perform a custom query using a javascript function. Optional conditions from a parent query are added as a extra argument.

FN( "function-name", "argument-1", "argument-2" )