Tracking Mobile Visitors in Google Analytics: A Checklist Guide for Mobile Insights

It is hardly surprising that “mobile” is becoming such a hot topic among digital marketers: comScore reported that in August 2012, 13% of total internet pageviews (roughly 1 in 8!) were from mobile devices (phones and tablets combined). And these numbers are increasing. Econsultancy’s Christmas Online Shopping Survey reported that 24% of shoppers used mobile devices for Christmas shopping in 2012. We are even starting to see a preference for mobile usage over desktops. A study by Nielsen and Google published last month found that 77% of mobile searches are performed in a location (home, work, etc.) where a PC is likely to have also been available.

This is good news for online marketers, as mobile devices provide increased opportunities (due to convenience and portability) for visits and conversions. But mobile users are also easily lost: upon arriving at a site which is not mobile-friendly, 61% of customers are likely to go to a competitor’s site. So what should we be doing to make sure that we’re ahead of the curve? And how can we be sure that our site is as useful for our mobile visitors as it is for those on a desktop? One way to gain insight into these questions is with Google Analytics (GA) data.

Part I: Overview

There are three types of web property for which you can track mobile traffic in Google Analytics:

  • a single URL site, which might be:
    • a non-mobile desktop site
    • a responsive/dynamically serving mobile-friendly site
  • a separate mobile site on a different URL (e.g.
  • a mobile app
For a single URL site and a separate mobile site, the important metrics are usually the same:
  • visitors
  • keywords
  • bounce rate
  • traffic sources
  • landing pages
  • ‘mobile’ report
These metrics enable us to answer major questions about our site and our mobile strategy, including:
  • Who are our mobile visitors? (mobile report)
  • What are mobile users looking for? (keywords)
  • Are they having trouble finding it? (landing page bounce rates, average time on page, page depth)
  • Are certain campaigns/pages better for mobile users? (mobile vs desktop conversions)
  • Is our site structure and information architecture mobile-friendly? (site search, goal flow, page depth, average time on page, conversions, site speed)
  • Should we optimise for separate mobile keywords? (keywords report - with device or screen resolution as a secondary dimension)
  • Does my site load fast enough to prevent users from bouncing? (site speed)
  • Is my site navigation easy to use on a mobile device? (landing page bounce rate)
  • Are we using the best mobile-friendly approach (i.e., responsive design vs separate URL)? (all of the above)
  • Are users behaving differently on mobile devices than on PCs? (all of the above)
  • Are my calls-to-action - content and/or placement - still effective when viewed on a mobile? (mobile vs desktop conversions, responsive click tracking)
  • Should I have an app? (conversion goals; keywords)
    • NB: apps should always be used in addition to a mobile-friendly site, never instead of one.
Some additional questions we should think about when using data from, or considering implementation of, a separate mobile site:
  • Are my mobile users searching differently and/or for different content? (keywords - organic search and site search reports)
  • Are my device-based redirects functioning correctly? (cross domain tracking compared with the device report)
Google has recently introduced (in beta) a feature for tracking a mobile app. The ‘Mobile App Tracking’ feature uses slightly different terminology, but the concepts are similar: 
  • Acquisitions = Traffic report
  • Users = Visitors report
  • Engagement = Visitors/Traffic reports
  • Outcomes = Conversions report
Questions we should be asking about our mobile app:
  • Are mobile users finding our app in the Google Play store? (Google Play sources)
  • How are they finding it (sources)?
  • Are they abandoning it after install or after opening? (users)
  • Is it crashing frequently? (engagement)
  • Is it performing better on certain devices? (users)
  • Are they converting better or worse than average/expected? (goals, transactions)
  • Is there a problem with the conversion funnel? (goals, transactions)
So how do we use these metrics/reports to find the answers to our questions?

Part II: Visual Guide

Desktop/Single URL/Responsive Tracking

The “Mobile” report on its own is somewhat limited. For example, you can’t view landing pages and also segment out just the mobile traffic (at least not with the default report). So instead, I often use advanced segments in the other reports to limit my data to mobile traffic.

The problem with the default ‘Mobile Traffic’ segment is that it includes tablet traffic as well. There is a default segment for ‘Tablet Traffic’ only, but in order to view mobile traffic without tablets you’ll need to create a custom segment.

The way we’ve done this for the Distilled analytics account is based on screen resolution and uses Regular Expressions (RegEx):
  • Name: ‘Mobile - no tablets’
  • Include: ‘Mobile (Including Tablet)’ containing ‘Yes’ AND
  • Exclude: ‘Screen Resolution’ Matching RegExp (1\d|[7-9])\d\d+x.*

What this RegEx means is that this custom segment should include traffic from mobile devices but exclude traffic from devices with a screen resolution of 700+ by anything. You may decide to tweak the RegEx depending on how large (or small) a device you want to include. (Some of the larger smartphones also fall in this range, but then again maybe these should be treated as tablets.)

Once you’ve got a sense for your mobile traffic, you’ll want to compare mobile and desktop traffic using multiple advanced segments. In addition to the non-tablet mobile custom segment, you may also need to create a ‘non-mobile’ segment: in this case, use ‘Include: Mobile; Containing: No’.

When we apply all three segments, we can see a graph which compares the difference between mobile, tablet, and desktop traffic:

This allows for a quick overview comparison of high-level metrics such as overall number of visits.

Using Analytics to Define a Mobile-Friendly Approach
If you’re just starting out to create a mobile-friendly website, you can use your regular site’s data to figure out the best approach (e.g. responsive, adaptive, or a separate mobile site).

Important metrics:

  • Visitors: what percentage is from mobile devices?

In the example above, only 5% of visits are from mobile, non-tablet devices. The proportion of mobile visitors can help in deciding how important a mobile-friendly strategy is to your overall online strategy. Even if this percentage is currently low, however, the numbers are increasing all the time, so whether it’s low priority or high priority, it should still be a priority. A low percentage of mobile visitors means you can afford to implement your mobile approach more gradually, with just the top pages, and build out as needed.

  • Keywords: is there a noticeable difference in search terms?

Organic keywords are a primary dimension in the Traffic Sources > Sources > Search > Organic report. A quick glance at top 10 mobile keywords compared with the top 10 non-mobile keywords should tell you whether people are finding your site using different keywords on mobile devices. If the mobile keywords are drastically different, you may want to consider a separate-site approach, which will allow you more freedom in optimising for mobile-friendly keywords. A middle-ground option is to serve different HTML on the same URL based on device, which allows you to have different meta descriptions and page titles when someone searches on a mobile device. If there is no real difference, however, you may choose to use a responsive design and simply re-arrange page elements to accommodate different screen sizes.

  • Bounce rate: is there a big difference between desktop and mobile visitors?

This data is in the Traffic Sources > Overview report:

If you see a significantly higher bounce rate for mobile devices compared to PC visitors, you need to make sure that your mobile visitors can access the content they’re looking for, and that they don’t have to pinch and zoom to do so. It may be, for instance, that your site navigation isn’t touch-friendly and so mobile visitors are stuck on the homepage/landing page.

  • Landing pages: which pages have a disproportionately high (or low) bounce rate for mobile devices? What about conversions?

The Landing Pages report is under Content > Site Content > Landing Pages. You can compare mobile and non-mobile bounce rates with those advanced segments. I also like to use the red-and-green ‘Comparison’ view for checking bounce rate because it’s a quick visual check of how each page is doing compared to the site average:

Remember that if your landing page is a blog post, as in the above example, a high bounce rate is normal. What matters here is not necessarily the actual metric, but the difference between mobile bounce rate and desktop bounce rate (although of course if you see an unusually high bounce rate across devices, that page may need to be investigated further).

To check which landing pages have high conversion rates, use the Reverse Goal Path report (Conversions > Goals > Reverse Goal Path).

  • Traffic Sources: is there a difference between desktop and mobile traffic sources?

This report is found under Traffic Sources > Sources > All Traffic. You will need to change the primary dimension to ‘Medium’. ‘Medium’ is the term for the highest level traffic types: direct traffic (here classified as “(none)”), organic and paid search, referral traffic, and any campaigns you’ve set up. This way (with your two custom advanced segments) you can see any discrepancies between mobile and desktop for the different traffic sources, and find out if you should be focusing on a particular traffic source with your mobile strategy:

This is an interesting example, because recently there has been an issue with some mobile referrer data. The iOS6 (iPhone/iPod operating system) uses Google’s encrypted search by default and due to the way Google delivers its mobile SERPs this search traffic is recorded as direct traffic. Android 4 is also losing some referrer data. Some of the discrepancy here (showing direct traffic as 40% mobile and 16% desktop, vs referral traffic 17% vs 23%) could be due to this. All is not lost, however: Annie Cushing has created this  custom report to help find your iOS6 traffic.

  • Site Speed: does your site take longer to load on mobile devices?

The Site Speed report is found under Content > Site Speed. The first thing to check is a mobile vs non-mobile comparison of average page load time.

We can’t take this metric too seriously, however, because it’s so easily skewed by a single instance of long load time. So if the load time looks unusually long, dig deeper into the particular page(s) having trouble.

If you do have a long load problem, consider whether large and difficult-to-load content (such as large images) is really adding anything to your site. If so, there are ways to minimise load time while maintaining the desired look for your site. Compress any images on the page as much as possible without lowering the quality, delete any unnecessary HTML or CSS, and compress those as well. If the mobile version of your site is responsive (or if you don’t have a mobile version at all), this will also clean up your desktop version. If you can’t compress enough elements to make a mobile page load in under 5 seconds, though, you may need a separate mobile site. Try testing your site(s) with the Google PageSpeed Insights tool. This tool provides information and recommendations for cutting down load times.
  • Goal Flow: Are particular steps in the conversion funnel causing a major roadblock for mobile users? (Conversions > Goals > Goal Flow)

  • What about tablets? All of these metrics can also be used with the ‘Tablet Traffic’ advanced segment, in order to decide whether you need a separate tablet strategy. General best practice is to display the desktop version of the site to tablets, but this may be changing as screen sizes become more varied.
For more ideas on useful data for choosing a mobile-friendly approach, Aleyda Solis has written a post on how to perform a mobile site audit.

Some Extra Data for Tracking Responsive/Single URL Performance

Responsive Click Tracking

If you already have a responsively designed site, you may want to check that your call-to-action buttons are visible and effective when the page elements are rearranged to fit a smaller screen. Henry Zeitler has explained a way to track mobile clicks on a responsive site, which he terms “responsive click tracking”.

Setting Up a Custom Dashboard for Mobile Data

We can also create a custom dashboard exclusively for mobile data, and/or to compare our mobile traffic with our desktop traffic and look for areas which are over- or underperforming. This allows a quick check-in on all the different reports at once.

You can create your own custom dashboard, using the data we’ve looked at above, or use these basic sample dashboards that I’ve put together:

Separate Mobile Site Tracking

Setting Up Tracking on a Separate Mobile Site

Many companies with a separate mobile site have not properly implemented their GA tracking code and are effectively losing the data from their mobile visitors. Mongoose Metrics did a recent study on the GA implementation of companies with separate mobile sites. Out of 75,000 websites, 41,344 had a separate mobile site and used GA tracking on their main website. Of these 41,344 sites using GA tracking, however, 37% aren’t tracking their mobile site at all!

Steps to implement GA on a separate mobile site

    1. Implement tracking code on mobile site pages (as you would with a desktop site).
    2. Treat this site as a subdomain and be sure to include it in the same account as your desktop site.Set up cross-domain tracking.
    3. Set up a separate profile which only tracks traffic to the mobile subdomain. You should have a ‘default’ profile which contains all of the data from both the main site and the subdomain, a mobile-only profile which is limited to data from the mobile subdomain, and a desktop-only profile which is limited to data from the main desktop site.

The important reports and metrics for a separate mobile site are roughly the same as for a single URL site. Keywords are extra important because you have more freedom to create unique mobile-specific content. Site speed is also very important.

A couple more:

  • Redirects: do a quick check of the ‘mobile’ overview report in your mobile-only profile…if you’re getting a lot of non-mobile traffic to a separate mobile site, something is wrong with the redirects. (The same is true in reverse on the main site analytics.)

You’ll find this in Audience > Mobile > Overview:

  • Comparing mobile site visits to desktop site visits: you can use ‘hostname’ as a secondary dimension in your site’s unfiltered profile if you want to see m. visits versus www. visits.
Mobile App Tracking

Google Analytics has recently released a feature called Mobile App Tracking. It is possible to integrate this data with your GA account with the new Universal Analytics, which has just been released into public beta.

The Mobile App Tracking implementation is a bit trickier to set up than web tracking: it requires the app developer’s input, using the Software Development Kit (SDK) for Android and iOS. The GA guidelines to best practice for setting up app tracking help you figure out whether to use multiple web properties to track your app.

Implementing the tracking code (2-part process):
  • Set up your new web property in the desired account: select “App”, and get your unique Tracking ID
  • Download the GA SDK for Android and/or iOS and add the code with your Tracking ID
Important metrics in mobile app tracking:
  • Acquisitions
    • installs and opens
    • Google Play sources
  • Users
  • Engagement
    • how visitors use your app (e.g. session length, crashes, screens viewed, behaviour, events, etc)
  • Outcomes
    • goals
    • ecommerce transactions

Questions we should be asking about a mobile app:

  • Where are your users coming from when they download your app in the Google Play store? (Acquisitions > App Marketplace > Google Play > Sources)

Note that this particular report only uses acquisition data from the Google Play store. If your app is not offered on GP there will be no data available.

  • Are they abandoning it, either after install or after opening? (Engagement > Behavior > New vs Returning, Loyalty, Recency)

These three reports (New vs Returning, Loyalty, Recency) will help you figure out whether users are continuing to use the app after installation.

New vs Returning report:

You can check this report to see whether you need to focus your attention on retaining current users or reaching out to new users.

Loyalty report:

In this example, the user engagement is high, with a majority of users returning over 200 times. So in this case, perhaps the focus should be on encouraging new visitors and new downloads, as the app itself is good enough to retain them once they’ve downloaded it.

 Recency report:


The report in this example shows that most users are using this app multiple times a day. The usefulness of the recency report, however, will depend on the purpose of the app. If it’s a social network, like Facebook, you’d expect visits multiple times a day. If it’s an app to find the nearest car dealership, you might not expect even the most dedicated customer to log in frequently. So remember that this report depends on context.

  • Is it crashing frequently? (Engagement > Crashes and Exceptions)

If your app is crashing frequently, you will lose users.

  • Is it performing better on certain devices? (Users > Devices and Network > Devices)

In this example, the iPhone sends 4 times as many visits as the iPad and iPod Touch, but also iPhone and iPad users are using the app for longer sessions than iPod users. These are discrepancies worth looking into.  

  • Are they converting better or worse than average/expected? (Outcomes > Goals > Overview; Outcomes > Ecommerce > Product Performance and/or Transactions)
  • Is there a problem with the conversion funnel? (Outcomes > Goals > Goal Screens; Outcomes > Ecommerce > Time to Purchase)


Well there you have it! Hopefully you’ve found this guide helpful.

Are there any other reports or metrics which you use for tracking mobile traffic? Or for making decisions about how your mobile-friendly site is performing? I’d love to hear about it in the comments!

Bridget Randolph

Bridget Randolph

Bridget joined Distilled in November 2012. An American born and bred (originally from a small Virginia town), Bridget came to the UK as a grad student in 2010, and didn't want to leave! In September 2012, she completed an M.Sc. in social...   read more

Get blog posts via email


  1. Michel Vennema

    Good info. Love the recency report! One question though. Why make the smartphone segment with screen resolution. I simply made in by including 'mobile (including tablet)=yes' and excluding 'tablet=yes'. Works fine.

    reply >
    • Bridget Randolph

      Hi Michel, glad you enjoyed the post! Great point about using exclude 'tablet' instead of Regex for the smartphone custom segment. That 'Tablet Traffic' segment is fairly new. One benefit to the screen resolution approach is that it allows for a bit more control over which devices are considered tablets (whereas I haven't yet been able to find out the criteria being used for the default 'Tablet' segment). But for most people, using the include 'mobile=yes' exclude 'tablet=yes' should get you the same result.

  2. Hi

    Fab post. I was searching how to compare mobile vs desktop in GA so this has been a God send, thanks so much. I am going to work my way through your suggestions concerning what to compare, thanks again!

    One thing, it seems GA now filters the tablet data as I applied the created mobile non-tablet and the default mobile segments, which returned exactly the same results. GA ignored the created segment in the graph.

    Damn Google, always fiddling.....

    reply >
    • Bridget Randolph

      Hi Jon, thanks, glad you found it helpful!

      That's very strange about the non-tablet vs default mobile segments, how did you set up the non-tablet one? Is it possible that you haven't had any significant tablet traffic? It's not impossible that GA is rolling out an updated version of the default mobile segment, as you suggest, but I'm just wondering...because when I read your comment I checked our analytics and I am still seeing a significant difference in our traffic numbers between the two segments.

      When you look at your 'Mobile > Devices' report with each of the segments separately, and then without any segments at all, does that seem to indicate that the default mobile segment is excluding tablet traffic? You've got me really curious now!

  3. up,up and away with your great blog post, very helpful to see your breakdown on what points to consider when looking at mobile traffic and conversions. Interesting move on the RegEx for screen size, I also just use what Michel suggested above but good to know there is a more precise method.

    reply >
  4. Great post Bridget - thanks!

    We are currently serverside redirecting from main site to mobile site based on device which seems to result in our Analytics reports being skewed - we're seeing a drop in traffic to the main site and a rise in direct traffic to mobile (because the redirect kicks in before the main site analytics tracking) - so it looks like organic traffic has dropped to main site and direct traffic has jumped on mobile.

    I'm sure we're not doing this the best way - do you have any tips please?

    Seems that by redirecting to mobile we're losing otherwise good data.

    reply >
    • Bridget Randolph

      Hi Rob,

      Thanks, glad you enjoyed it!

      It sounds like your referrer headers are getting stripped out before the analytics tracking gets to it. There's a couple possible explanations for why this might be happening:

      1) iOS6 and Android 4 operating systems were recently losing some referrer data because of the way they were automatically using encrypted search (I mentioned this earlier in the post). It's worth doublechecking your analytics to see if you're only losing the referrals for those devices, but seeing mobile organic traffic from other OS. If that's the problem, you can try using Annie Cushing's custom report:

      2) If that's not the problem, there may be something about the way your redirects are set up which is causing the server to strip the referrer data. So you'll want to dig into that a bit deeper.

      Hope that helps!

  5. Rob Holmes

    Thanks Bridget - it was not device specific (Option 1).

    I'm planning to fix this (Option 2) by sniffing the original referrer tags as part of the redirect script then building the original tags back in to the target URL on the mobile site to force the referrer details back in to the analytics script on the mobile pages. Currently experimenting but I'll let you know how it goes!

    reply >
  6. I solved the problem with javascript help. Wrote guide how to do it:

    reply >
  7. Valentina

    Hi Bridget! Your post is very interested because it's the first one where I've found some info about the referral issue on Android 4 (in addition to the arleady well-known on iOS6). On your post, you have explained that this is affecting visits from search engines: "search traffic is recorded as direct traffic". But I'm experiencing the same issue also on visits from other referral.
    The strange thing is that the problem doesn't occur if I use a non-native browser, for example Chrome or Firefox.
    Have you had the chance to delve deeper into the question? In your opinion, what I'm experiencing is a matter of native privacy settings on browser or server redirects? Thanks :)

    reply >
    • Bridget Randolph

      Hi Valentina, glad you found it helpful! I'm afraid I haven't gone much deeper with this issue myself, although there are quite a few good articles about it. If you're seeing the same issue for all mobile devices, my guess would be that it's a problem with the redirects. That said, since you're seeing it occur only with native browsers, and the problem isn't there when you use a non-native browser, I'm inclined to guess that it could be an issue with the native privacy settings. What devices/operating systems are you seeing it for?
      Let me know if you find out any more details about this issue.

  8. Valentina

    I'm seeing that using Android4 and iOS6.
    Tha strange thing, as I told you, is that that issue occurs also on referral not only on Natural search - Google. Have you experienced the same?

    reply >
  9. Hello Bridget..

    I have been reading your Post, but i have a question here:
    You say: "When we apply all three segments, we can see a graph which compares the difference between mobile, tablet, and desktop traffic"
    Excuse me for my ignorance...reading the post ( step by step ) i see 2 Segments:
    Visits Mobile - no tablets
    Visits non mobile traffic

    Bridget...where is the other one??? Visits tablets traffic?????
    Could you explain me how can i get it??

    Thanks for your help.

    Andres M.

    reply >
    • Bridget Randolph

      Hi Andres,

      No problem! The 'Tablet Traffic' segment is a default segment so you don't need to create it separately, you should be able to access it in the advanced segments list.

    • Hi Bridget...
      Thanks for answer me.
      You should write more here about Analytics.

  10. Great article and research.

    My toughts: modern smartphones all have a high resolution screen. All mobile traffic providers in my country sell their plans with high resolution hardware (tablets or smalrtphones)....

    Small resolution mobiles are allready HARD to get - and those which still are in the market (nokia c2 as example) are sold for the ones which dont want a monthly plan or simple use their mobile just for calls and sms (prepaid) .

    Bases on this observation i dont recommend anymore to build responsive designs except for ecommerce shops. Instead we build our sites for 1024 resolution (960 mobile screens)... and optimize for touchscreen navigation and use LARGE fonts!

    One version, BETTER conversion and the regular visitor can surf the layout which he allready knows (most surf mobile AND desktop!).

    So for me, the whole mobile discussion is outdated, latest in one or two years when all the 320pix mobiles are replaced!

    reply >
  11. Hello Bridget...

    Could you help me with a couple of questions?
    1. When you did ( Traffic Sources Report ) you changed Dimensions from Primary to Medium...i should make it each time that i want to compare Mobile Vs. Desktop??

    Bridget, On the Landing Page Report ( Bounce Rate ). On that chart i can see ( Red and Green Bars ). Could you explain me what does it means??? I mean.....-46,44% ( Greeen Bar )??? or 21,51% ( Red Bar )????
    Bridget, looking my own Reports show me two items:
    Not Set
    Not Provided

    What does ( Not Set ) Means????

    Thanks a lot for your help

    reply >
  12. Great article Bridget, although mobile segmentation has finally been updated on Analytics - smartphone and tablet are now treated separately.

    reply >
    • Bridget Randolph

      Thanks Chris, and thank you for sharing that link; I'm really glad that Google Analytics is finally recognising that there's a significant difference between the two! Now if only that knowledge wasn't being undermined by the changes in the new Adwords Enhanced Campaigns (which don't allow for as much granularity between mobile device types)...

  13. How can I see campaign reports for mobile app tracking?

    For my web property the acquisition menu includes reports for campaigns, traffic, channels and more. But for my mobile app google analytics only shows new users, app market place, and google play under the acquisition reports.

    The iOS SDK does include tracking campaigns. (see link below)

    Any idea how to get campaign, channel and traffic reports for mobile apps?

    reply >
  14. Excellent article! I believe Mobile SEO and marketing is the next big thing. Thanks for sharing.

    reply >

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>