At Google's annual I/O developer conference, an important new feature to Google Analytics was announced that's going to turn mobile marketing on its head.

Built on the back of the new Universal Analytics platform and its baked-in ability to track users across sessions and devices, Cross Device Measurement is like Multi-Channel Funnels for devices.

And it's huge....

The new features were unveiled by Nick Mihailovski, Pete Frisella and Andrew Wales in a session titled Optimize Web and Mobile Apps, Across Devices, Using Google Analytics.

You can watch the whole presentation here, and I recommend you do: it's forty minutes well-spent, and covers some additional stuff that isn't covered in this post. (It's embedded at the foot of this post). 

What is covered in this post is, in my opinion, the most significant part of the session as far as digital marketing is concerned. Let's get stuck in. 

Cross Device Measurement with Client ID 

Cross-Device Measurement is related to one of the main ways we think Universal Analytics will change the way businesses look at their data. It's all about how Universal Analytics gives us a better understanding of how visitors use multiple devices to access our content.

To illustrate this with some screenshots from the presentation, take a look at how "regular" Google Analytics will track three visits from three devices as three Unique Visitors, even if it's the same person using all three:


You can see the problem: each visit on each device initiaties a Client ID (cid) which tells Google Analytics who that visitor is. With three visits from three devices, we've got three Client IDs which, as far as Google Analytics is concerned, is the same as three Unique Visitors. 

Clearly, this isn't optimal! 

Luckily, Universal Analytics gives us a way of understanding when three visits is one person, or three people: 

With Universal Analytics, we get a new parameter - uid, or User ID. This lets us tell Google Analytics that it's the same person using mobile, tablet and laptop.

We're no longer seeing this as three Unique Visitors, but as one Unique Visitor -- so the distortion has been removed. 

By itself, this is a huge change. We all know, from our own behaviour as well as that of our customers, that people are using more devices and switching from one to another with increasing frequency.

From a digital analytics perspective, having a tool that can't tell the difference between three people and one person visiting three times is completely insufficient, and if your tools aren't up to the job your analysis won't be, either. 

See, this is why we all need Universal Analytics! 

Device overlap

To continue with the example, let's say that someone accesses our website on a tablet, and then they switch to their laptop computer to complete a purchase. 

With User ID, we know that only one person was behind us, but we're still not quite getting the complete picture. 

Here's another slide from the session showing how it might look:

On the face of it, we've got one tablet visitor with $0.00 revenue, and one desktop visitor with $3.99 revenue. That makes tablet traffic look pretty suboptimal, and you might decide to ditch tablet users and focus on desktop users instead.

Once again, a suboptimal tool isn't going to get you anywhere in life, let alone in actionable data analytics. So, how can Universal Analytics help us?

Cross-device Measurement

In a word: Cross Device Measurement. Okay, that's three words but once you see what it does, you'll forget my inability to count. Here's how Cross Device Measurement would give us a more accurate view of the same situation:  

Bam. Universal Analytics is giving us a picture of multiple-devices and how they're used by our audience to collectively contribute to our overall conversion volume and our revenue. 

If you're thinking that this looks and feels a lot like Assisted Conversions, you're exactly right. Just as the idea of Multi-Channel Funnels gave us an understanding of how certain marketing channels co-conspire to drive conversion, Cross Device Measurement tells us how our users hop from device to device, and tell us where we should be paying most attention.

But wait. It gets better. 

Device paths

Take a look at this: 


Universal Analytics is going to give us device-based conversion paths! Again, analogous to cross-channel conversion paths, this is really going to blast open our understanding of how users are interacting with our content. (Apologies for the fuzzy image by the way, it's the best I was able to get.)

The session then homed-in on a couple of angles as to how this can give us insight.

First up, we might discover that a certain type of screen-to-screen behaviour tends to elicit a lot of conversions: 


In this example, users that switch from desktop to tablet spend a lot more on-site than users that complete the reverse journey, or that stay on desktop all the way through. After patting ourselves on the back for having such a great tablet-optimised site, we might decide to put something on the homepage to tell desktop users to try us on their tablets. 

Which, to take a step back for a second, is pure digital analytics. Observe a behaviour, look for ways to encourage it. Perfect!

Secondly, how about this: 

It's reminiscent of Assisted Conversions again, and it tells us some pretty important information. On the face of it, with only $0.99 in revenue we might decide that mobile isn't a terribly important platform for us.

However, when we bring in transaction data from other platforms, we can see that the user that accessed our content on a mobile phone made purchases totalling $17.98 on other devices. So while smartphones might not be great as a final touchpoint, we're getting an understanding of the importance of the supporting role they play in the process. 

And if that's not made you dizzy, think about where this could go. We already use Multi-Channel Conversions in our reporting, when we have Cross-Device Conversions to play with too, the insight is going to increase exponentially. 

We'll probably see ways of being able to splice the two sets of data here and there; so we might see that a certain campaign played a pivotal role in early-stage awareness for users on smartphones, but that a completely different campaign did the same on tablets. 

As I said before: bam. There's no timeline as to when we can expect this to be available publicly, but still: bam!

Turning mobile on its head

Digital marketing continues to become more and more data driven and with additional features being made widely available - in a free product, let's not forget - the quality and depth of that data is always increasing. 

So, are you as excited about this new feature as I am? How do you envisage this new dimension of reporting being useful to your organisation? 

Damion Brown

Published 20 May, 2013 by Damion Brown

Damion Brown is Trainer & Consultant at Data Runs Deep and a guest blogger on Econsultancy. You can connect with him on Google+ and on LinkedIn

2 more posts from this author

You might be interested in

Comments (13)

Save or Cancel


This is interesting stuff but Google itself says it's for 'signed-in experiences'. (Fancy signing in to every website you view? No thanks.) Which means it's not going to be usable in this format for a large chunk of sites out there.

Theoretically Google would only need users to sign into a single site once on each of device for it to recognise that those devices are associated with that user (in fact they probably know this already for a lot of folk given that so many people sign in to Google).

But I am guessing that there is a legal or strategic reason for not just having this as a generic web-wide 'this is Paula Jones' cookie which is loaded onto the device and can be read by any GA account.

It would be handy though - for sites which would otherwise require big changes in order to get reasonable number of users signing in.

about 5 years ago

Gemma Holloway

Gemma Holloway, Digital Marketing Executive at Koozai

Excellent post Damion! And to answer your question, Yes I am as excited as you.

Universal Analytics is going to revolutionise the way we view data. This post covers so many exciting new things we can now determine with the use of and that's before even taking the offline tracking into account.

I appreciate the use of screen shots too (even if they were from the original presentation), they really helped your explanation of each point.

about 5 years ago


Jim Banks

To pick up on Kirsten's point. The "signed in" number is now as high as 75% for some sites and with the continuing signing up of people to Google+ that number is only going to get higher as each month goes by.

Utopia has always been the opportunity to close-loop transactions and this moves us closer to that. I still think there is some work and a ton of education/adoption, but it's exciting.

about 5 years ago


Peter O'Neill

I am with Kirsten - this feature won't have that much impact for the vast majority of websites. Yes, for websites where the majority of visitors login, this is going to shift to Visitor analytics and those sort of multi touchpoint reports become very interesting.

But most ecommerce websites only have login from the Checkout process (< 5% visits), most media websites are not behind a paywall, B2B websites have very little signin options. For these websites, cross device measurement will still not be possible.

The features look cool - and it is fair to say this will have huge ramifications - but with a big caveat of only for a small proportion of companies.

about 5 years ago

Damion Brown

Damion Brown, Trainer & Consultant at Data Runs Deep

@Kirsten -- you're quite correct, it does rely very heavily on uses being logged-in. Don't forget though that if you're accessing content on an android device, you're effectively logged-in to your Google account all the time, and the same is probably true if you're signed-in to Chrome on a desktop device.

Regarding web-side sign-in, you're on the money:

Potentially, by having a prevailing Google+ sign-in as we browse the web, the user ID will become a defacto.

@Gemma -- thanks! And yes, I agree that the offline tracking is where things start to get really interesting :)

@Peter -- good point. I think we'll definitely need to move away from signing on to every site, as people just won't do it.

Have a look at Cross-Platform Single Sign-On which was (coincidentally?) also announced at Google I/O last week:

If there's an infrastructure shift to get everyone signed-in all the time, it's going to remove a lot of the barriers to cross-device tracking working in a practical sense...

about 5 years ago


Tom Belgeonne

Hi Damion

Great article - bring on an engaging and relevant 2013 for consumers.

The Offline tracking is superb we're already able to help businesses and brands understand Traditional Media's contribution to online sales far more accurately! Epic!.. And perhaps help the traditional ways hold on longer with heightened accreditation.


about 5 years ago

Damion Brown

Damion Brown, Trainer & Consultant at Data Runs Deep

@Tom -- Thanks! And that's an interesting point about how digital insight might help traditional media's longevity.

Digital, the saviour of offline... Who'dathunkit?!

about 5 years ago


Peter O'Neill

@Damion I had joked with ppl on twitter that this would all be easy if we only could use the Google, Twitter or Facebook cookie as the visitor ID. But we assumed that would never happen due to privacy concerns.

Using Google login on more websites is only a partial step there and if people get used to authorising most websites to use Google for access, then everything does change. It could become a habit to click on a little button whatever website you are on. Still have privacy concerns, be interesting to see ToS on what data Google is allowed to collect from these logins.

about 5 years ago

Anna Lewis

Anna Lewis, Google Analytics Analyst at Koozai

I'm really looking forward to this data, I just hope enough websites have the functionality/technology available to get the most from it otherwise it's going to be a shame to know this data could be collected and used and isn't able to be.

about 5 years ago

Carmen Mardiros

Carmen Mardiros, Digital Analytics Consultant at Clear Clues

Peter and Kirsten are right, this does benefit only a small number of businesses. But that is precisely why I am so excited about it.

Look at innovative, technology driven startups. More and more put sign in at the core of their business model (particularly SaaS companies). For them it's not a case of coming up with incentives to get users to log in -- they architect the entire experience in a way that sign in is baked into the entire product experience and value proposition. For users signing in feels like a natural expectation.

Think it doesn't apply beyond SaaS?

Look at Quora. Huge growth and content consumption when they make to log in if you want to read more than 1 answer. Look at Deal of the day sites which require you to register & and log in to access current deals. There are many more examples.

Yes, you are still limited in analysing the first part of the acquisition funnel where people make multiple visits (cross device) as prospects before they sign up (and most never do) but there are ways to get more insight there. Look at the Freemium model, the barrier to conversion is lowered so much that the path to trial/purchase (the bit that's most muddled by multi device browsing) is shortened considerably (reportedly Evernote free users sign up straight away but take an average 129 to convert to paid!)

I'm not saying that these techniques will work across the board. What I'm saying is that with the emergence of tools like Universal Analytics comes the need to innovate traditional (dare I say tired?) business models.

Google know that and they're investing considerable resources in building technology able to power those innovative business models. And perhaps it's my focus on startups that makes me so excited about how Universal Analytics will help grow small, agile and disruptive businesses. They're the giants of tomorrow.

about 5 years ago



Have i missed something but isnt this what IBM have been doing with IBM web analytics (Coremetrics) for over a year already? i.e. de-duplicating users and being able to provide unique visitor counts across domain / device ? But with the added ability to then re-target these users through cheaper channels than adwords ?

about 5 years ago



I have tried to link same user using CID. That is online session to offline conversion. But GUA isn't linking user session properly even though I used same CID. I guess this feature isn't available yet. You could find the my test here -

about 5 years ago



now linking user using CID is working properly

about 5 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.