This post tests which browsers are most susceptible to unique browser fingerprints and configurations you can hide behind.

First, let’s re-visit what a browser fingerprint is

Panopticlick and a number of other sites have discovered that by taking all of the information about a user’s browser (such as the web browser itself, operating system, installed plugins, screen resolution, bits per pixel, system fonts and other elements about their configuration), it is possible to generate a unique ID for a vast majority of browsers accessing their site. The idea behind it is that it can uniquely identify a browser (without the use of cookies) throughout the internet – whether or not a user clears their cookies or even has them enabled.

Test your browser’s uniqueness on Panopticlick’s website.

Browser fingerprints as I captured them, contain the following information which is then stored as an MD5 hash in Google Analytics (alongside no other unique IDs, except country, city and ISP/service name):

• User agent
• Screen resolution & depth
• Timezone offset
• Local & session storage
• Browser plugins (Skype, MS Office, Adobe software, Steam etc)

Here is what my browser fingerprint roughly consists of:

“Mozilla/5.0 (Windows NT 6.2; WOW64) AppleWebKit/537.31 (KHTML, like Gecko) Chrome/26.0.1410.64 Safari/537.31###1080x1920x32###-600###true###true###Shockwave Flash::Shockwave Flash 11.7 r700::application/x-shockwave-flash~swf,application/futuresplash~spl;Chrome Remote Desktop Viewer::This plugin allows you to securely access other computers that have been shared with you. To use this plugin you must first install the Chrome Remote Desktop webapp.::application/vnd.chromium.remoting-viewer~;Native Client::::application/x-nacl~nexe;Chrome PDF Viewer::::application/pdf~pdf,application/x-google-chrome-print-preview-pdf~pdf;Google Update::Google Update::application/x-vnd.google.update3webcontrol.3~,application/x-vnd.google.oneclickctrl.9~;Intel® Identity Protection Technology::Intel web components for Intel® Identity Protection Technology::application/x-vnd-intel-webapi-ipt-2.1.42~;Intel® Identity Protection Technology::Intel web components updater – Installs and updates the Intel web components::application/x-vnd-intel-webapi-updater~-2-0;Microsoft Office 2013::The plugin allows you to have a better experience with Microsoft SharePoint::application/x-sharepoint~,application/x-sharepoint-uc~”

Under an MD5 hash, it could resemble the following:

3c674d0f6ccce277b37370bc8ca6bxxx

I then store it in Google Analytics as a custom variable and then I mash it together with IP information available in Google Analytics (hostname, city and country). So, given the following fingerprints, I count all three as being unique:

3c674d0f6ccce277b37370bc8ca6b878 (Brisbane, Australia, ISP: Telstra)
3c674d0f6ccce277b37370bc8ca6b878 (Brisbane, Australia, ISP: ACME Internet)
3c674d0f6ccce277b37370bc8ca6b878 (Gold Coast, Australia, ISP: ACME Internet)

Again, it’s possible this visitor IS travelling between access points and cities (particularly for mobile users) but let’s save that for another day. With the methodology out of the way, now you know to take the data with a healthy dosage of salt.

But it’s not all that unique or reliable…

Certain configurations are simply too common to be able to accurately pin down (e.g. iPhone). Secondly, so many factors can alter your browser fingerprint that it’s simply unfeasible to use alone as a method to reliably and uniquely identify users to your site. For instance, the following things can alter your browser fingerprint:

• Updating your browser to the latest version
• Installing new plugins
• Using multiple browsers (sorry porn browsers, incognito/private mode is not a different browser)
• Changing your screen resolution (or using a laptop to plug and play monitors all day)
• Enabling and disabling cookies or other browser functionality
• Disabling JavaScript
• Switching timezones (including daylight savings)

Every time Chrome updates, your fingerprint will change

Just look at how often Google releases new versions of Chrome! Thanks to Andre Mafei for this excellent visualization of Chrome updates:

Each one of these blips indicates a new version – and with each new version comes a new fingerprint.

Browser uniqueness

Browser fingerprints for browsers are surprisingly unique for desktop based browsers. For mobile devices, particularly those on iOS (thanks to Apple’s iron grip), are not very unique when you look at the variety of browser fingerprints. Don’t get too giddy Apple users, as you will soon see this doesn’t matter much at all:

Look what happens wen you mash it together with IP information…
Boom! Unfortunately (or fortunately for that matter), mashing fingerprints together with geo IP and network information available in GA (network domain, city, country) paints a very different picture:

All browsers that were once considered relatively anonymous are now >99% uniquely identifiable.

Operating system uniqueness

No surprises here that you have virtual anonymity when using an iOS based device.

Now let’s add a dash of IP related information into the mix
Muahaha… Once more, the moment you pass in IP information, the playing field is leveled:

One application: Plausible solution for cookie-less tracking

One of the more prominent web analytics vendors, Adobe has introduced cookieless identification of unique browsers which it achieves through a combination of browser fingerprints and a user’s IP address. I’m not certain how they perform this, but I would imagine they use something similar. Even if it’s not 100% accurate across sessions, it should give marketers a clear picture 90% of the time. And this is OK – remember, no web analytics tool is ever going to be completely accurate.

If you have a high proportion of visitors who do not allow cookies (a 2009 study revealed roughly ~5% of users do not allow first party cookies), then you may also have a use for this.

I would also be surprised if Google and other big ad networks WERE NOT testing or using fingerprinting to track visitors throughout the web.

Future exploration of fingerprinting

There are a number of things which I would really like to explore further in the realm of browser fingerprinting, even though I have essentially sidelined it as a useless dimension.

Using data in GA to develop your own fingerprints
I haven’t tried, but it may be possible to use GA data to generate fingerprints. E.g. combine all the browser configuration information and count the unique instances.

Tracking it alongside Google Analytics visitor IDs
This would allow you to see if visitors are clearing cookies and what proportion of them do this. Alternatively it may provide insight into how unique fingerprints actually are.

Tracking IP address and Browser Fingerprint together
Even more granular than network domain in GA is the specific IP address. Unfortunately even though this is more granular, it’s susceptible to dynamic IPs which will add to the volatility of your fingerprints.

Adding Adobe Flash into the mix
As Panopticlick data shows, one of the most uniquely identifying bits of information – fonts installed – can be accessed through Adobe Flash. This will greatly improve the uniqueness of fingerprints.

Perhaps this was Adobe’s evil master plan for analytics domination.

How users can avoid being uniquely fingerprinted

• Use a common browser configuration (e.g. get the most popular phone in your region and use the default browser)
• Avoid using Adobe Flash (Flash allows access to reading system fonts installed – one of the most uniquely identifying sources of information)
• Use a different IP address regularly (e.g. proxies)
• Update your browser and other software regularly
• Disable JavaScript (to avoid any probing scripts)
• Fake your user agent (personally I like to browse sites with IE 5 on WIndows 3.1 just for shits and giggles)

How browser fingerprinting can be augmented

1. Never use it alone – always tie it back to a user’s cookie, IP address, customer ID or all of the above
2. Rather than taking a hash of the user agent and a host of other identifying features, collect all the pieces of identifying information individually and account for small updates here and there.

Anyway, I hope you found this analysis on browser fingerprinting useful. Obviously there’s a lot more to it than what I have covered above but it may help answer some questions explored on sites like Reddit, Github and Stack Overflow.

What are your views on unique browser fingerprints?

Like this post? Do me a favour and share this article...

Web Browsers Most Susceptible to Browser Fingerprinting is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

I was exploring creating my own split testing JavaScript the other week until I stumbled onto this amazing piece of work by James Yu, called Cohorts.js. With a couple of updates, I’m excited to start using this on my own sites and clients’ sites with a low budget.

To sum it up, this post details how to implement a poor man’s Optimizely.

First, how do JavaScript split testing tools work?

Breaking it down, a split testing tool must handle a bunch of jobs:

• Determine which visitors see which variation
  • Randomly assign a sample of visitors to view each treatment/variation
  • Have they been previously allocated?
• Serve the correct variation to visitors across different pages and different visit sessions
• Track the results in an analytical tool

If you’re like me, you especially want to measure the experiments in Google Analytics, SnowPlow Analytics or Site Catalyst.

When you think about the various split testing tools out there, such as Optimizely, Visual Website Optimizer, Webtrends Optimize, Adobe Test & Target and so on, they’re all JavaScript based. It’s just an easy and effective way of running split tests.

How Cohorts.js works at a high level

Essentially it allows you to configure A/B or multivariate split tests with a pure JavaScript framework. To get it up and running, you may need the help of a developer or someone with a bit of technical clout.

• Publish the cohorts.js object in the header: Download
• Create the testing object and configure it to work with your analytics tool(s) of choice
• Analyse the results in GA

Check it out on Github.

Running a split test with Cohorts.js

1. To get started, you will need to have published the script within your document’s header i.e. before your closing tag:

<script type="text/javascript" src="cohorts.js"></script>

You can download it from here (right-click, “Save as”).

For best results you can serve this code from a speedy CDN like Amazon S3 and Cloudfront. Though, keep in mind this code is already very lean and once it’s minified and gzipped, it should come to a kilobyte or two in weight.

2. All you need to write is a few lines of JavaScript to set a test up:

$(document).ready(function() {
	var a_test = new Cohorts.Test({
		name: 'a_test',
		scope: 1, // Sets the scope for the test and custom variable: 1: Visitor, 2: Session, 3: Page 
		cv_slot: 5, // Sets the custom variable slot used in the GoogleAnalyticsAdapter
		sample: 1,
		cohorts: {
			variation_a: {
				onChosen: function() {
					$('p.message').html('Hello!');
				}
			},
			variation_b: {
				onChosen: function() {
					$('p.message').html('Sup?');
				}
			},
		}
	});
	$('p.message').click(function() {
		a_test.event('Clicked on message'); // Track any evens with your storage adapter
	});
	// Add more test objects as you need
});

How easy is that? Very easy (in case you didn’t know the answer to that rhetorical question).

Breaking the test object down:

• a_test = new Cohorts.Test: Can be named anything you like e.g. wave_9

The values within this object can also be configured:

• name: Insert the name of the experiment here. This will categorise your custom variables and events in GA, SP or whichever tool you’re using.
• scope: Set the scope of the experiment and your GA custom variables. I.e. make the experiment and tracking persist over a page (3), visit session (2) or visitor’s lifetime (1).
• cv_slot: Here, you can set the slot number for the experiment, useful for if you ever decide to run multiple experiments on the site at once.
• sample: How many visitors you want to expose to the experiment for 50%, you would set this to 0.5 and so on.
• cohorts: Here you can define which variations will show.
  • onChosen: Inside the function for this object, you can make any change you like to the page, using JavaScript – hint, for big changes, try using Ajax.
• storageAdapter: This object holds all the values for tracking and measuring your experiment.
  • nameSpace: This will categorise your events in your analytics tool.
  • trackEvent: The function that runs your tracking in each of your tools.
  • onInitialize: When the experiment runs, this function also runs, to start the tracking.
  • onEvent: Whenever you run “test_name.event(‘event name’);” this function will run.

3. Preview the test by including the following in your URL.

• #{test object name}={cohort name}

e.g. for the example above it would be:

• http://www.optimisationbeacon.com/#font_test=big

4. To stop the test, just remove the code.

5. Analyse the results within Google Analytics along with my split test calculator:

First things first, you will want to jump into Google Analytics and run this custom report:

This uses unique visitors – as this is what most split testing tools use to calculate success.

As the name suggests, you will see unique visitors by custom variable. Anyway, download the report here:
Unique Visitors by Custom Variable

Once you have the report up, enter the values into my test calculator here, like so:

Here you can calculate the results of your experiment.

…and it will tell you auto-magically, if you have a winner or not. Clearly, I need to keep running this experiment for a statistically significant result.

Pro tips to make the most of it

• Serve the script from a CDN, like Amazon Cloudfront and S3 to ensure minimal page load time impact (not that you should be concerned given its tiny footprint)
• Set the expires headers for cohorts.js to a small value, so that you can quickly pause the test
• If you want to preview the script in a production environment simply set the sample variable in the test object to 0
• Beware that you can only fit so much text in the custom variable, so keep your test and cohort names short yet meaningful
• This tool is like a swiss-army knife when you’re using jQuery – I highly recommend learning even just a little bit of this
• Make sure you’re not overwriting existing custom variable slots in GA – remember, there are just 5 slots
• Serve all your experiments from the cohorts.js file, and use domain.location.pathname to determine which URLs you want to run your experiments on

The verdict

In summary, it’s a very clean and simple tool. What it lack in bells and whistles, it compensates for it in speed, control and price. Of course, it goes without saying that you need some development clout to make the most of this simple tool.

Kudos once more to James Yu for writing an excellent framework. I hope my post brings this tool to light and spurs on future development work. There’s tons of features that can be added that will make this tool exceptionally powerful.

Benefits of Cohorts.js over other tools

• 100% control over how the code works & any integrations
• Leaner than various other split testing scripts
• Able to be self-hosted, making it even speedier
• Completely and utterly… Free

Cons of Cohorts.js (at this stage)

• Need to perform your own analysis
• Not built for marketers (though that may be easy to change)
• Doesn’t offer the same rich feature-sets offered by paid tools

To do

I think a number of things would really help improve this tool and boost its adoption. Here’s a list of things which I think would really add value to this tool:

1. Easy to use test setup (similar to what Optimizely and Visual Website Optimizer Offer)
2. Segmentation tools – for bucketing users and serving them particular messages
3. Page targeting script – i.e. simply use regex to determine which pages a test should initialize on

Any savvy developers out there should try contributing to James’ project. In the meantime, you can simply fork cohorts.js from me and I’d be more than happy to pull requests that help with the above goals. Be sure to reach out to me on Github.

Let me know how you might use this in the comments…

Like this post? Do me a favour and share this article...

Cohorts.js: Open Source JavaScript MVT Split Testing Framework is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

“Oh, it’s those people who type in your domain name. Yeah that’s it.”

Or:

“Umm… they’re the people who bookmarked your website.”

Sure they’re two possible sources of direct visits but as I’ll explore in this article, they aren’t always. This is kind of a shame because direct traffic usually has a strong intent which you’ll see in terms of behaviour and conversions. Unfortunately most people just don’t know the definition of a direct visit in GA.

The true definition of a direct visit is any visit that lands on your site without a referrer. A direct visits will not override an existing campaign (search, referral or others).

Therefore, Direct is not always Direct in GA or Site Catalyst

Even if someone visits your site directly, they’re not always going to count as a direct visit. Confused? You should be (it confused the bejesus out of me when I first encountered this). Take this situation for example:

Johnny visits your site through an AdWords campaign and returns a week later directly through a bookmark (aka. while the _utmz cookie still exists). In this case, both visits will be attributed to AdWords.

That’s why 3-month-old AdWords campaigns still refer you traffic, long after you’ve paused them.

Anyway, here are a few different sources which you may not have considered. Is this list exhaustive, Rob? Nope… but it covers most bases.

1. Type-in Traffic

These are usually your existing customers, loyal readers or people that simply heard about you through some offline channel.

2. Links sent through instant messaging

Whether it’s SMS, Whats App, MSN Messenger, AIM, Yahoo Messenger, Skype, Kakao Talk or Google Talk – all of these instant messaging programs will not provide a referrer, and show up as direct. The exception to this is web-based chats which are not done over HTTPS or running through Java/Flash/Silverlight etc.

Even web-based chats appear to send direct visits (at least under HTTPS)

3. Un-tagged emails (Outlook, desktop clients and secure web-based email)

Whether your visitors are sharing links with their buddies or you send out an email without any URL tagging on it, visitors coming from secure web mail or desktop email clients (Outlook or Thunderbird) will also show as direct visits. Fortunately, quite a number of people don’t enable HTTPS for their web-email clients, so you may see GMail, Yahoo Mail and Hotmail showing up under your referrers section.

This traffic can show as direct – especially when untagged emails are sent

4. Opening a link in “Incognito Mode”, “In-Private” or similar

Potentially a big source for those… ahem… dubious websites, could be “Open link in (incognito|In Private) window” – of course visitors . If you were to open a link in a new tab or in a regular new window, however, the referrer will still be passed (at least under Google Chrome).

This is potentially a bigger issue for sites with dubious content within the site – think “18+ sections.”

Opening in a new window will provide a referrer, but not opening in incognito mode

5. Bookmarks

Typically if you’ve bookmarked a site, chances are you’ve been there before. This means that if your previous campaign is still available in your cookies, this won’t result in a direct visit. If it’s been 6 months since you last visited the site, you’ve cleared your cookies or you’ve imported your bookmarks from another browser, you will certainly show as a direct visit.

Opening bookmarks do not provide a referrer – and therefore can lead to direct traffic

6. Use of rel=”noreferrer” in the previous link

The HTML5 specification includes rel=”noreferrer” which basically means that anyone following a link with this attribute and value will land on the following URL without a referrer. Don’t expect this to be a huge source of direct traffic, however.

Coming in HTML5, rel=”noreferrer” will also be a source of direct visits (arguably very few though)

7. Visitors from HTTPS websites

Part of the HTTP specification recommends that browsers do not pass along a referrer to non-HTTPS sites. The wording of this is slightly vague, as it reads, “should not” rather than, “must not”. This is certainly an odd point as it sometimes does provide a referrer.

For instance, searching from SSL Google.com will result in a referrer on the URL you land on. So this isn’t a hard and fast rule, but it’s generally the case.

8. Links within mobile and desktop apps

Lastly, if you’re still scratching your head about a huge number of direct visits, could they be arriving through links within your mobile/desktop application?

EDIT: A note about iOS6 and other mobile operating systems, thanks to Alistair Lattimore in the comments section below. Many apps use the Safari browser and links will therefore pass a referrer to your site. Analytics distinguishes this traffic under the browser “Safari (in-app)”. This is also the case with other devices, however I’m unsure how in-app browsers show up in GA though.

9. Browser homepage set to your website

If you’ve got a large enough company that sets its browser homepages to default to your website, then this will surely influence your direct visits as well.

10. Hacks/extensions to hide referrers

A very small number of visitors will intentionally hide their referrers using extensions like this one for Chrome and others for Firefox.

Also, some browsers will not carry a referrer from some JavaScript initiated links.

A (very) subjective guess on the liklihood of direct sources

This is just my subjective view on how likely you are to receive visits from particular direct traffic sources.

As good as direct traffic is, if you’re not tagging all of your campaigns, it’s going to be polluted. At least now you can.

“…but Rooooob. I really want to track direct shares.”

You can! Well, kind of…

If you’re familiar with the concept of dark social (that is links shared through email, instant messaging and whatnot which would otherwise show up in your GA as direct), then you can kind of track this. In the case of The Atlantic’s article, dark social accounts for a large majority of social on the web.

That big blue chunk isn’t Facebook – that’s dark social traffic to The Atlantic.

It’s important to track right?

Well, yes. And it can be surprisingly simple to do. Here are two ways:

Tracking direct shares of URLs using advanced segments
Realistically, you’re not going to type in a long link such as “http://www.optimisationbeacon.com/analytics/using-brower-fingerprinting-in-google-analytics/” directly into your browser. Therefore, you can assume that anyone landing on a page other than your homepage is coming through dark social. Here’s an advanced segment I cooked up earlier:

Download here: Dark social (or bookmarkers)

It can help you answer valuable questions such as, “Which content of mine is being shared most?”, “Which content is getting bookmarked by visitors?”:

After conducting an analysis of dark social traffic, I now know which of my posts are drawing bookmarkers and direct shares.

Tracking direct shares of your URL using code (caution: experimental/pseudo code)
According to Richard Fergie, you can track copy/paste shares on the web very simply:

<script>
function addTrackingParameters() {
    if (history.pushState) { //feature detection
        var currentURL = window.location.href;
        var trackingParams = '?utm_src=ctrl-c+share&utm_medium=direct';
        var newURL = currentURL + trackingParams;
        window.history.pushState(null, document.title, newURL);
    }
}
</script>

This works by adding URL tagging to the end of the link after the page has loaded. When a visitor copies the link, the URL tagging will go along with it, carrying “ctrl-c share”.

…and Bob’s your uncle – you’d probably be tracking direct shares of your URLs in GA. Beware, as I haven’t tried this yet. Let me know in the comments if you get around to trying this one out. I would love to get in touch with you.

Like this post? Do me a favour and share this article...

10 Sources of Direct Traffic & How To Track Direct Shares is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

A few years ago, you may have read my post about Tynt – a service that lets you track copy/paste in their interface. I liked it, but I didn’t love it because:

1. The hefty script added to my page load time
2. It saved all the metrics in its own interface (which I never logged into)
3. None of the metrics were visible within the context of all my other valuable analytics data

I just had to find a better solution…

Hunting for an Alternative

This annoyed me, so earlier this year I did some Googling and found an excellent jQuery solution on Onderweg. This script had everything I needed but it was missing a few things:

• I didn’t want another jQuery plugin to keep track of
• I wanted to track more than a handful of letters in the clipboard
• New lines and paragraphs were not represented at all
• The length of the copied string was never tracked
• The events were counted in the bounce rate calculation
• Page paths were not tracked inside the event – making some analyses difficult

Therefore, I hacked away at it, tailoring it to suit my needs. And this is what I came up with:

Tracking Copying to Clipboard in Google Analytics

Simply add this script after your GA script and jQuery library (Note – Requires jQuery and async version of the Google Analytics script):

<script type"text/javascript">
// Track copied content adapted from Onderweg & Tim Down by Robert Kingston - http://www.optimisationbeacon.com/
// Get Selection Text function by Tim Down - http://stackoverflow.com/a/5379408/458627
function getSelectionText() {
    var e = "";
    if (window.getSelection) {
        e = window.getSelection().toString()
    } else if (document.selection && document.selection.type != "Control") {
        e = document.selection.createRange().text
    }
    return e
}
jQuery(document.body).bind("copy cut paste", function (e) {
  var content = getSelectionText();
  var contentClean = content.substring(0, 499).replace(/\\(n|r\\n|r)/gm, "\\n "); // Represent new lines
  var length = content.length;
  _gaq.push(['_trackEvent', 'clipboard', e.type+' location: '+document.location.pathname, contentClean, length, true]);
});
</script>

This tracks copying from anywhere on the page, including line breaks and up to 500 characters of text. There’s room to copy much more (the POST to __utm.gif supports up to 8,192 bytes per request), but I figured, 500 chars should suit most purposes.

Hey, you can always bump this up, yourself.

There you have it! Easy as one, two three. Here’s what it’ll look like in the Google Analytics interface:

Whether they copied, cut or pasted is captured in the event action alongside the URL path.

Contents of the clipboard are captured in the event label.

Under event value you will find how many characters were copied at once.

… and for the sake of organisation, copy/paste/cut events will be filed under “clipboard”.

Just for the hell of it, here’s an example of the content people have copied from my site. No surprises, here – it’s code!

What sorts of applications would you use this for?

Let me know in the comments…

Like this post? Do me a favour and share this article...

Track What Visitors Copy From Your Site in Google Analytics is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

Maybe one day it will look this good.

Update 09/01/13: Improved script to exclude MSIE 5-8 as they don’t play nice with the jQuery viewport stuff I’m using to calculate scroll depth. At least it’s possible to filter these out of your reports if you do get dodgy browsers with scroll depth >100%.

One of the most useful reports in tools like Mouseflow, ClickTale, Crazy Egg (?) and the like is the “Scroll Reach” report.

It tracks how far people scroll down a page – fancy that! As simple and useful as this metric is, it’s a forlorn metric as far as Google Analytics is concerned … but I hope I can cheer you up.

I popped together a little script which gives us a rough idea about the scroll depth of visitors as well as the height of their browser viewport in pixels. It’s free, easy to install and you can access all data from your cosy little Google Analytics account.

Win, win and win.

How it works

It’s a little different to other scroll depth trackers – such as Justin Cutroni’s excellent engagement tracking script (I highly recommend checking it out). When the page loads, the window view port size is calculated and as a visitor scrolls down the page, it calculates the lowest point that you can see on the page and stores a cookie. When they navigate to the next page, the scroll reach is transmitted to GA using Events Tracking.

Given how far down the scroll bar is in this example, the scroll reach would be 29%.

I’ve used jQuery as it appears to be compatible across browsers unlike the various plain JavaScript methods I looked at. I also used Scott Andrew’s Cookie functions from the Quirksmode website for reading/writing cookies. All data is collected in Google Analytics and is visible in the events tracking section of the website.

You will see a percentage of how far someone has scrolled on a particular page in the event tracking section.

Sure it’s not perfect. For one, it’s doesn’t take into account exit pages (just like the time on site metric) and tabbed browsing is still iffy. Luckily, it will give you an accurate enough indication of scroll reach by page (particularly for eCommerce sites or other sites with low exit rates).

Perhaps other platforms, like SnowPlow Analytics or Piwik, with heartbeat tracking could implement something to work around GA’s 500 hits per session limitation.

Requirements

• jQuery library
• Google Analytics’ Asynchronous script

Installation instructions

Did I mention how easy it is to install it? Well, it is.

Step 1: Paste this snippet after your GA script… and you’re done:

<script type="text/javascript">
// Scroll reach tracking (WTFPL licensed) by Robert Kingston - http://www.optimisationbeacon.com/
// Cookie functions from Quirks Mode by Scott Andrew - http://www.scottandrew.com/

function mkScrCookie(e,t,n){if(n){var r=new Date;r.setTime(r.getTime()+n*24*60*60*1e3);var i="; expires="+r.toGMTString()}else var i="";document.cookie=e+"="+t+i+"; path=/"}function rdScrCookie(e){var t=e+"=";var n=document.cookie.split(";");for(var r=0;r<n.length;r++){var i=n[r];while(i.charAt(0)==" ")i=i.substring(1,i.length);if(i.indexOf(t)==0)return i.substring(t.length,i.length)}return null}function rmScrCookie(e){mkScrCookie(e,"",-1)}jQuery(document).ready(function(){try{var e=parseInt(rdScrCookie("scrReach"));if(e!=undefined&&document.referrer.indexOf("//"+document.location.hostname)>-1){_gaq.push(["_trackEvent","scroll reach","viewport: "+jQuery(window).height()+"x"+jQuery(window).width(),document.referrer.toString(),e]);rmScrCookie("scrReach")}}catch(t){}var n=jQuery(document).height();var r=jQuery(window).height();var i=jQuery(window).scrollTop();var s=Math.round(r/n*100);mkScrCookie("scrReach",s);var o=/MSIE (5|6|7|8)(.*Trident)?/g;if(!o.test(navigator.userAgent)){jQuery(document).scroll(function(){n=jQuery(document).height();r=jQuery(window).height();i=jQuery(window).scrollTop();var e=i+r;var t=Math.round(e/n*100);if(t>s){s=t;mkScrCookie("scrReach",t)}})}})</script>

In an ideal world

We would be able to generate a heatmap overlay with the data. i.e. for each percentage point of the page, we could colour it based on how many visitors saw a particular part of the page. This would make my day. There’s also no telling how it works with those people who open 50 tabs on your site.

We would also be able to implement this (or something similar) into a platform with a heartbeat on the page that tracks away every 15 seconds or so, whilst the user is active. Too bad that it would chew through all of your calls to __utm.gif, faster than the cookie monster could devour a Mrs. Crockett’s factory.

Regardless, these can be developed later. It’s WTFPL licensed, so you can do what you want with it (except for Scott Andrew’s portion of the code). It’s also on Github, so…

…fork it on Github

You can follow it, fork it and add your own modifications. Take a peep through the link below, if you’d like to see the fully commented version of the script.

Visit GA Scroll Reach on Github.

Insights from scroll reach tracking

Identifying poor performing posts
My two posts about how to build conversion rate heatmaps, were very popular when they were posted but suddenly – due to the old API expiring – both methods broke and visitors became disengaged. Because of this, I noticed visitors were just not scrolling – so (along with my readers’ encouragement) I immediately fixed the issue and scroll rates have improved.

Your ideas here…

How will you use this script?

I’d love to hear some of the creative uses for this script. Any interesting ideas etc will get featured in the section above along with recognition in the form of a link.

Like this post? Do me a favour and share this article...

Track How Far Your Users Scroll in Google Analytics is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

WTH is a browser fingerprint?

Using information available from your browser configuration (through JavaScript and sometimes through Flash), we can get all sorts of interesting information, such as your user agent, fonts, browser plugins and other settings. There is so much information about your configuration that, when combined, creates a fingerprint that is unique to your browser (or in theory that’s how it works).

Panopticlick has a brilliant demonstration of this on their website.

Test your browser’s uniqueness on Panopticlick’s website.

What can it be used for?

While “privacy advocates” would scream bloody murder over browser fingerprinting, there isn’t much reason for concern. Because, what can you do with a fingerprint? Not a hell of a lot but here are a few possible uses:

• Tracking sessions that traverse multiple domains/properties/brands (where cross domain tracking is not setup or is not an ideal solution)
• Stitching sessions together where a user may have deleted their cookies or opened an incognito window

OK, so maybe the last one is a bit iffy as far as privacy concerns go, but as far as uses for browser fingerprints, there’s not a whole lot to them. Unless they’re placed in the right (or wrong, for that matter) hands, of course.

Note that when your browser upgrades or gets a new plugin, your fingerprint also changes.

How to do it

Well dear reader, look no further than Carlo’s jQuery browser fingerprinting script and Joseph Myer’s MD5 algorithm on Github. Combining these tools allows you to generate a fingerprint and develop a hash of it.

Simply chuck the hash of the fingerprint into a GA custom variable slot and boom! You’ll have a unique ID for your visitors (or round abouts).

Read on if you’d like to install it on your own site.

Requirements

• jQuery installed
• Google Analytics’ asynchronous tracking

Installation guide

I recommend that you checkout the Gits of Carlo’s Browser Fingerprinting and Joseph Myer’s MD5 hashing script. I understand you may not be willing to spend the time grabbing the scripts and combining them into a single file and whatnot. I too hate combining lots of JS files into one.

If you’d rather not whack the scripts together yourself, follow these steps:

1. Upload browser-fingerprint.js to the root of your site

2. Add the following code before you Google Analytics script but after your jQuery script

<script type="text/javascript" src="/browser-fingerprint.js"></script>

3. Modify your Google Analytics script to include the following line before _trackPageview() (updating the custom variable slot as you need):

_gaq.push(['_setCustomVar',1,'fingerprint',window.md5(_rawFp()),1]);

… your Analytics script should roughly resemble the following (along with any other customisations you have):

<script type="text/javascript">
  var _gaq = _gaq || [];
  _gaq.push(['_setAccount', 'UA-xxxxxx-1']);
  _gaq.push(['_setCustomVar',5,'fingerprint',window.md5(_rawFp()),1]);
  _gaq.push(['_trackPageview']);

  (function() {
    var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true;
    ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';
    var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);
  })();
</script>

4. Find your fingerprints in your Google Analytics’ custom variables report.

That wasn’t hard now, was it?

Notes

You can increase the uniqueness of your fingerprints by opting including fonts as provided by goodwink on Github. Your mileage may vary depending on the volume of traffic etc. Here is an example of the fingerprint that the script above will generate:

My Chome Browser
"Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.11 (KHTML, like Gecko) Chrome/23.0.1271.97 Safari/537.11###1024x1280x32###-660###true###true###Shockwave Flash::Shockwave Flash 11.5 r31::application/x-shockwave-flash~swf,application/futuresplash~spl;Chrome Remote Desktop Viewer::This plugin allows you to securely access other computers that have been shared with you. To use this plugin you must first install the Chrome Remote Desktop webapp.::application/vnd.chromium.remoting-viewer~;Native Client::::application/x-nacl~nexe;Chrome PDF Viewer::::application/pdf~pdf,application/x-google-chrome-print-preview-pdf~pdf;Adobe Acrobat::Adobe PDF Plug-In For Firefox and Netscape 10.1.1::application/pdf~pdf,application/vnd.adobe.pdfxml~pdfxml,application/vnd.adobe.x-mars~mars,application/vnd.fdf~fdf,application/vnd.adobe.xfdf~xfdf,application/vnd.adobe.xdp+xml~xdp,application/vnd.adobe.xfd+xml~xfd;Java Deployment Toolkit 6.0.260.3::NPRuntime Script Plug-in Library for Java(TM) Deploy::application/java-deployment-toolkit~;Java(TM) Platform SE 6 U26::Next Generation Java Plug-in 1.6.0_26 for Mozilla browsers::application/x-java-applet~,application/x-java-bean~,application/x-java-vm~,application/x-java-applet;version=1.1.1~,application/x-java-bean;version=1.1.1~,application/x-java-applet;version=1.1~,application/x-java-bean;version=1.1~,application/x-java-applet;version=1.2~,application/x-java-bean;version=1.2~,application/x-java-applet;version=1.1.3~,application/x-java-bean;version=1.1.3~,application/x-java-applet;version=1.1.2~,application/x-java-bean;version=1.1.2~,application/x-java-applet;version=1.3~,application/x-java-bean;version=1.3~,application/x-java-applet;version=1.2.2~,application/x-java-bean;version=1.2.2~,application/x-java-applet;version=1.2.1~,application/x-java-bean;version=1.2.1~,application/x-java-applet;version=1.3.1~,application/x-java-bean;version=1.3.1~,application/x-java-applet;version=1.4~,application/x-java-bean;version=1.4~,application/x-java-applet;version=1.4.1~,application/x-java-bean;version=1.4.1~,application/x-java-applet;version=1.4.2~,application/x-java-bean;version=1.4.2~,application/x-java-applet;version=1.5~,application/x-java-bean;version=1.5~,application/x-java-applet;version=1.6~,application/x-java-bean;version=1.6~,application/x-java-applet;jpi-version=1.6.0_26~,application/x-java-bean;jpi-version=1.6.0_26~,application/x-java-vm-npruntime~;2007 Microsoft Office system::Office Plugin for Netscape Navigator::application/x-msoffice12~*;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::application/sdp~sdp,application/x-sdp~sdp,application/x-rtsp~rtsp,rts,video/quicktime~mov,qt,mqv,video/flc~flc,fli,cel,audio/x-wav~wav,bwf,audio/wav~wav,bwf;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::audio/aiff~aiff,aif,aifc,cdda,audio/x-aiff~aiff,aif,aifc,cdda,audio/basic~au,snd,ulw,audio/mid~mid,midi,smf,kar,audio/x-midi~mid,midi,smf,kar,audio/midi~mid,midi,smf,kar,audio/vnd.qcelp~qcp;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::audio/x-gsm~gsm,audio/amr~AMR,audio/aac~aac,adts,audio/x-aac~aac,adts,audio/x-caf~caf,audio/ac3~ac3,audio/x-ac3~ac3,video/x-mpeg~mpeg,mpg,m1s,m1v,m1a,m75,m15,mp2,mpm,mpv,mpa;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::video/mpeg~mpeg,mpg,m1s,m1v,m1a,m75,m15,mp2,mpm,mpv,mpa,audio/mpeg~mpeg,mpg,m1s,m1a,mp2,mpm,mpa,m2a,audio/x-mpeg~mpeg,mpg,m1s,m1a,mp2,mpm,mpa,m2a,video/3gpp~3gp,3gpp;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::audio/3gpp~3gp,3gpp,video/3gpp2~3g2,3gp2,audio/3gpp2~3g2,3gp2,video/sd-video~sdv,application/x-mpeg~amc,video/mp4~mp4,audio/mp4~mp4,audio/x-m4a~m4a,audio/x-m4p~m4p,audio/x-m4b~m4b;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::video/x-m4v~m4v,image/x-macpaint~pntg,pnt,mac,image/pict~pict,pic,pct,image/x-pict~pict,pic,pct,image/x-quicktime~qtif,qti,image/x-sgi~sgi,rgb,image/x-targa~targa,tga,image/jp2~jp2;QuickTime Plug-in 7.7::The QuickTime Plugin allows you to view a wide variety of multimedia content in Web pages. For more information, visit the QuickTime Web site.::image/jpeg2000~jp2,image/jpeg2000-image~jp2,image/x-jpeg2000-image~jp2;Microsoft Office 2010::Office Authorization plug-in for NPAPI browsers::application/x-msoffice14~*;Microsoft Office 2010::The plug-in allows you to open and edit files using Microsoft Office applications::application/x-sharepoint~;Google Update::Google Update::application/x-vnd.google.update3webcontrol.3~,application/x-vnd.google.oneclickctrl.9~;Shockwave for Director::Adobe Shockwave for Director Netscape plug-in, version 11.6.1.629::application/x-director~dir,dxr,dcr;Silverlight Plug-In::4.1.10329.0::application/x-silverlight~scr,application/x-silverlight-2~"

My Firefox Browser
"Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:20.0) Gecko/20121201 Firefox/20.0###1024x1280x24###-660###true###true###Microsoft Office 2010::Office Authorization plug-in for NPAPI browsers::application/x-msoffice14~*"

If you’re relatively technical, be sure to checkout SnowPlow Analytics for a full featured open source analytics platform.

Is this method of tracking overly creepy?

Let me know in the comments below.

Like this post? Do me a favour and share this article...

Tracking Browser Fingerprints in Google Analytics or Other Tools is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

SEO tools are great for competitive benchmarking and general analysis of backlink profiles: Open Site Explorer, Google Webmaster Tools, Bing Webmaster Tools, ahrefs and Majestic SEO. But they only offer a snapshot of where your links are coming from.

Mining your web server logs can give you a pretty complete list of backlinks too, as some crawlers and most visitors will report the referrer to your HTTP server – even if they aren’t running JavaScript. But this is practically impossible for most people to do, so best to steer clear of it.

Google Analytics can also show you a fairly comprehensive list of backlinks along with visitor metrics. If you subscribe to the view that well trafficked links generally have higher value, then visits and user engagement are good metrics to help judge link value. Sure it won’t tell you how many of those links still exist, what anchor text they use or how many are rel=”nofollow” but that’s why you use tools like Niels Bosma’s SEO Tools for Excel once you have the data.

Best of all, the GA method, takes you no time to pull the reports you need. “How long, Rob?” you ask. Just 5 minutes or about “…………….” yay long.

The problem with Google Analytics

Typically SEOs shy away from using backlink data out of Google Analytics. Most SEOs may know about Analytics’ ability to show referrers, but they don’t use it because out of the box, the data sucks.

You simply can’t get a full URL out of the “Full Referrer” dimension (of all places)… Just look at this:

OK – Look away, dear reader. Let’s look at how we can fix this up.

Don’t worry, you can apply a simple, yet crafty, filter to fix this in a couple of minutes.

Step 1. Setup the Full Referrer Profile Filter

Here’s how:

1. Create a new profile or use an existing one if you have the bollocks (remember, apply profile filters permanently alter your reports data, there’s no “undo button” if you mess up your main profile).

2. When you’re in the profile, follow these links, “Admin” in the top right > “Filters” in the profile tab > “+ New Filter”

3. Create a filter , with the settings I’ve used in my example below:

This profile filter will add full referral URLs into your user defined value in Analytics.

Optionally, you may want to create filters to remove certain parameters from the Full Referrer such as utm_source and other popular ones.

Step 2: Loading up the Custom Report in Analytics

Simply open this custom report with a decently large date range, including the date range from when you setup the filter. And ensure you’re looking at the profile for which you setup your filter.

Download this report.

You may need to customise the report a little to exclude some nonsensical links, but once it’s setup, it’s quick and easy to use.

Step 3: Export & Analyse the Data in Excel with Niels Bosma’s SEO Tools

First of all, if your report returns more than 500 rows, use Sumo Baby’s handy boomarklet which makes it possible to export up to 10,000 rows at a time.

Then, hop on over to Niels Bosma’s site for analysing all this data in his handy SEO Tools plugin for Excel.

This amazing tool will make the data much more useful for SEO.

Comparison of using GA to other tools

Unfortunately I haven’t got a lot of retroactive data with this filter which I applied for my own site (it’s a bit of a case of the gardener’s garden), so I have reverted to using the inferior full referrer dimension. Even then, it demonstrates that Analytics is capable of finding a healthy number of links.

Google Analytics doesn’t show the highest number of backlinks, but it does show a reasonable number of them alongside useful metrics.

Google Analytics (without full referrer hack above): 501

ahrefs: 372

Open Site Explorer: 283

Majestic SEO: 1698

Google Webmaster Tools: 1211

Bing Webmaster Tools: 160

Server logs: Did not measure – I should have, could have, would have but I didn’t.

Pro tip #1: Track a rolling total of links or new backlinks created Month-on-Month

Google Analytics V5 allows you to bookmark and share URLs to access unique reports. If you want a rolling count of the number of backlinks, simply change the date range in the URL (formatted as YYYYMMDD) to a far off future date and bookmark the URL.

Simply change the date in the URL of this custom report to something well into the future. In this example, I have it set to 31 August, 2019.

Likewise tracking new backlinks each month can be done very easily by the following formula:

Pro tip #2: Use Google Analytics’ Built-in Social Activities Report

Another option you can take, as detailed by Daniel Waisberg is to use Google Analytics’ own social backlinks report which combines shares and links in the same report.

Simply pop into Google Analytics and follow these prompts: Traffic Sources > Social > Pages > Activity Stream > Events

From here, you will see a range of backlinks which Google Analytics characterises by how it was shared:

Happy backlink checking, folks!

Like this post? Do me a favour and share this article...

Find Backlinks to Your Site With Google Analytics is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

Read my original post on how to create a heatmaps of analytics data here.

Thanks to a smart Finn called Mikael Thuneberg, I’ve managed to whip up a quick and dirty little Google Docs template which builds these Google Analytics heatmaps in seconds (much easier and faster than the old method):

Hour and day of week heatmap of conversion rates using Google Analytics data

How it works

Basically Mikael Thuneberg developed some JavaScripts that pull GA data from the GA API into Google Docs. All the code is readily available for you to inspect here and inside my template.

How to do it

IMPORTANT: Google disabled the API, so please read my old post on the manual heatmap method (it’s still reasonably quick .

Watch the instructional video below (I’m a YouTube newbie, so you may have to bear with it) or follow the steps outlined afterwards.

Step 1: Load the template from Google Docs and save a copy to your Google Account (ensure only you have access to this as your password will be stored in the spreadsheet).

Pointing out all the little bits and pieces in the spreadsheet.

Step 2: Pop in your username and password, into the fields on the “Data” spreadsheet tab. Feel free to hide your password in the spreadsheet.

Step 3: Grab your Profile ID (Note that this is not your UA number!) from the URL in your Google Analytics interface and pop it into the corresponding cell in the template. You’re looking for the number in the URL which looks like this:

https://www.google.com/analytics/web/?pli=1#report/visitors-overview/a5147757w23349371p21652238/

Therefore, you would use “21652238” in the spreadsheet.

Step 4: Set your date range – you’ll want a decent date range so that you have a couple of hundred thousand visits at least.

Step 5: Pop the ID of an advanced segment into the spreadsheet. Some are already provided under the inputs area. Otherwise “-9″ will work for most sites with goals setup.

Step 6: Jump into the reporting tab at the bottom of the screen.

Step 7: Copy the tables you’d like to use and pop them into Excel, and dress them up in a table with conditional formatting.

Either download my example conversion rate heatmap template for Excel or use conditional formatting and table formatting icons up in the ribbon (note you’ll need Excel 2007 for this):

Done!

Whereas the old process would have taken 30-60 mins to aggregate in the GA interface, this less than 5 minutes – and much faster with several cups of coffee.

Like this post? Do me a favour and share this article...

Build a Conversion Rate Heatmap by Hour & Day of Week in Google Docs is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

What is Ethnio?

It’s a service that uses your website to recruit users for research purposes. Decide which visitors you want to target, prompt your users to participate in research for an incentive, screen out the ones you don’t want… and boom. You’ll be recruiting quality participants in next to no time.

Following is how I used it recently on Optimisation Beacon to conduct interviews. Read on for how I used Ethnio (and its features) in excruciating detail …

1. The Opportunity & Objectives

I knew very little about my website visitors and was concerned that whatever I was going to write about was going to alienate some of my readers. I had little understanding about what kept bringing loyal visitors back to my site (outside of top landing pages for return visitors) and I wanted to learn what new visitors thought of my site. On top of that I really wanted to know what people found useful, struggles they face in their jobs and what would encourage them to subscribe and share my content.

2. The Methodology

With little knowledge of my visitors, I opted for exploratory research involving simple unstructured interviews with my website visitors. Sure, I could try to go out and contact subscribers, or write a blog post about it, but I would only reach already-loyal subscribers, so I chose to recruit via my website.

My budget for the study was a mere shoestring (90% the costs I incurred went straight to the respondents and the rest were overseas phone call charges – note to self, “Do not call Israel.“) so choosing Ethnio’s free package was a no-brainer for me.

3. Setup

The simplicity of Ethnio cannot be understated – it was dead simple. Right from the get go, I was able to create a custom screener that would allow me to phone up and interview loyal readers and new visitors. Here’s how easy it was:

Choose the design of your screener modal dialoge

When you create a new screener, you can jump in and select the screener type. Choose from pop-over modal dialogues to a little message that sits in the bottom corner of the users’ screens.

Select the screener design you want Ethnio to use.

Specify where you’re recruiting from and for how many respondents

Simply pop the URL in here that you’d like to preview your screener on and let Ethnio know how many respondents you need for your study.

Specify where you're recruiting participants from and how many you need.

Add your company logo

Customise your screener with your company’s logo, so visitors know the research is being conducted for your site.

Add your logo to the screener form.

Position the screener on your site so it pops up where you want it to

Ensure the screener pops up where you want it to and make sure it’s not covering anything up, that you don’t want it to.

Position the Ethnio screener to sit on the page.

Customise your screener text and set the incentive

Here’s your chance to set the incentive and customise your screener’s text. This part is particularly important. You have two options here: offer a universally valuable incentive (e.g. cash) or give away something cheaper albeit with niche appeal (e.g. download an eBook). I think it’s important to give an incentive of universal appeal, as this will ensure your study appeals to everyone whilst your screener weeds out the people you don’t want.

Set the text and incentive to be shown in the screener dialogue.

Create the screener questions

Perhaps I asked a few too many questions here (you can’t blame me for being curious, can you?). This allowed me to learn more about my respondents and be picky with who I chose to interview.

Provide questions to screen out users you don't want to have participating in your research.

Customise your thank you message

A little thank you page to show visitors once they’ve completed your screener.

Customise the text at the end of your screener.

Optional: Automatically recruit users and pass them to your online study

By default, Ethnio will recruit people and wait for you to screen them in, when you login to the interface. Branching logic allows you to automatically screen in respondents based on their answers.

I didn’t use the branching logic in this project, but if you’re recruiting for usability testing or online surveys for example, you can link screened respondents directly into the study, so you get your data as fast as possible.

Publish your screener via your site, Twitter or a public link

Once you’re done, Ethnio allows you to share your screener with a link provided or setup the screener to show on your site. I can see this being popular with any businesses who have massive social media presences.

You can share the screener through your site or tweet it out to your social media profiles.

4. Data collection

After just over a month, and a poor response rate to my $40 Amazon gift voucher incentive, I had a number of candidates which I could choose to interview over the phone. Obviously in the below instance, I hid the responses, so you can’t see my visitors’ phone numbers…

Here you can collect and manage your respondents in an easy to use interface.

The interviews

All I had to do was setup a time in which the interviewee and I could sit down for a quick chat and call them up. In this regard I was a little under-prepared. I basically called up with a rough set of questions and let the respondents tell me a bit about themselves. The calls were recorded for my future reference, all the while I took notes.

5. What I learned

The main points I learned from the interviews were as follows:

• Visitors to Optimisation Beacon sit roughly within one of two segments: Developers or Analysts
• Whilst analysts may not necessarily have the technical know-how, developers struggle with the marketing/business application know-how (though some have the best of both worlds)
• Analysts are drawn more toward posts that help them do their jobs and get more insight (i.e. How to maximise the insight they get from GA, calculate split test significance, and posts like this one)
• Developers are generally more interested in the codes and how to posts that will enable them to affect business outcomes (i.e. Track GWO tests in Google Analytics, setup content targeting with custom variables etc…)
• The tools I have developed encourage visitors to come back to the site on a regular basis (particularly the split test calculator)
• Developers, whilst being more interested in code are also interested in learning about website optimisation, but more introductory-level posts whereas analysts are looking for more advanced posts

6. Actions

Obviously, it’s great to learn lots of interesting stuff, but it’s even better if you can do something about it. Here’s what I’m working on now and for the future for Optimisation Beacon:

• Post meatier, authoritative articles/resources that help (and engage) analysts and web developers do their jobs
• Develop more conversion and maths related tools
• Post more consistently (OK, perhaps I failed a bit there) and develop my instinct on what my readers will respond to best

The Verdict: 9/10

Pros

• It’s freemium! And because of this, I’m now a paying user at work.
• Setup is lightning fast and beyond easy.
• They’ve covered just about everything with the service, including incentives, full customisation of the text and screener questions.
• The interface is sleek and quirky, without getting in your way of setting up your screener or recruiting respondents.
• Use Ethnio to pay incentives to users (for a 15% cut)

Cons

• The screeners’ branching logic is a little inflexible as you can’t send different respondents into different studies based on their answers.
• If you go premium, the cost is a little on the high side. If it were slightly lower, I could see this tool as being one of our permanent tools. Perhaps in the future they may provide discounted annual accounts or agency deals.
• Doesn’t show you abandonment metrics for the screener (say if you ask some really personal stuff – occupation, name, email, phone etc – you want to know if it’s affecting the response rate)

OK, so you know I love this tool but should you use it? Well, if you’ve never had personal contact with your customers (or your clients’ customers) and want to understand them, I highly recommend you check out Ethnio to recruit for interviews. If you’re looking to recruit for another remote study (e.g. usability testing), I can’t quite say, as I haven’t used it for that purpose yet.

Let me know your thoughts about Ethnio in the comments…

Like this post? Do me a favour and share this article...

Ethnio Review: How I Used it to Interview My Visitors is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.

To date, I have built a calculator for graphing and measuring statistical confidence on your tests (alongside the feedback tool and GA content targeting script). It’s the first of its kind to graph split test results freely on a website (at least that I know of).

Check out the split test calculator here – share it and bookmark it if you find it useful.

About the calculator

An example of the graph produced by the tool.

It’s built with Jquery and the Google Image Charts API to calculate statistically significant changes in conversion rates and even bounce rates. It’s approach is fairly similar to Google Website Optimizer’s so it should yield similar results.

Use it to calculate:

• Landing page testing
• Email marketing split tests
• PPC ad testing
• Calculating changes in bounce rates in your experiments
• Split testing scripts in a call centre
• … and practically any other binomial data

After days of “hard yakka coding”, I really hope you find it useful. As always, tell me what you think of it through the feedback form in the corner. Enjoy!

Like this post? Do me a favour and share this article...

New: Graphing Split Test Calculator for A/B Testing is a post from Robert Kingston who blogs about Conversion Optimisation and Analytics Tips.